<%BANNER%>

Design and Testing of a Self-Powered Wireless Hydrogen Sensing Platform

xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID E20101118_AAAAEG INGEST_TIME 2010-11-19T02:18:09Z PACKAGE UFE0014378_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES
FILE SIZE 5976 DFID F20101118_AACNUR ORIGIN DEPOSITOR PATH jun_j_Page_090thm.jpg GLOBAL false PRESERVATION BIT MESSAGE_DIGEST ALGORITHM MD5
748fcca18575454b2cb5fe80367825e5
SHA-1
d60d1b3a5e7813e3f9d3b2b75a79060289139055
23097 F20101118_AACNVF jun_j_Page_082.QC.jpg
11e581f3d50669ac969e507d4464d271
1d7e70f7065e52f90c52484f3f5e2d6e5b8fb835
75427 F20101118_AACNUS jun_j_Page_032.jpg
daa3deb99e5febf25c98ed1d5add55e5
f9c3d5bcfcb2a226ad6b1320b9853bbc3a3afc99
61331 F20101118_AACOBA jun_j_Page_023.jpg
cba835babb685499ae14abe6da88ec84
7fe3579cd2d25ea428e63e790ba07b1d98ce3e33
10387 F20101118_AACOAM jun_j_Page_002.jpg
da40226ef7b051c47b55e53a40c71453
19c5ddb04690dc13933db69bd6a4a81061ee2eba
2072 F20101118_AACNVG jun_j_Page_118.txt
72d964062863da19c3e451523089d215
cc47c18cb0e2fa8cf7203d47306d4fe8faaff42a
1053954 F20101118_AACNUT jun_j_Page_030.tif
c468d4230a6ac9bf4452f74d627d3dff
feb773d8a6acec6645c38835b962b099c083bd45
74114 F20101118_AACOBB jun_j_Page_024.jpg
fe04e612c2931d548c17195460a33f2f
bd186d0d6a3ef73f8e575b3038491a961385ab21
50389 F20101118_AACOAN jun_j_Page_003.jpg
befbe8a19601596fb4bebf02583e3af3
2484459eda22ddce1827d679f30fa0be8fa37280
25271604 F20101118_AACNVH jun_j_Page_074.tif
e60a9293f37c5691ee6345465f2049a4
90760ec8f18f654eb59c7119b45dd9600a4c8caa
69834 F20101118_AACOBC jun_j_Page_031.jpg
7daa40bd790b19b263486b8265ff4dc7
0be98b85c792194044c6cd2698e29912841340f9
15254 F20101118_AACOAO jun_j_Page_008.jpg
745ee9d1302a8d1bd5a7e56d356f6533
02d50ab9282712ef57065f2142e87b6d5707d27c
5836 F20101118_AACNVI jun_j_Page_027thm.jpg
665846645f9a99c48d14a645d6927d8f
9f4ada4e6f6c9dba1ebd4a05fdc67c5eec6d9bdf
F20101118_AACNUU jun_j_Page_106.tif
bc665122f54348d7ed25e3bea2891d53
13b9a5e4b7cf7bad8cead047987afe4210977a35
66886 F20101118_AACOBD jun_j_Page_033.jpg
8f8383888a5f7939cd645ab6f16980b6
84028a7d7232b57aa5981adf94ec2c4d7a87c062
73184 F20101118_AACOAP jun_j_Page_009.jpg
417fa517ab1ec3a7a7f8427fe018ab7c
d151e0f36979607c4bbbe1429ae41ae3c36e23e2
1461 F20101118_AACNVJ jun_j_Page_055.txt
3af57071e96ea8f932fd6105390d9b44
cb9e160b0c8ce4aac5e5875d74ba235e27f47a5f
5832 F20101118_AACNUV jun_j_Page_086thm.jpg
822ed85b3b945250254e4ec5389cbb8b
2657115524dfd8e17f95656661611b6cc82bb880
71865 F20101118_AACOBE jun_j_Page_034.jpg
651c59dba25b5fa4dde04ded43bf0987
1b185e0712b2d63e39067a003d52756b5aebbc9b
85956 F20101118_AACOAQ jun_j_Page_010.jpg
1c0fa79110c9ff4777e93f72dd78c1ad
73154b85264c22142bb1277f1fdaf3f61c6fe07a
18655 F20101118_AACNVK jun_j_Page_123.pro
b9fb024542ca2882827b0c530d159fa5
f9561eb0942a6962f833079dace85f1cc42455a4
57814 F20101118_AACNUW jun_j_Page_062.jpg
3c8e743f79ac0b55f3592073b079e85a
0007ce1a6c6134287987662607170da841439547
56326 F20101118_AACOBF jun_j_Page_035.jpg
0c7e1a07ffb882cb91b9979416ff9ea5
780fc81ff958a9eb38a16b85c4b4114508bde8e9
83930 F20101118_AACOAR jun_j_Page_011.jpg
7fbff189cac5796be7d5713a567ae880
f05946c37357bba06c1c9d322d481c758dccb2a6
4903 F20101118_AACNVL jun_j_Page_085thm.jpg
066d7f1e74e9adbd10ea690224dc29ad
66df1b3e03983cc28216ce727613249837c438ce
68370 F20101118_AACNUX jun_j_Page_044.jpg
27337a9627cc7cb2b8a7794add80e25e
011f3bfc854d891d23a448c2b6355158306cf33d
66874 F20101118_AACOBG jun_j_Page_041.jpg
9fb4fbce40fd877e80857915e8368ce9
ad0e7a6559a3012e217c2d80d36d036d1a57e4f3
47393 F20101118_AACNWA jun_j_Page_038.pro
586510b2cda77da90209e75df38b6e75
5be43f2543d80112c5fda65b77f198b917cd92b0
43351 F20101118_AACOAS jun_j_Page_012.jpg
5cbb6d703e648961ed3cbd2984e56cc9
2b368f065fa29a415732ab0379295d9fbbc66389
85867 F20101118_AACNVM jun_j_Page_104.jp2
fc7787ce8f14b808abe82e58a08183e2
ed34865a81fc5d0205f15099ff7de9b14ac5d37a
F20101118_AACNUY jun_j_Page_093.tif
2a7e08839e2f36bd51c139ece45ff7ff
9793eb14eeb55d29df57a7b0d9df46cde5da7372
67100 F20101118_AACOBH jun_j_Page_042.jpg
5264a15a6f199e449a2ef347b074a862
3b2ce7886c1568faa4eebf2ca4d0a25806904b42
69480 F20101118_AACNWB jun_j_Page_026.jpg
72d8ef0ddbab04397732a7b9f899c07f
9e3d96132cc09a024d8327e6b0c4b9e270aa1d88
54913 F20101118_AACOAT jun_j_Page_013.jpg
dbb6812e3ab6dc94dbbab240dfc87e65
bc8d008e0c0937296ba361d46bb1c1bf5774065d
6429 F20101118_AACNVN jun_j_Page_079thm.jpg
0e63e0df133a9e8c3c14af964aab554b
e372afee701c812d55dc3b14dd54f33cb814e41c
19609 F20101118_AACNUZ jun_j_Page_091.QC.jpg
9cfb71e73496ebad4c9d153f7c95a08a
cf028ca95948338e60f4874a6d300729e338e193
53053 F20101118_AACOBI jun_j_Page_046.jpg
0558c7c3e99b95e5e172fbd73a67dcae
e05757b16cc99e9625d4d2407a29f2a3613cdf44
1592 F20101118_AACNWC jun_j_Page_067.txt
0d5df7d08a8e5447cd813ba87fa2562b
ee9bc6654f32d8dba2b321f5a4a2d0b367c01b9b
22148 F20101118_AACOAU jun_j_Page_014.jpg
ba35667f466e7554faa9199f19aa5bb3
163b2dac554b6dd10e2aca3887a03119bb2508b7
17739 F20101118_AACNVO jun_j_Page_071.QC.jpg
42c9429e9233b0b79b49a20845afb924
5491ac6cd0793c748bf27174b0cb2c3a7724f6be
65487 F20101118_AACOBJ jun_j_Page_048.jpg
f81ecd031a9de13908614de329b68ebd
c87154ef2377b6594ae92c3087e24a0e780b221a
6290 F20101118_AACNWD jun_j_Page_111thm.jpg
408e5efb3a2cbeb5656cc54812ce52a6
3536eae5e96740c63ec14ba8070d6703c098f89d
63532 F20101118_AACOAV jun_j_Page_015.jpg
76f7539d922ffe89c88ccbd55af2a07e
2a6e522886b82e6d43de0eaebf9515a42c52b715
45843 F20101118_AACNVP jun_j_Page_108.pro
db6eafc10a328a8739dae0e798e8fe3a
9b1d1d4f0f424c17d812856c6412a527ab03f3e7
45268 F20101118_AACOBK jun_j_Page_049.jpg
a5801078c1f440d8768b56e53c9d7baa
49c5d642f77ebd962985a21c3ac06f5914ab160f
47001 F20101118_AACNWE jun_j_Page_103.pro
b58b87f45143b6f1d84fca7b7d4650f0
b64c2db48cb27cb27e792611fbcfff31c8cc5292
68109 F20101118_AACOAW jun_j_Page_016.jpg
5dab9df6e5af28387640abccbc058db2
4ba0a29ccea438f21f655eea331fd597fcafd5c2
2335 F20101118_AACNVQ jun_j_Page_094.txt
d0bd85a964d1f033d943d280b674a53e
96300f5c4595a21fc838810ee208e949fe81c67b
60692 F20101118_AACOBL jun_j_Page_051.jpg
70c833abd3a19db25df26cd57bc9e8f6
5d14d7fdc9f1fd48b2e09c883064e3cc6001ed80
F20101118_AACNWF jun_j_Page_045.tif
453849c86df90a41d5fed451fab661f5
44e035b51d5197c938c771551ba93ccdfe48e347
43284 F20101118_AACOAX jun_j_Page_017.jpg
214da8f9963d38bfff36667f41899a5f
88ab92e95c3afc0f1e4379c5b384266f2f666e4a
46425 F20101118_AACNVR jun_j_Page_105.pro
a57e0c21b39e4ac9ebfaa657cff22262
383a218a6a17598fbc56fea2661c5342d49ce4f0
74154 F20101118_AACOCA jun_j_Page_074.jpg
63787719ea2eac79ae6019ed7d6f98b1
90fa03683ab4a38d10a53678f70a002148d1b1c7
70285 F20101118_AACOBM jun_j_Page_053.jpg
24fb78bcd9515b2f163388be241fc30d
14e103f3aa77ecc613d3524998304cb48735d179
20023 F20101118_AACNWG jun_j_Page_080.QC.jpg
1a680ef8980a0bad7538b3b745592ef8
e9a448a2a74e785b714f2dbe7cfdc1419b43a0b3
67430 F20101118_AACOAY jun_j_Page_019.jpg
fbe9572d84940e7c352dd6f2ca00151c
467c5b4de8b8e2072b17f9741791a624c7f013c1
23286 F20101118_AACNVS jun_j_Page_034.QC.jpg
d6a3ce1d9c3bbe040a9f6da4d37e8218
7b75b5e7f3b2314d9307124440d02812279d1f72
66368 F20101118_AACOBN jun_j_Page_054.jpg
890b1dd9219b91cfa08c4cae8e2807e4
9dac8448b2b9941b69dbfeceeb7e6fd80e86c8c5
5723 F20101118_AACNWH jun_j_Page_029thm.jpg
a91a94ed710680d3368d42f00138c5c9
2ed688bb7c27038e0871dfb0ad00221c8c71a9d1
60055 F20101118_AACOAZ jun_j_Page_020.jpg
f8099328875b52f6fe865de96b0444ce
b77c38c0367d4390285f880bd5468643fa3d6251
66007 F20101118_AACNVT jun_j_Page_108.jpg
312b20b0296a1f31c72f4655947a7ba5
9ce939ccd8ddb65b8a7bfa96dbc666b437572d32
53979 F20101118_AACOCB jun_j_Page_075.jpg
e1424154dcb3bc086aac51c358f88f41
b096217ecabbbfbbc67cfb787d80174afd74c1cc
52286 F20101118_AACOBO jun_j_Page_055.jpg
d6ccaa4c7344d417f6a541882b54c23f
f65732e7ab6fa30188e8e356a8d2069d72c7bdf9
2127 F20101118_AACNWI jun_j_Page_065.txt
45c37b7fbdea27197ef07c3d8d398b92
a38d653796a91eb6d8eb4828b7e5c66a31757cd5
F20101118_AACNVU jun_j_Page_027.tif
45e2a8d2623784697ff2820a3de18b2c
54777ba4030c81309eb353cae3974c1d1ec7d97c
60322 F20101118_AACOCC jun_j_Page_078.jpg
8e8a51c62b73ebeada3fd97baafecb1b
8731fd5d91619a608e309421b575b468398b9b1d
58762 F20101118_AACOBP jun_j_Page_057.jpg
dd4b99937b883a27584de82507a99053
2fb07c56e024276a178c915d3006b0e8c2e4bd34
915269 F20101118_AACNWJ jun_j_Page_081.jp2
b6f96a647b3f8156151e906497767ccd
416a1a56d896537630f0de0e226e624dc909a895
74167 F20101118_AACOCD jun_j_Page_079.jpg
07f6629bfdbdfb9b9843edf6da5107de
d6f9e799c73368be466ffe4caf8a7b8ba1187a8c
70563 F20101118_AACOBQ jun_j_Page_058.jpg
2777161681a9861bd8c49e83e743c25e
7a926f68e23b26a8a5b016b3f4eb7f3619d0ac28
1589 F20101118_AACNWK jun_j_Page_070.txt
702504c2344c789ec3ce80c2c40cd173
5a357b476de53b95a420aee573d9b69becd374b1
96706 F20101118_AACNVV jun_j_Page_064.jp2
e62f6edbc0030f71852991ca4e1f66cd
8db39129d65645204b4856011979fbb6da318a96
62919 F20101118_AACOCE jun_j_Page_080.jpg
5475d901ef065f27ace163227c93d171
4e5800bcb99be04f6605cc7b542184e3a0f2829c
40579 F20101118_AACOBR jun_j_Page_059.jpg
f492a10559ef1b1f1035ae9dc4bdc48a
369c2325dbc340a28b36fe4dc7a2213fa3e79708
F20101118_AACNWL jun_j_Page_107.tif
c5a99d0b834e862e58db04fe6dca2a09
108bc1c9b43cf541e2ff74a7c07bbf229cc4a8f0
1051976 F20101118_AACNVW jun_j_Page_120.jp2
c37f7210d6920eba018451a43469e997
6df2e35cc68781dbaaefe54f06e785c5ea786eca
66833 F20101118_AACOCF jun_j_Page_083.jpg
d5beda4b4da6203d0d313d35413a5ab2
823ffecb632dc95a3e56c5fadaaf2e9215519dbd
23755 F20101118_AACNXA jun_j_Page_040.pro
11d899da78e2fa0381bb9ac0bb35c83d
41142b534c769b842e90858dfd3c1dad13901f24
73815 F20101118_AACOBS jun_j_Page_061.jpg
ec83cbb12b95f660570e644d49c01be0
cfe52ac0494f8c57e2d9d90a1c004bd7498eaf7a
F20101118_AACNWM jun_j_Page_006.tif
4d24276f4c290211f3054635a56268d3
57fb63e1ddce8e336eee7ae5a3a02fb4abdd568d
18293 F20101118_AACNVX jun_j_Page_052.QC.jpg
4ccbeb6f947622e798711f3332bb725b
e115665c8289a6bc013cfdaf76294167b706349a
52071 F20101118_AACOCG jun_j_Page_085.jpg
0e554502fc275f7a3b1df499f3cee8b9
fcc3d5117ba53e62df751dc1d42002705100253f
1752 F20101118_AACNXB jun_j_Page_060.txt
8a57a7f2051d311886840629d74c3f3f
1bcc299a134d67a9f4f47cf85f2196c0bb166a09
40780 F20101118_AACOBT jun_j_Page_063.jpg
6bff3bfd77dfbfb8fb5a80c4a9771e70
77eba08f05f8a83583e7e1447905a7238b6b44fb
5821 F20101118_AACNWN jun_j_Page_112thm.jpg
d7576f3489bd657a0fe0a9db8d820238
bdd94d3641309a2f65ab541f1bc27ff90da69996
60605 F20101118_AACNVY jun_j_Page_119.pro
85961348c37a407c44d68f6e4ccf0673
9e874c0e20ba9836b17c6720dac22b574a44f88f
61395 F20101118_AACOCH jun_j_Page_086.jpg
17487e09cf0679175ab7f6a43cb77cbe
42c8a21bdf46d2cdf2ed0ff081f1baba63e5cb09
34414 F20101118_AACNXC jun_j_Page_028.pro
781dc688ff4c9b310ad9f709610fb1a2
e555dbfc04b265d6735451e2741298a83529c234
64225 F20101118_AACOBU jun_j_Page_066.jpg
1ff67bd8773c0a42488d62a7eb10f248
113ed4ffb2e8f7900488b718a7631ef5fd8f069d
19279 F20101118_AACNWO jun_j_Page_073.QC.jpg
7eff80bc2d4d90fc40ccf3539fb4e985
01784de8e2fd75ca7f98bbfda08dcf97e75d4356
5949 F20101118_AACNVZ jun_j_Page_018thm.jpg
158919c56b4440b54da761e7e232ca93
c3e3216d471e7bf30aa7a1f6538add159073c4c3
71483 F20101118_AACOCI jun_j_Page_087.jpg
bb45798d763ed98723e160fbe482360b
90edc4d17896c4b27d728efc068733ed65766cd2
41999 F20101118_AACNXD jun_j_Page_018.pro
ca2128794f01102cd6e8ad92dd024902
216e4b89a51ce4a4d590c5eaad8c2fd253951e72
57122 F20101118_AACOBV jun_j_Page_067.jpg
634570407c442390a30595c9ea6a1a41
89b8042ee12812bf53296c109cb9cb17614f8d8a
46211 F20101118_AACNWP jun_j_Page_058.pro
244f323e0fd6dd65ab06165d423c0fb2
0df5e9cd362ceabecd6b6dc7cbad56f6f7b8ab53
61114 F20101118_AACOCJ jun_j_Page_088.jpg
2f632105b7419a3f05a2bfabb7bf853d
e2ad7503c5b55703f9638214e993ccebd3456388
F20101118_AACNXE jun_j_Page_100.tif
4bd3df898c79a5f315f8e1b43ef81edd
546d6817fdf7daefaad6f55e585cc5bd3b4ce77b
61022 F20101118_AACOBW jun_j_Page_070.jpg
0cb48d5c1edd0b8025a2971605140fea
60ff240a9262cb70fb53f7bb8dc8d28c2e6ec693
94171 F20101118_AACNWQ jun_j_Page_027.jp2
daaba6386ead14259bd831cb8152f5ef
a8f6b5afe039f2ffef06d1b602863bd821fce98c
56611 F20101118_AACOCK jun_j_Page_089.jpg
647fce7c1d5dee138358d23e67bbb463
eb04897bab807c29a81683e2bd2e28a2cc091bc0
869707 F20101118_AACNXF jun_j_Page_109.jp2
c8b00f64e4b3331881f6deba8542e195
5bdf8c22f01e86d2cc3875bfabbb84020b8882ff
55902 F20101118_AACOBX jun_j_Page_071.jpg
9b17e1ab342fa699311f46947603f27e
7f2cd1ef96a25ce158b59a0b57b017308bc6930a
F20101118_AACNWR jun_j_Page_099.tif
eafced61c2fd900b0ec933b4bb72bb0c
876700c82eef1723b06745c608e39cb047b5ff86
71467 F20101118_AACODA jun_j_Page_116.jpg
5c8e6c657c80eff7691897e94bb9727c
78581140ab70866951356f260b8f8195c84c9455
73197 F20101118_AACOCL jun_j_Page_090.jpg
0876434bdd32264e0f72571fca76c930
bf5719a8ab457d2e4da9a8d21be5cbafc2dc80a8
70651 F20101118_AACNXG jun_j_Page_082.jpg
412c55824c169d9afd5dcb22d00d6331
42c5f81f6e9805496e949ee92841949ae8bd5442
46505 F20101118_AACOBY jun_j_Page_072.jpg
d930f5cbe298da6500428208663141f5
ae6694055bf417fe1b04900a2a50ea4c8d3d8de1
85742 F20101118_AACNWS jun_j_Page_029.jp2
136318653d401e691f731f80dd7c2af4
8cc6fa827bc713dbe05b45512b3c5634662e2465
106954 F20101118_AACODB jun_j_Page_119.jpg
abd4a5ec0468c529940c469423659c06
40a1668f2c48c82b5ac2ad1341bdc59eb1585bf9
61141 F20101118_AACOCM jun_j_Page_091.jpg
95c051859411fd388b3a033626b48c61
d3ed4a524933e7ba5029d1e7315bdd150a6e3b2a
1624 F20101118_AACNXH jun_j_Page_008thm.jpg
824135a5744c9f813ccbfaebf8b61868
62d479054bd485435267c14cd7c6fef0e9c3b09a
65556 F20101118_AACOBZ jun_j_Page_073.jpg
5067d0eca6ec4e4d58811bd69849d96b
52c33ce982d9f75fbf55d15bf04244b2f33fe1ec
20911 F20101118_AACNWT jun_j_Page_114.QC.jpg
be8fb6aef359820c1f804c881052809c
85ef736674411af347f0b6e1c39f8c47fbb47fe8
72715 F20101118_AACOCN jun_j_Page_093.jpg
64c714d751a8778d12f9f5e4febfd3e9
89888b8966643ff7b2ae053c264ba00dffe1aa24
66395 F20101118_AACNXI jun_j_Page_056.jpg
a9657190663717ba88c125e21f3003f9
2dfe771eb4a5526837dd5a3ce7f103b1f08b4f13
54410 F20101118_AACNWU jun_j_Page_007.pro
39f2c7d106875c5deba4b1b777513a1e
7513af3a7eaacaea5452f6478dd3e497771e7a4b
102128 F20101118_AACODC jun_j_Page_121.jpg
f7669b7db90adfa77c1a7da9748455a7
6c0e62f6e43d5ac28c0c0f2f71ac87d54b7fb3b1
92802 F20101118_AACOCO jun_j_Page_094.jpg
f31e89de0f50208a8877facb54a8c103
bc80e99f07229e58a2c9141347993a68757c348a
F20101118_AACNXJ jun_j_Page_116.tif
798e3ff5057b3205b383fa85d7e4e6e9
71837e4f168c4ea018fc624573334e374e4c6f4d
14421 F20101118_AACNWV jun_j_Page_017.QC.jpg
2f352e916e06c1599b35ac259f3aaf12
aa3cf9d5aff74f622b778df731de1b1ad26ed003
32614 F20101118_AACODD jun_j_Page_123.jpg
4b29536810c1f3fef161a6aee65003f1
a4db6dd6f253aeeff899d687da7aeb6c1df6054d
67617 F20101118_AACOCP jun_j_Page_097.jpg
bf2d749264a32bbae6e2cb1a5c3043bc
7788ff03065eaa5aecea5c86d6d72b0bb5556af2
1783 F20101118_AACNXK jun_j_Page_079.txt
62782914c5cb86286c4c1736ca5fab3c
00c3cb400f289730e6b438c86c82baf47282ae2d
23115 F20101118_AACODE jun_j_Page_001.jp2
c1cf83f0b20d073b3e81ae0abf4c49bf
e50055717e5560fdbf807d9db0e553fccf39dd16
63910 F20101118_AACOCQ jun_j_Page_099.jpg
ae8631ede60a1b909c77de4442cc7e32
e4bc3e2ca97405fe41d88b94f0342cff916cca77
56650 F20101118_AACNXL jun_j_Page_093.pro
aca31f7264c673573afcd84c33a7ca55
73d8aa045280c0fc7dcf1052ed9a26faa9941ff3
33395 F20101118_AACNWW jun_j_Page_090.pro
15e114afe27b743276652bbc4db90268
b4b5d8baed6725079b920705ebe21a7e23c60ede
1051971 F20101118_AACODF jun_j_Page_004.jp2
cd936f70efcd358d8b80689cebadc013
2bd3e65b318a412008470dd4399dc9b9a424f123
65222 F20101118_AACOCR jun_j_Page_101.jpg
79cd48d3f95cb0ee502adb2330e2a9cb
aec8a849cdec9afefd228c5414d74bb3158fe703
76079 F20101118_AACNXM jun_j_Page_122.jpg
a3c6567ecc463d65d071ca3d01825f20
c2755961799b0d8ecab3a28bae92765ef2a2e84c
582394 F20101118_AACNWX jun_j_Page_049.jp2
e083692175044365f4ffd5dcaeb10002
cd896b6263a3498400c72b71e810dade9998dca2
1051983 F20101118_AACODG jun_j_Page_005.jp2
519d48b1361a63be9a9ed318614633df
fe031d76fb6addbd920d711411fb93b2d8d7e6da
22887 F20101118_AACNYA jun_j_Page_077.QC.jpg
c7c0624d34f122821594ac5ee95d8bc3
ad51515363b9ed97f5e8ee1c2e4248db2f71dc2b
77357 F20101118_AACOCS jun_j_Page_102.jpg
0d613b99ffe8d9d52b4a7239bc0c449a
bf4f54a4db4d9a87418b1ee468f116b8554daef0
51685 F20101118_AACNXN jun_j_Page_040.jpg
4b2638d43934048ed2c5c89346df9c38
2266d752e54b8f77202f5c30099ad8f1d74b23de
68538 F20101118_AACNWY jun_j_Page_115.jpg
54d37aa465037637a82839641b5d6154
da067729db2f10b2729254267baffe9d156fbed8
413869 F20101118_AACODH jun_j_Page_006.jp2
db607b03aebd475909b1448ba2dac9b1
c07600a9ba7b663ed2a7b70d9d9b5a6b68489f6c
798947 F20101118_AACNYB jun_j_Page_062.jp2
b883d064d1fa13f85be4026fceb952fe
a6bdbda48a5c07b6a2b831bf38c60ce7f19e5df7
57297 F20101118_AACOCT jun_j_Page_104.jpg
f8f244a865bd3721dbe445c4b0978ca2
5895aec5634096bbadcd3058c3cf809c7e21eb05
14923 F20101118_AACNXO jun_j_Page_049.QC.jpg
972f0bca2b3d638502084c12573071df
4efe6bb847180741bdcf7de9100ccdc47b48eb0d
F20101118_AACNWZ jun_j_Page_017.tif
0dec9d50c9fe7f56e4417fe4b8921179
06989be8dfbabd79d6f1ffa2af441a4ab1f0d78a
1051981 F20101118_AACODI jun_j_Page_007.jp2
b2a7905695f2aec4c08ddcbb056459d5
2e426f88a1fdd770f2853d58d467467e71df7806
19900 F20101118_AACNYC jun_j_Page_070.QC.jpg
1f6a98b79248770859b93c95d3ffb15c
5ad975d03000b5ccb0bff37770259b87411a6b1e
59194 F20101118_AACOCU jun_j_Page_106.jpg
67f43dc0b8cc1c8f7508a6700c7846c5
5c8238a057613bb2b097ae2087c143ad04afbe11
26238 F20101118_AACNXP jun_j_Page_052.pro
0200b58f6f392635ab700a5ed0ed177f
d346bcf2e776abfad930de82b390743d5874ff55
210478 F20101118_AACODJ jun_j_Page_008.jp2
344bebef36ae00453674143c1830b7ee
a96a366f155fcdb07d5a7ba4766cde1dd0ccb3bf
F20101118_AACNYD jun_j_Page_122.jp2
439d01973d5e464c52a2695d37e169ab
6d1fa30e88873e468a946d777fb0d032baae37c7
28533 F20101118_AACOCV jun_j_Page_107.jpg
3896767ba6a8f7ec91a61899eb15c4d9
d1f4e403b1db66c398cf9925c058a2f31e6bc225
72194 F20101118_AACNXQ jun_j_Page_077.jpg
528e14e9265edf438ba4d55fd3a7d0f0
6445e30a385ec6aa557c7b0cbf922787fff02b73
F20101118_AACODK jun_j_Page_010.jp2
ec1f594bd0035623be1b9ba8a03d6cff
1537d86b43d1c2d316fe0212e570905d7aa893e0
72093 F20101118_AACNYE jun_j_Page_098.jpg
adc0ee0faf30cd57ca66e6fb1f863740
820c26a4e5d4d4fb1695cbc0edda0fd2780dc9a1
56848 F20101118_AACOCW jun_j_Page_109.jpg
1696813b0140601fcacf44f90de88c2e
b3a32a0097f76fa36b0b67a31b0220e519f4dd11
82500 F20101118_AACNXR jun_j_Page_069.jpg
90c09577e8e6eb376cd1955086535c24
7ab73cfba90e59c6016bac989e3befca22dd29ab
895081 F20101118_AACOEA jun_j_Page_043.jp2
35575258eeeb173f2b573f038739f56c
ecdcdf5acadccf1a74d4d1a4c19ed94a7252ddf6
F20101118_AACODL jun_j_Page_011.jp2
9041d851a9a81108799b5442aaa779be
8500b280f9c00c4003153bdc993ed648ee2b0dd4
34998 F20101118_AACNYF jun_j_Page_042.pro
b74ca414248448863781af1df8741cc8
1391d8abc48dfe633d037092621eecf39e97008d
66280 F20101118_AACOCX jun_j_Page_110.jpg
292c383fca201c2f585263af2413171e
8e5cc93346ff7650fa74494f7dbd186054c8702a
20253 F20101118_AACNXS jun_j_Page_015.QC.jpg
cb687af73e7a0224e44c45ef8161621c
64641a9931d99fb2e8314d27c5de0fec03f46f54
870708 F20101118_AACOEB jun_j_Page_044.jp2
bd574a19d76cc5fe664d8a7652d46dd4
763e1d52916bdc2906bf6dd98a011f521a056f69
1051961 F20101118_AACODM jun_j_Page_012.jp2
a02e5c3a357596ab0f3d1145ee4b69df
bf0e9ee74c711a2df896f0876034974ae1af87d0
19535 F20101118_AACNYG jun_j_Page_021.QC.jpg
1b60070069c261388417838d674e8eda
cb835d3675faf16565e6d6a2d6517275113c7ead
67852 F20101118_AACOCY jun_j_Page_111.jpg
0e6b10a47673a3a75a6cc439eb0cb263
3da5c7110e5a235b4b393bd19855c27f16d1408c
77847 F20101118_AACNXT jun_j_Page_105.jpg
886de8be13c9fb65eaf113e6a2f9dbd0
ef2ba66d894007a5bc79732476318839efacc831
101928 F20101118_AACOEC jun_j_Page_047.jp2
71c93a660f8ff1337d1c1bb57f1699ec
37d63b6ff97a535f28fb23a1644a1ef830fc6f63
78889 F20101118_AACODN jun_j_Page_013.jp2
e613fe6dbb4fcbec077272c8a3a1aae3
2ea5d958c97f8d49f2175d88460bcfab38be247d
74131 F20101118_AACNYH jun_j_Page_052.jp2
4bbb11d11928cfee4f3d26cb3ff7c59c
ec07fa158a521ad0aedce3e3942ae388a0e6239e
64305 F20101118_AACOCZ jun_j_Page_114.jpg
d71c3ba6b0c19d15f767d521b39fe964
e6619cf13655faf8c8522c510c1285992589d8a8
93909 F20101118_AACODO jun_j_Page_015.jp2
de0487f2e29a912d2cfbb2149242cfe4
5fd9388b64fd7befc60ccb752553ed9914d2d504
17360 F20101118_AACNYI jun_j_Page_006.pro
b43ced55b2c3eb95a49aa54ebf0e7a25
93d9b4e161a11851e3576ae3deedf344ecc6787a
3862 F20101118_AACNXU jun_j_Page_063thm.jpg
dc2ff4b805811dcfabdad8469a60db9e
6a3b4a2bb6a88298e7a1dc89ed21dd89457820cb
483067 F20101118_AACOED jun_j_Page_050.jp2
66f512cdee7d02dce6d9f3b8246e100d
55f2d6019dd8683b894109c1c7ae174a90f294b0
92271 F20101118_AACODP jun_j_Page_016.jp2
be08a89c99c47489c2175111e530c84b
69720d78476b3a708c9ba037a51e43cd6b836516
1139 F20101118_AACNYJ jun_j_Page_017.txt
e0fa99c95615785fe7e9d15420572fbf
4ca12323f7aed0e7fcb3b900c0ab7bb8ddf0a4a4
20431 F20101118_AACNXV jun_j_Page_066.QC.jpg
6e08581deb15af6cc994155aa8fc9a69
7223d5022a1559c735662a8d721696e94977e5b8
95680 F20101118_AACOEE jun_j_Page_053.jp2
a01a03ff8e9fd19594daaf8a21c12269
dc2b9934bd2187787736fbc16c5bf732cfffd80a
1051889 F20101118_AACODQ jun_j_Page_020.jp2
6b59de8ee6b57c74d6f7f96370bef747
38825ec029107728ee743131eae1f5aaf5f654f3
1871 F20101118_AACNYK jun_j_Page_097.txt
af7aff1313a92909c93aa3857459fb2e
2add3cfb30991ada9a08ec2c620ab46ae5956f9a
6507 F20101118_AACNXW jun_j_Page_026thm.jpg
496d2a1ada62920992e70bd36a535be8
92d49f2c5ca35bac33e862f337f22569a030188a
890151 F20101118_AACOEF jun_j_Page_055.jp2
f140e82f5a689d9703c45a639d056a59
4c8daf08f9952cabe6a62eba7cd138a93c62a26c
1029004 F20101118_AACODR jun_j_Page_021.jp2
b51277bbb73155aa1d397d10165b9803
6b644997331e602ed217cf774f7b2e500a356bcd
64681 F20101118_AACNYL jun_j_Page_076.jpg
65a4b218f1eacb854fd8667458ba6923
0ff6b079c0a96788eb8c03e7228b290b0a5925d0
928359 F20101118_AACOEG jun_j_Page_057.jp2
55f90cb6c2c88a0dce9dae6d5e5f9912
11e8d89d32ea91c8286f7b904b5e5e90e65a55e1
29642 F20101118_AACNZA jun_j_Page_021.pro
6b7d30e028ece7df9d90ce6a9588e2cd
e8f3864c56387013170aae661af6cbe4f96ef0d0
831995 F20101118_AACODS jun_j_Page_023.jp2
fc7c545b85797a8adcd130ce931b07b2
9df74bf61cc86646f2ccd9e5840d4c004f3b545f
23745 F20101118_AACNYM jun_j_Page_098.QC.jpg
80f53c356cc957b01754928aea2a4f66
f1e97eb9cf0fb3593c9cbd1a678057df6da54ceb
F20101118_AACNXX jun_j_Page_070.tif
af36c12f8dff81cd21e4c436e5d4ceb9
6d13443655aa27290876445b364f6dce928c1d0f
98220 F20101118_AACOEH jun_j_Page_058.jp2
e036bb693ac30a120ce99e67ca894349
d1f50c92e23d83d2eea09a71d2c4afdb5edbdf6b
42618 F20101118_AACNZB jun_j_Page_015.pro
5bd92c18453fbb03f92abd1d5b022bbd
e955553ec64917533540147f5329239b6fe0e952
64542 F20101118_AACODT jun_j_Page_025.jp2
f0bf8bf9fc7616ca018b38fffd2e9936
0714b218187213b461b68e65315b236cf748dca3
1800 F20101118_AACNYN jun_j_Page_015.txt
40e33ae16608aba8e78ab27b803863ac
6ba241d26c5b8a287b0c8c319690ab1a43d8fbab
23954 F20101118_AACNXY jun_j_Page_053.QC.jpg
c79e51ef8dd2ed71c4c0754a20bb3c68
cdfeb1b641a26a85b9381a59c45a72149b876ee6
537332 F20101118_AACOEI jun_j_Page_059.jp2
05bf02bc28486bf68f5be1007e17af02
854bba929453ae2072e8b9233b045a90f5b6d514
6374 F20101118_AACNZC jun_j_Page_110thm.jpg
5443ef7de84d45b2a514a5873abf9fea
4b49b7df72fc70f1fbf8a61933b4d5e69d54694e
78078 F20101118_AACODU jun_j_Page_030.jp2
3e94497c2d0874b3a184b420b2548997
3e64b9fec26918ae39128a24255f85073436d661
104182 F20101118_AACNYO jun_j_Page_120.jpg
305762fba558e165b67881d7151a87df
6530b676af6ce9400c8a30b8c4f1a296247baeda
979764 F20101118_AACNXZ jun_j_Page_072.jp2
887575e89ad3d3f4a019e86ddc76ae08
b165ff2eaeff472653e3c5606f21df1f22acbf3b
508368 F20101118_AACOEJ jun_j_Page_063.jp2
b8b491622624bcb9e4d250692d710499
3b87684136ee78916778e6d1d3f09e0ea0152d31
1811 F20101118_AACNZD jun_j_Page_096.txt
47456b248e7b5169e173672956697b8f
ae1f3eef54143b47b05546f5f7cad8382557c4f9
F20101118_AACODV jun_j_Page_032.jp2
732c1b07a553537a7734d4c292dd6e23
2a52674edced8d48a34bae3379368069752e7f91
19551 F20101118_AACNYP jun_j_Page_086.QC.jpg
054b0c4dfdeab790acccea2cd48d91ff
5143642b0ac629686dfd6d32e3c59831d62564d0
1017812 F20101118_AACOEK jun_j_Page_065.jp2
6c7082e57c70c32440e2ab941b58a771
844a6e74bf5e3a2d354f4f679975c2a9a88d83c0
F20101118_AACNZE jun_j_Page_018.tif
b9c6706ac141fdabba448435e05d04d9
49609e0d663e03cc91abd98f3cbc04bf928039d1
101215 F20101118_AACODW jun_j_Page_033.jp2
5dd563f452497d1fc9bd2560ab3aa325
286153bcd83a2324cee129c386eeca3e4e7d692b
5262 F20101118_AACNYQ jun_j_Page_025thm.jpg
a2c3ca4c2b57b606cb0070ccb6057bd1
490820809735c871aa634485930be7c8a7da4414
862219 F20101118_AACOEL jun_j_Page_066.jp2
ab8dc7c7a9ae2f99dae6a4951c54d6d2
fcf46ac1291b839148418a04647e209587c8d3ed
F20101118_AACNZF jun_j_Page_050.tif
cc5a3629bf295f540b6b8c7431302fee
0d4e3e5d6d95d0c47eceef1e2daad6a3b0bb5615
107613 F20101118_AACODX jun_j_Page_034.jp2
718fb5bc46d43939a5f1f75b7bba3e9e
976b0dcca6eea8109ee3db4a97830dd4e37d0eeb
24171 F20101118_AACNYR jun_j_Page_010.QC.jpg
bdf2675772f811a448db80ebb95bf3cb
1a7fc102e14f02922100d2dd4c3420240c41f383
98208 F20101118_AACOFA jun_j_Page_087.jp2
5eeb2dccfe15e3e2aafdedd525cfcaac
7cce8d2225823669678cdcfad52ab19995e30d51
781452 F20101118_AACOEM jun_j_Page_067.jp2
2315e5dc8d7a4dc4dd2c8e751aab4d2d
66633e644f808d6ff16d497574dd93c106707162
1674 F20101118_AACNZG jun_j_Page_081.txt
f37705b9d530ca9ab9d3536f198ad57f
824029c5a20482c8731f6392702a22175bc12669
107185 F20101118_AACODY jun_j_Page_038.jp2
f707c42b20554877f3631f1b8cbceb1c
d508cfea5df8fd042bc1932da49e0b66708acc09
1535 F20101118_AACNYS jun_j_Page_021.txt
89359ea88418cf159ac349aef1e5854c
962da2895f5c2f2b3de3c62c789fe10ecec2474a
866719 F20101118_AACOFB jun_j_Page_090.jp2
5d8441bb49bd4457aa5d61cb53c0f8e4
57535d4f4f3c5cc063c7a73da84ce91ebc36a87e
104354 F20101118_AACOEN jun_j_Page_068.jp2
ffe4cbcd851dd500d27f20e7da9f2f20
3b8a11a136e685d87faadda30c9eb46d91954a2b
F20101118_AACNZH jun_j_Page_040.tif
c96fdcf4149b7ea0fab206b24c92d140
44a67a7853dfc2bd94fc75660c3c08c34d4c739a
1051913 F20101118_AACODZ jun_j_Page_039.jp2
d6a83f65bb8b4b7219c56f7171f7b956
5d519e45799eaf2d110498270e0937f037ecb0ed
112556 F20101118_AACNYT jun_j_Page_024.jp2
6b8d9ae528ae7f8e984f141fe8a0a1d0
91776ab817b4fdf50ef3bbbcb65dfe6eae418bbb
816974 F20101118_AACOFC jun_j_Page_091.jp2
d866cbc26047b3cc5c92b571ddecddf4
1605dfe440a88c18b32ab8eb5579b5733a6795a9
109342 F20101118_AACOEO jun_j_Page_069.jp2
62485d73840621ca4a7ee302dd23ae9a
465d5937e14a42cc75d398d1da29f86e720f8e27
463 F20101118_AACNZI jun_j_Page_050.txt
92ff94e0739e3d0b173e5580ddaa4732
d91913e58a8692e9c8e2b717b747f12523f9d9a0
68778 F20101118_AACNYU jun_j_Page_068.jpg
30346e99463ef7df56d6d886e3e01297
7a5e999276ccbf2554c5e48301fccf5d79d95646
84721 F20101118_AACOFD jun_j_Page_092.jp2
97f057a26993a70eab1b298a6e28004a
139bef999cfb9dd7367d0b3761181881bcf8ff95
1051882 F20101118_AACOEP jun_j_Page_071.jp2
8108a5c539890054717f645abd0413ac
c10ae44e9beb5681218ecb852c68bc28423cf37c
861356 F20101118_AACNZJ jun_j_Page_106.jp2
df461cf3e4b5b55ed34048f17b9692d6
220de298d653a405ef4af7abd6d1cc676113fd5d
4278 F20101118_AACNYV jun_j_Page_059thm.jpg
66cfe035b5220189b72f3774bd9017d9
27c5b659cd4f33957c378476ea44fd60bafb452f
1051984 F20101118_AACOEQ jun_j_Page_074.jp2
3cd7fd29c496ede1d0dba8fac18d0f61
e598a135bf9231c84e3220bdff4462f103c7c670
50967 F20101118_AACNZK jun_j_Page_025.jpg
2994d5f3961071b5fa4c77ea5de1ad09
4e226a904e79767b0ff26ca8306748079963dc0e
F20101118_AACNYW jun_j_Page_008.tif
e8ef484835fe952fa7a8ce5dd7a449f9
beb75ec48fc7e23f7e506b1c3000c6c7557afd2f
F20101118_AACOFE jun_j_Page_093.jp2
5c576cbd83872e6dce000d42aa15c1e4
8a93c1e02f71445fc25d91f687b2ecfdf19e5176
F20101118_AACNZL jun_j_Page_002.tif
f6b28753604cf1025c45ffbddad20d3a
a10e7ec1cabaa36d12fb045151396aee478e3730
6420 F20101118_AACNYX jun_j_Page_006.QC.jpg
deb3398726f662edbde97d2ebc6e55b8
95c4c377b653b3403865ba819f157372a0ca0719
884073 F20101118_AACOFF jun_j_Page_095.jp2
bc2ae2aeb888eaf46d260f9cb26c1c10
c1d9cf79a0bf060a42e58b5c1a57066d029fec30
704793 F20101118_AACOER jun_j_Page_075.jp2
3b6f80d4828869707a7fcf6e8c45c544
2860492d1f77d28c2e966109d1914bf11e57dfef
103488 F20101118_AACNZM jun_j_Page_045.jp2
ef3c15acf83a2dc97d873e9c40dd7407
54e400b54bb0756fbc49eec0c50669509cb471e9
732277 F20101118_AACOFG jun_j_Page_096.jp2
a6238ce69f2e81a2728b23656a0d5042
3dd0c3b5688361cc594fe5e19c5b519b5dd90d8e
864201 F20101118_AACOES jun_j_Page_076.jp2
3a43e0cc4d19ce910691d79b479744e7
e0bcb0347f4644126ab449beb81f85658185e762
6198 F20101118_AACNZN jun_j_Page_039thm.jpg
68a2a5c3e476cff091982d7b5ac7d469
09730c52fcc5d1eec2a7b30f0e9329ad2d5d3933
43558 F20101118_AACNYY jun_j_Page_122.pro
f2cb981abe0081b95ba834d00e3f8a2d
f341b6b59a9521cddcffbdb24cd5f723df2f80ac
100160 F20101118_AACOFH jun_j_Page_097.jp2
7b5ea56c9d2d09eeeeb596b8fc0773d3
f65fe74cb4d7747a85eb1c5714146501c8bda21a
1051970 F20101118_AACOET jun_j_Page_077.jp2
4502c34a02eb428aa5ccd507c0f44fb3
525af3efa6cd211b24fb8f3022256abe5f906dd7
F20101118_AACNZO jun_j_Page_120.tif
3a8cee022161b642610d0f0103d01e97
a2d408775fbe46f0f81dfb267f70b70312130278
1019983 F20101118_AACNYZ jun_j_Page_022.jp2
aa9c15703e860af893366844510c5765
ab6e9df1a0d0e16f3f7dfeac1de7e37b097835c4
108981 F20101118_AACOFI jun_j_Page_098.jp2
6da7729c3277fb67caeb2b48ddcf7b2b
bf1e131594e369bf990438e21cf5d544191e7491
777498 F20101118_AACOEU jun_j_Page_078.jp2
443fcb408550165bea9d77968f04adb6
e8d897d850bc981f819ab512260a212c055c0363
93212 F20101118_AACNZP jun_j_Page_037.jp2
419bcf0c1bf67bfd753ff42ba93d1fa5
05e7c915513c76ab08ac77207cd6195d48f5cd0c
1051969 F20101118_AACOFJ jun_j_Page_100.jp2
ff8443f4d6a29528335476cbcb3d1bcd
747a00c0fc536df104e955193317213b88e8b875
964863 F20101118_AACOEV jun_j_Page_079.jp2
456b2eeb136c1d69e855db0630a94626
244a642ce823d9ac7fb3819ea14052c99bb44222
799645 F20101118_AACNZQ jun_j_Page_035.jp2
4cb4f6d1f95102ef0cedc2d5cd67c490
d78f2e5f3f5d8bc19bb8fc377dc0a68ea41eb536
95922 F20101118_AACOFK jun_j_Page_101.jp2
9daf5db63dbfd26260600b4c9f15889b
471ff007e9e6e6abe1dc8a55ea8a76e06d30a346
106212 F20101118_AACOEW jun_j_Page_082.jp2
6c5cfc6c88dbf26fb8fddf095afa4318
12d31e457334c4108d8d6014ae5e10e7b1e1ff3f
1487 F20101118_AACNZR jun_j_Page_083.txt
b2a29b9441e6ff2a36e28180a77ee3c7
73f4f57b03927aea599ded5e51a5d34a2058d13c
F20101118_AACOGA jun_j_Page_015.tif
cfec638c8188a4dac577b95a66b608a2
8fb3ae4828cf5d9385c073d8967d05caffe78cdb
38413 F20101118_AACOFL jun_j_Page_107.jp2
2231bdda43484e8cfe06022eaeb44cc7
d5ea568bc7cb0d9c608500f889255a42e3d47bb3
913163 F20101118_AACOEX jun_j_Page_083.jp2
a80e83275d618774ea1cd2aa8df76552
33cd1d206636487f543775b72bc482b12dbf0606
61599 F20101118_AACNZS jun_j_Page_018.jpg
8daf09d0900c2af5208b429f53befc62
b4ccbbaa8902cf1dcca201754e91cf7d20745ba2
F20101118_AACOGB jun_j_Page_016.tif
b91ed70cf6c26b97653ba92376e4bb98
427ea8ce1fca6cb86f9df5fe0dec4b4ca9ebf874
987587 F20101118_AACOFM jun_j_Page_111.jp2
26d084313c34e1c6bd198d94a5a1f9ad
85d4cb598bdf94a53518f45a54a158b213de08fc
724625 F20101118_AACOEY jun_j_Page_085.jp2
a2ba3714ac629de83ada57a3664ff24e
861f07dec5277ded3b62572f88f3a65ad002a569
45881 F20101118_AACNZT jun_j_Page_115.pro
2abbc206c5acc378c96ea2bc24392de2
d42bae25e4c47dd75252e47e7ccd7a5a3b6c1001
F20101118_AACOGC jun_j_Page_020.tif
8e3b2676140a9da9b87be1e3c9cc21cb
7efe32b1d34b395e87c35370402c296c8b103780
670613 F20101118_AACOFN jun_j_Page_113.jp2
0ee87e1a5ed14abdaad4e94bd27051f8
b5a543a2462905de56c72bf09e0d4f6d4e02b990
1051978 F20101118_AACOEZ jun_j_Page_086.jp2
eeabb1d4f8d4d74b7d8552b092101b81
a4574e7f4525685b00c023c73e7932c842a20429
F20101118_AACNZU jun_j_Page_094.tif
b038b2ec6e187eef1649e4fbb8ef6b65
3378caa132b0ddc29487559a594bc27ce235ee4e
F20101118_AACOGD jun_j_Page_022.tif
098aeff51c4e61a7ba09c143489168a2
797e726f9edbf2803a2651c4b6b9af54a7a8d564
929771 F20101118_AACOFO jun_j_Page_116.jp2
15cbdeb1cdfa8cab3fff5505843435d5
0d43c6ba1534645795c11b7b2b57fc2bb2e29e1e
73714 F20101118_AACNZV jun_j_Page_022.jpg
9d5c2cfe6c0711a3ec2b785bd0bb2a55
5f4b5844109b8d3e91544c0dea23394a08a87a75
F20101118_AACOGE jun_j_Page_024.tif
7878844dcf64299600ebabb5acae2cf5
fee038ff7c1af4f8479230ae244bf2381f3dd93b
499401 F20101118_AACOFP jun_j_Page_117.jp2
130b21c1b237db6dc88b0a8e55cfda8d
1bdccb2b96090b70af0b3ffb6046d9073f27bf5d
22232 F20101118_AACNZW jun_j_Page_031.QC.jpg
256532662ba9a27e5833bd9d4e082174
a7b3aa541a29c9f89bf4bacd1c9b75e2f700bcee
1051985 F20101118_AACOFQ jun_j_Page_118.jp2
d6c4a4e4c3c5423cf1e9a627f39fd7eb
aec3eb1939dfb0881bddd8328e861059c24fe719
1978 F20101118_AACNZX jun_j_Page_092.txt
ba79abfc7dec7bfc528392fb69cc0290
81f33b284c98bbaae79c6cd6608295fa325f232d
F20101118_AACOGF jun_j_Page_025.tif
00cd2aa3f7a00e911b605e4d9b51f2ce
960629c385b4214f6c3e7f1ad37db5b1fcdc35f7
1051977 F20101118_AACOFR jun_j_Page_119.jp2
d3cb942c9cb5d708471fbb3049be3161
fd5664e8fe1fe17f9aea7639f025c0f69cd2e4bb
40223 F20101118_AACNZY jun_j_Page_114.pro
fe20ff13c1efabe9bc543990454aba7d
ed4244cbc9bba11be09362f7d3cf6d37b0a1950f
F20101118_AACOGG jun_j_Page_026.tif
e5978d1bbe63ee0b3de7ea2666d79596
4b27ac886466bcaf61aff0f5593d5467bb4c794d
1051965 F20101118_AACOFS jun_j_Page_121.jp2
e875495dce887bce836a6e1c45637252
86b090f06f9572efc2ce5cd955c6d4ebac17d88d
F20101118_AACOGH jun_j_Page_028.tif
6407ef322b25e62add6e5fb4c653843f
9c09a4b3fd2975a18496589d2b99cd0a54ee20c4
43692 F20101118_AACOFT jun_j_Page_123.jp2
9c222f331122dc3de052d3898d242e1a
3501cd70f24203053a9ff9c36854330134acc856
5809 F20101118_AACNZZ jun_j_Page_073thm.jpg
ca5a62a679f0af430acf7ff4f9dab67b
ae38bbc8d974bd1cc4cee0c99130d3473b3d0cc2
F20101118_AACOGI jun_j_Page_031.tif
e0fb73a96dcf7772cacc66aa7cfe21e9
0469483786d0dd9f8e150c00926602a3f8922542
F20101118_AACOFU jun_j_Page_003.tif
ddd98273795c27c41ca3e4d248c2ee59
85ebcf571b05e464ac0e7a43c732d7a0e68e77b1
F20101118_AACOGJ jun_j_Page_032.tif
2613bf0e40cae88e4833abff30faabad
a867298c8a2b37376b7c3446d0b2b2b10f58b23e
F20101118_AACOFV jun_j_Page_007.tif
82f950339e16cbd4793777be09995016
d188655537932a9a3971d398398944b7a9d6e0c2
F20101118_AACOGK jun_j_Page_033.tif
fd2e009245a2ece0faf963acbfcfeb4a
d82142e54b276665923e59d0a91cef111e890b12
F20101118_AACOFW jun_j_Page_009.tif
ddd81932a1fbe2cf37a1594cbfe4fb2f
f0a50f9469f34a648fff30dfc9f667f95d82cf5a
F20101118_AACOGL jun_j_Page_035.tif
4355d512e2c45de2e70fab45de45c550
2951b7ad2ebb23c9bc38b222cd3e4e8e554921db
F20101118_AACOFX jun_j_Page_011.tif
fd65bd609208200f7ef950fa320f6e13
c4f1c86c6fcfcaee6939bd573ec14d1887192a73
F20101118_AACOHA jun_j_Page_059.tif
3c062cca4bf5a4d194ae80591a23b269
79ceb941ea0336316af4c35f3c638648629cadf2
F20101118_AACOGM jun_j_Page_036.tif
fd234d1bb5632a622874204128d8de85
f1b2554ca4bb381b57a02c3a5b48e4698397503b
F20101118_AACOFY jun_j_Page_012.tif
64a74156c02eea1715782d101f4501d2
9fcf705c2491a6c23f1fa28d8e104aaa62951f80
F20101118_AACOHB jun_j_Page_060.tif
f1f9ca8d24622983e0b58b9bbd8a0275
96ca0d5deadba39033c0efc9837ac495e8b95810
F20101118_AACOGN jun_j_Page_037.tif
16b55f67721e9e78da9dd45300954461
b8a9380e03f00bb3e9f74cff3763c7017df82a24
F20101118_AACOFZ jun_j_Page_013.tif
acd561b99b9c987d5ff5041378299b36
62fa7b781933f47f779015768cc88a23c427847c
F20101118_AACOHC jun_j_Page_061.tif
6cde7d7d473e6ef4643835590b9f9f67
bb9f6fc8e88b179f98c58afa4abc651478d4a488
F20101118_AACOGO jun_j_Page_038.tif
bfd219da0dfb86e7f3f62c7632051821
eb4dfd2ddd83554010204118830d75f4722b7d17
F20101118_AACOHD jun_j_Page_063.tif
5b94acfd1b7a0400b7f0ff3daf567707
e7b60ae16683b5376bc6738bcfa27e8f2b06455c
F20101118_AACOGP jun_j_Page_039.tif
cf01b75d8fc4d36d5388b10a1b1c88ad
c8c10fad0929a4482372afd0c13a85f6558a0b60
F20101118_AACOHE jun_j_Page_065.tif
6a3f6aedbbe2e2ddf2c3c2b09fa5f24d
16e3b7bed39f43c3591ba253bb89f87232f987a3
F20101118_AACOGQ jun_j_Page_041.tif
3082c0d6750b9148fe42d327845eb860
474f719ab145e962d38b93ca3e5810c005e359d7
F20101118_AACOHF jun_j_Page_066.tif
5fc67d8dd1e4336dffd0b101bfe1f6ea
7250eea2d30e4385471a88970cb9e7d52f5b19f3
F20101118_AACOGR jun_j_Page_042.tif
f2c0375a6ad4cd78f63cc4b2df7e762a
6497b50c37848719cccc2e9036923953b845d459
F20101118_AACOGS jun_j_Page_043.tif
2a2b5b477e0dd9b1a21d1ef0f4d11a2b
fb7f24c7cc411801e05ab73836e732bf67c7d970
F20101118_AACOHG jun_j_Page_067.tif
4d36bb3088e7a86d44aa2acdfb9e9cdd
6d288b0853142985830971e18f6d184fdd94ca14
F20101118_AACOGT jun_j_Page_044.tif
87c71972ddc3db0b1fe302475d181e0b
8620a343af0139f46baa118c10e392c0af835547
F20101118_AACOHH jun_j_Page_068.tif
769f269d9419560108edf9967ced3405
3b557045c428293a53ccf784cc092517307a645e
F20101118_AACOGU jun_j_Page_048.tif
a54ae9a93673ab97c22b9b6b0b98856f
060c09c49fc460a591b18012fc4e8009374bfaaf
F20101118_AACOHI jun_j_Page_071.tif
c465fa39d9741388bf52f38ab1daadca
3d12f1c51945db4fc036baa0a183c2e68fca5553
F20101118_AACOGV jun_j_Page_053.tif
e2d136514d8ed4b11bc3f21220ea2a8e
cb323d1b52acf0ad552bf9a1126cbffa20172b01
F20101118_AACOHJ jun_j_Page_072.tif
09784dbeece062eac6df3ca18ebb63b5
49a84a37da3a52bc570c37b2d4d548c713916301
F20101118_AACOGW jun_j_Page_054.tif
43f35c2b3269bb919a85e7ffa03fc3dc
ec0abce666d670223210cdd084636256c08d150e
F20101118_AACOHK jun_j_Page_073.tif
d2dab2fb949e30db615a16fc94ac57fc
05f771edc806471a9b60bd7ee3f77c9c27ef5c54
F20101118_AACOGX jun_j_Page_055.tif
b0ce40e66c889987dbdd15090a8ead49
8246c0ad106ce5e2cae74646abe7fad75f04e8b3
F20101118_AACOIA jun_j_Page_095.tif
9827da47e9c25ce6908cc464a4a36a13
19ec9278514cf28e00930fa0d64744348c9cdaf0
F20101118_AACOHL jun_j_Page_075.tif
cf50fa81489011386061a6795ce73551
d96630b435c8c4f7ec40630a716dfa120e17b15d
F20101118_AACOGY jun_j_Page_057.tif
515e47b87cee6bf5bcf29018416b33d0
d98a8426cc3483b745d3f2ec9cdf99457c188294
F20101118_AACOIB jun_j_Page_097.tif
df8d25ee0b1eb85682e14f2e63289122
c253ab0a292f0046ffb9a6af48e241f44768bda3
F20101118_AACOHM jun_j_Page_076.tif
18e5c2676ae3af5af65f0cf1d03a9d82
b03c69f372d41585c0c784c67fe5c82c1383602f
F20101118_AACOGZ jun_j_Page_058.tif
f91d5933ae8cf7d536c437080f781583
05fff3768f6f04082067194c415035660d3f3790
F20101118_AACOIC jun_j_Page_101.tif
6d145422da953a0efb1e082dc8b3f023
1e5586d407c8a368577688333b911c71ef934f68
F20101118_AACOHN jun_j_Page_077.tif
4eedcfeaeabcde78bdadcc17c82e2b9a
7b7416f4eb9a72447d0bfb0dca3458f40b365bf2
F20101118_AACOID jun_j_Page_103.tif
3de20438c0d831490f955435db69c0f5
b3d674238442f5a90aef03899892c58e8b26ad1d
F20101118_AACOHO jun_j_Page_078.tif
a8b3ef01a9736e85c0448bfb4496882f
ff5f3c4c0ef833a87f61557839fb74918549104f
F20101118_AACOIE jun_j_Page_104.tif
84dfba47ad29d3cbb165df3517147ff4
bdfc5efb4787dd551eb07c4a9c5fc08438ead5ca
F20101118_AACOHP jun_j_Page_079.tif
9b99cd023902c0a45c854178e9353fc8
20aaf854ac48bf46a09f9c7ba41d45080774a763
F20101118_AACOIF jun_j_Page_108.tif
ef5768cf5576a86f1dbf47bc360fb8e4
dabdd9bf4c674ebcf2a20265a35a5ba3bb227f3d
F20101118_AACOHQ jun_j_Page_080.tif
6923e917c7afc32d718c644b4ae66793
af26a3e1d1122708451075f9b0084bea4576edd9
F20101118_AACOIG jun_j_Page_109.tif
f5d432144b555bd4014e671514c2ee18
14516ec6f3697292cf670a0f588ae7dd14cea731
F20101118_AACOHR jun_j_Page_081.tif
62707bbd01d9f21d57aebb606ad00bf6
0d97c3dedf5ce9d72c832b6f8b363f29dd516a1d
F20101118_AACOHS jun_j_Page_083.tif
156ec8bd91494926c1185cc399622db0
ea37fcca871971c87921fb709ceeafeff579ae65
F20101118_AACOIH jun_j_Page_110.tif
a4f73e110bdca0272a5110a232ae8b1e
884beadc40796de0cdeb9f1cfc1f4d115deda536
F20101118_AACOHT jun_j_Page_084.tif
9bfda92047382ed340de048a3cb1bb9c
cb1f0dc2e3fe0a9285b910f5957debacd442224a
F20101118_AACOII jun_j_Page_111.tif
8414bdf030e2baac8763431f8255707a
c90f7e681d32904f844b086da047c9e2f0213a73
F20101118_AACOHU jun_j_Page_085.tif
4298d5c1bb106c33f9695f72643d44fc
c8ca9a6ab91d97b223809dca56a1979a7a739e4b
F20101118_AACOIJ jun_j_Page_113.tif
271f1c2ec0b0f087ef7e564639b43287
225911a2907abc45f0289b49ed7426f511e9429a
F20101118_AACOHV jun_j_Page_086.tif
46b771bed4b3a1c61494f8eb380d586b
b9937ddd137269972ed796c8dc29e221b5e57915
F20101118_AACOIK jun_j_Page_115.tif
0d854301f815b366cf1540d941d62925
7ef6e29d06936a3e9f4dc2bafdbbca60855bf15c
F20101118_AACOHW jun_j_Page_088.tif
94d9e9ad53838468726a000da462f7ad
3c44c7520bc36aff8128c473ca43074bacdff79d
74518 F20101118_AACOJA jun_j_Page_032.pro
64200e3fed9c005c02f4cac8ff28e8f8
a616ef4d743db7ae59738f21b40c27ab6c06c408
F20101118_AACOIL jun_j_Page_121.tif
9a9e14134798baa015efb5bc293471ab
201cdf1cd472fd6960a07f46d47ee9aa7e6cb9e8
F20101118_AACOHX jun_j_Page_089.tif
e00380e4e2f8fae81e7057236043c24c
df70bbd704bf630acdb3095ad722c9aebfb775fe
47254 F20101118_AACOJB jun_j_Page_033.pro
df759af6999a1058441511c9efdb22d6
280c9781face3141a9cad974cdc5ae81d9e2aa91
F20101118_AACOIM jun_j_Page_123.tif
90bf8ee9dd7d23bbbd054aad5eefb95f
0937d16216d6298824dfc2278ac5029494df00e2
F20101118_AACOHY jun_j_Page_091.tif
c362981a289ccd8659a3607ba953f5b9
534cf0090c59ebcbba01babbcca55ac6a897618b
50303 F20101118_AACOJC jun_j_Page_034.pro
f1bcfcb878cd92a651786253cbf9cc24
f16d0784e6435fb0b094a4c16a1e8ae9439efeac
73599 F20101118_AACOIN jun_j_Page_004.pro
32b3f50b7bae3bf85694bff20f716295
e0244abac345ff7eec1e9e3c4e02cbeae2e27153
F20101118_AACOHZ jun_j_Page_092.tif
3ceda21ee07fe8cbf6555c42d29020aa
98fe6fc866b324f96d6a9527b3203bb3af793332
18127 F20101118_AACOJD jun_j_Page_035.pro
acef03d5e71607a93d15c3cc91254ee1
b56124b2f562d215c7fd20e124a8c425fd2753a3
5084 F20101118_AACOIO jun_j_Page_008.pro
09d924ed92737020c9e6c959bfcdaf13
2ed105690d3fdd92f5808ea52b5143c3c66ce2fd
36301 F20101118_AACOJE jun_j_Page_036.pro
6c23fdec00b6e40165376d7eb56aae3a
13eccd7022e19c1564f582f0f21217f1dbfd1581
58784 F20101118_AACOIP jun_j_Page_009.pro
96c38febd74245476328222bbb4c3755
6ad239a4ecfaea464ad4849478c116871bcf6cc3
41344 F20101118_AACOJF jun_j_Page_037.pro
1f43f5ddc62e73d984b3da4812826fc0
27d03649b5b3f7856b77c7741ac25645aab130d1
65072 F20101118_AACOIQ jun_j_Page_011.pro
37c7a58a370434148a3461771d20ad35
6797fb98de6107a0a95ee486d66e9a312d3f0ad2
34760 F20101118_AACOJG jun_j_Page_041.pro
b9eb76a2c1925345b53eb30eb714eb26
1974a564a999c6cfc90b17c7627990da9faa2897
11107 F20101118_AACOIR jun_j_Page_014.pro
6096eeea29a71b18379f7b5ecf31f867
0838001d7cf15c121adfbeecdc90b5779a6bd8b8
36456 F20101118_AACOJH jun_j_Page_044.pro
2348ce22de1d9a7729b756e85980c877
09f6be0265dcc6c465e2dfdd575ae0fa133cffc4
27369 F20101118_AACOIS jun_j_Page_017.pro
d07ed7bb0b735371295d3c7bfbb00c6a
e539135d513665c64a67fd6db8259900a8eccf9a
36692 F20101118_AACOIT jun_j_Page_023.pro
2ef5bf89185348600fcb8b8068216fe6
79e6ced3d22252436b341086f5a759140f083121
47453 F20101118_AACOJI jun_j_Page_045.pro
c6345db991bc4cda22d6c82c4127817c
e7a8af1199b55aea3c5056fc2fe07351cd4d8f81
52742 F20101118_AACOIU jun_j_Page_024.pro
0b41a29caa46a5761861ce32ccb15abd
21f65f81bc2d93fb0479e7cfdf1d24740da8b68f
29885 F20101118_AACOJJ jun_j_Page_046.pro
fdf12acdd9e86c8a7900a9ab53f694d9
a7e836def00db3b113cf976e26bad9eda29b0ab7
27759 F20101118_AACOIV jun_j_Page_025.pro
9a774b67381be91a57894bafbc255afa
6e3122aeef2a34abc9194a96c2b49f14899d0260
46313 F20101118_AACOJK jun_j_Page_048.pro
42e0c22183edac2fbf827916c4cc4c6f
08766cdf923c39e28b44eae166c41c60c2ae3458
47555 F20101118_AACOIW jun_j_Page_026.pro
35f5033f243a58a34c334a869e70e19d
1a7879ce6f0c298c05d33c4524f768ac315ac512
23704 F20101118_AACOJL jun_j_Page_049.pro
c1a77cf39cf86cbafa7aab844a6e9fce
ef432a355291efb38f85ef30eab2c7b6d7e9ef34
43107 F20101118_AACOIX jun_j_Page_027.pro
a618840c8755c9692bf992707b961b40
7d24eefc89a7e2b7191a802482e5c0ce9ed07527
13903 F20101118_AACOKA jun_j_Page_072.pro
cf7399d2d2420bb20227cf3001f32c16
1d053c0d5aac739568760b94c0b4ab6f7f8d1b27
11392 F20101118_AACOJM jun_j_Page_050.pro
56a3368425bce301dd27bf11f13090bd
e5f34425a637343d2c44e9247abd1e9a06966547
37893 F20101118_AACOIY jun_j_Page_029.pro
7a2322eefbd39d7e706f50c427ba016b
233416e65d5dc2aad80847f49a14ca0e60d267ca
31800 F20101118_AACOKB jun_j_Page_073.pro
4302c37c1d58bb4070f96015db8f8635
45a1425a690e0b0a0e401a512b87c75f50c47dd4
40724 F20101118_AACOJN jun_j_Page_051.pro
614377838e824845ff3f8d1012f28b3e
ce6a8010b743105e3d9b4e6469e4238823cd94b4
37340 F20101118_AACOIZ jun_j_Page_030.pro
71af9c6197b298d252a1a254bf695811
49f9996978f683141492f814aa985f662b289a03
40492 F20101118_AACOKC jun_j_Page_074.pro
0205e2b19c9215dfcd8fc91239303fca
c00b92a61d1bac90bb2ca1907808f3d78588477e
26028 F20101118_AACOKD jun_j_Page_075.pro
4ec9bcd227dce3ce95206c185c1a0c47
cca5255bc7ad73ef70f9e927c738bf48f2eea48b
33014 F20101118_AACOJO jun_j_Page_054.pro
519c9c2c3db20f775192b3d718385367
1e947b17ddd8e4c4514c0c100cd8cc7659740939
36794 F20101118_AACOKE jun_j_Page_076.pro
60a6b91ab8d601070208d001bd8221d9
058902bdc9c459695828f0c118c931f0e5cd1908
23724 F20101118_AACOJP jun_j_Page_055.pro
ee3f14edc335545e981159527c501626
1aab06bbb0851b6f44c23a093712726b57a3a0df
36749 F20101118_AACOKF jun_j_Page_077.pro
bb9c6b29023ce33c73e8057d76813a9a
56d7c8a2c1cdc388ecac4d05032520d87261eb69
19765 F20101118_AACOJQ jun_j_Page_059.pro
1c6a0da4d821b8fb31877f3456542759
a9082eb0a4334b59ca3d504037102cc20134d3a8
42894 F20101118_AACOKG jun_j_Page_079.pro
31a4296157d29bde81b6af41f75ab5e4
8b6a4c263895ba8f597000b25a54361cfb644e1b
41756 F20101118_AACOJR jun_j_Page_060.pro
f419a18bd02dbc03c24abd8e616c3763
195ee65020a60ec1b01bfd8e96f7a5c9b6f11092
48776 F20101118_AACOKH jun_j_Page_082.pro
d5b867d3c4204bca71d0d71f298c289b
f949d8acab49f258674ec016024b8ddb06b5d768
46415 F20101118_AACOJS jun_j_Page_061.pro
efe38576f479a0419a0f1fe49ff7869a
5cb33a7b0892fb2a881a32dab07fc80f72626c53
30470 F20101118_AACOKI jun_j_Page_084.pro
deebe7a29c63c49bec29f92d292c395a
9fddfe552b7abab7323e46ab8c91c7cfbf83aef7
44908 F20101118_AACOJT jun_j_Page_064.pro
b86a82baf54d993c342d9b9815c2b7ea
d3e71d4dc8b6a76b5511a09e1cedf25147791a54
44816 F20101118_AACOJU jun_j_Page_065.pro
9d19adeade653a5fe1406e07c596242a
4406bbc9eb80b2986549235eb6799355eb225aa2
47914 F20101118_AACOKJ jun_j_Page_087.pro
b977cff14bf5ccf11f732d5d6e8591a1
a43fb1929a639d34b326f6b81ca6cfc480dc3057
40146 F20101118_AACOJV jun_j_Page_066.pro
1468dec4a3bd8698a42b55356124d527
a61a8a7cab2127f7a72dd8ef94e9be900032e1d2
33671 F20101118_AACOKK jun_j_Page_091.pro
464e155b3361e093f283a8d290cab738
826a5307d73400642ef17921158cea4db03e7a42
34129 F20101118_AACOJW jun_j_Page_067.pro
a9a57b91c219c7f93f8ac223b57922b8
0e441758c8859dc653c29109afe4d8fc295341bf
2176 F20101118_AACOLA jun_j_Page_007.txt
6efce88aa26b084e01cd448d7a745ab5
41fb3e930490a068d5dee4667e0660c432ad1866
41910 F20101118_AACOKL jun_j_Page_092.pro
137d7db31b9f0c65afa40c3923a6e348
09071971452c527d68b6391ea10b5e6337302a63
47110 F20101118_AACOJX jun_j_Page_068.pro
fd0c8359b30002102f42d82e5ac72ccb
aa912b8a88bc4a626c403089d88cc53c56f6eab9
206 F20101118_AACOLB jun_j_Page_008.txt
217b62945f6b9fc2b91e32867952bf2c
642276fecaaaa54b2dc753fa0338b4d6dcc0056b
46447 F20101118_AACOKM jun_j_Page_095.pro
d9c79f1df4e0007c62d129ca3b1ee379
d48de7ef593f77cb030a762db24ebf4b3e1dd9eb
52877 F20101118_AACOJY jun_j_Page_069.pro
bea4a2449411c0edadfb69b29618fda7
124b6b7c187170d8e8639f522cee54e8e2ee8d5b
2334 F20101118_AACOLC jun_j_Page_009.txt
4d1880f49e896808c8998be7e06bb4a3
7789a7466f51b2d6cbd9a52b902b0a5d70c9388d
37229 F20101118_AACOKN jun_j_Page_096.pro
d8162481b6dcf45a5bf94521373f2fdf
d86cdd94e0acc742e2ff94f2def2c661b948c1f2
25117 F20101118_AACOJZ jun_j_Page_071.pro
0b401e8cb2217a73dcce14e257070978
5333566f5b0290756182e12fc1ff4767a60b2bdb
2701 F20101118_AACOLD jun_j_Page_010.txt
58281a95fdd6ff3aa0636498417d1e3a
8aae324e6cc1f11881c857a5a1f80bfc927a2fc3
34368 F20101118_AACOKO jun_j_Page_099.pro
9bf8bb7795e03269fe31550a377c471d
e5f3e62e166e60cf07fc24bcde161ebe16770f7e
2543 F20101118_AACOLE jun_j_Page_011.txt
32cf3e421926b92bb4a0223f82005265
30408ecc10085ff315355923b297c872a6a37214
38666 F20101118_AACOKP jun_j_Page_104.pro
76b612fb8d07e0a69c17e8c98cdb726e
f86076486234c6fa27c996a66faa8379fc0202ea
1102 F20101118_AACOLF jun_j_Page_012.txt
f4974be7324439da43d213a7ecd4c685
6ac519feb97198debfb8a48e57398710a9640c78
34996 F20101118_AACOKQ jun_j_Page_106.pro
3be183057ef7f9e7e050386955d87161
64b3b1f657b5c6d4044d432aff895fe03f0770d9
1594 F20101118_AACOLG jun_j_Page_013.txt
7b8e71647f73b56225e513378cd6331e
7ec65f2f4b2efaeef1c2f125ddfd3aa56bfcdfca
31098 F20101118_AACOKR jun_j_Page_110.pro
b1d390bd968c7b1619189cbe6cf77f82
9ca9730f200a8173b255cb8ef197582bc2e94f3d
1679 F20101118_AACOLH jun_j_Page_016.txt
c1054f1c1d6512ece471a7f4b7bab643
59fd6e82ac58ed5b4e1ad7468c3ab6a902d30c66
38084 F20101118_AACOKS jun_j_Page_111.pro
f1a217d5c0e0bab840d588f3f892ef98
23916eb665a44e469a9aa26a5ed470d8a7a13dbb
1793 F20101118_AACOLI jun_j_Page_019.txt
de4c6930bb0304b4e5a4c62bc3968a6e
8d5fbd919eca4a8ca52452433e6578ae69264ee4
17385 F20101118_AACOKT jun_j_Page_112.pro
7b53a52822a45cbbe4396e59abe94edd
50c3bd5894524aa7c8a0bb2cc71bf8e1baf85371
1500 F20101118_AACOLJ jun_j_Page_023.txt
250fed8ecdabf60003050837b4834b4b
244b24f3bc66c0995be395957a3f81aaeac0b0df
2363 F20101118_AACOKU jun_j_Page_113.pro
e1068ee6c32b6ed1d9fea19d1e173add
9a8deb1f6ba3f03779ad1bb592246b30851a8aee
116 F20101118_AACOKV jun_j_Page_002.txt
4200f7ad280900d0dec61b3f76ef7457
ae192d25b3377939369bed4ad14e75d70c24378b
1636 F20101118_AACOLK jun_j_Page_025.txt
a6dbeaa2ed4de6e2ea83840d20aea609
2300657e70b160fdfa4ac3c465372492345830a8
1312 F20101118_AACOKW jun_j_Page_003.txt
701695c03398bcd3c9ef2003d3164c83
9e6794086fa3a00c53731b224f93d79e8698b298
1906 F20101118_AACOLL jun_j_Page_026.txt
cd7cac3aa8cfdeec9661b71dbe40c457
9a5ac9cd40c1ea34ede580658077c48fcb361e6e
3075 F20101118_AACOKX jun_j_Page_004.txt
05346a18e0294f62a3b385605389ac13
d39327783cb70a0ef855230766ac3aaef15d108a
980 F20101118_AACOMA jun_j_Page_049.txt
229d4003fad12589a7ac596e65af1a79
422465a1465ea1b4e05464d9116e45892b160e98
1719 F20101118_AACOLM jun_j_Page_027.txt
bc12a3caf03dec00f826cda2e8d81b24
8983ab518993e7adbcfe2d1b71c366d939ee9fc2
4262 F20101118_AACOKY jun_j_Page_005.txt
62169d09a3372d680c96668c32be430a
dba46bf51b301ea953126ffc308ba5165117184c
1754 F20101118_AACOMB jun_j_Page_051.txt
5982e4f6c817fcd2ffcc4142ebc4309d
5efb9011f513dc78f22c65c696f9067f0fd92cdf
1600 F20101118_AACOLN jun_j_Page_029.txt
099ab54630d0660270088a3a2f8012a8
a1e8a42de8f94eb04f279062982da37308f56a15
704 F20101118_AACOKZ jun_j_Page_006.txt
b810f8a8c049b873a610ce68d85bcbd5
e80cb2385da769b026fd3559c940d568bc29bd42
1045 F20101118_AACOMC jun_j_Page_052.txt
bb3642a6f4211e2f5ae6179e12c572d6
d7537cc997c671acbeb1ad4db77a3eadb4b5c62e
1839 F20101118_AACOLO jun_j_Page_030.txt
61a80c0d9c23a2e902f64935c29bf4ab
bb6f6374808c692a68e21df7a3d74cd2ea4d3bd6
1405 F20101118_AACOMD jun_j_Page_054.txt
0a6b2a6e8cb5270ff35212b3c7a95306
319e4c5cee38d2b1ed98df3c5ffb19b0b0e2f37f
1718 F20101118_AACOLP jun_j_Page_031.txt
6ab20b2efebc3d26a5c42bdcfd5791ce
f98c40f58f2b585d9c1760e690bd89a03f8a926d
1254 F20101118_AACOME jun_j_Page_056.txt
641940e28abd1b60860a636cd596e8c4
4e619151a3accb3f62290d1763eda95744718750
2025 F20101118_AACOLQ jun_j_Page_034.txt
f4713eda9ad75826559463fdc2f7ec9e
caec7510647c9630d11defcfde47b06533c48cb5
1958 F20101118_AACOMF jun_j_Page_058.txt
957be4d29492e3faa7bb4c888af32ada
87365bed157f0484f23dcb4fd037e0d20a8fbc1b
938 F20101118_AACOLR jun_j_Page_035.txt
e085d0a9d39e43f79aacf8d9bb4a824d
3472f76aee086874a20adce8dcea019444e9693a
855 F20101118_AACOMG jun_j_Page_059.txt
2e80e35b5f8c0aa9b534b1863951a1c9
160eb4915cc6cbc18945516c8c70375a699a21d2
1689 F20101118_AACOLS jun_j_Page_037.txt
5bc4643d32bf34eff2bb258e82dd3ebc
72dade27a788e8c1ebd21b4c933eb2a2944b1a02
2065 F20101118_AACOMH jun_j_Page_061.txt
56c7cfb073123667a10a507f3d0d86f8
76bb47d7af7667f3ca2de64f3f939f21a71879ef
1873 F20101118_AACOLT jun_j_Page_038.txt
fccbf0c05a08cbaa486c8af6db5b27a7
0d87487297a0ba1a923123215a620763300d1bd4
1946 F20101118_AACOMI jun_j_Page_062.txt
5ad0acdf5852a871e7ddb4012118f6d1
31e27ffcba56f808df74506d66c954837d53516e
933 F20101118_AACOLU jun_j_Page_040.txt
3c15904f9880f855db0a46152bf81d87
109ec48f9c79609209371915d3a3b242ad06c2d4
1788 F20101118_AACOMJ jun_j_Page_066.txt
32e6ca7331436245ea4465e1b552730b
16a1ddafd5ba96bad57e92abbe33b0d44b58d9a5
1534 F20101118_AACOLV jun_j_Page_041.txt
626dd3ac55d92db63cd83e239f6412e2
decfc4e51f84d07bd97eb8d93552f2de4fd3a9f1
2604 F20101118_AACOMK jun_j_Page_069.txt
b270e26d4b74428c0c39d4a1eac87b5e
881d7ac6a21fe6d532c0544289f61748f83c7b9e
1527 F20101118_AACOLW jun_j_Page_042.txt
26f7e7fb258f3940698e51ed47908c61
4ede7a928f89730e28890689e02d396a596a0bf0
1877 F20101118_AACOLX jun_j_Page_045.txt
cd742f4c7cab6717ff3ee3d7f7efaab3
95898a4734012be20192197ab08bd33fc007ff5a
2251 F20101118_AACONA jun_j_Page_105.txt
c2295b66faf65e9be5b2a74cc994ccf4
8859b0e3c616b1fc4bca47431566b09fd98235e7
1332 F20101118_AACOML jun_j_Page_073.txt
f6513930dbbc4fdbbf3b2ff26781ad57
836367b2530384eb7cf69a9ae016fd5c6237c13b
1263 F20101118_AACOLY jun_j_Page_046.txt
bd9b72787c5a6f5bfe790d01762ef0c0
9cb4fdf99c3f46c7fd432b3e9f9081130720b720
1641 F20101118_AACONB jun_j_Page_106.txt
3f0e0256d1c8d06ba6f60f9374b7d0ee
b8e713cdb9fbc4e2e5443dc8242714fd47ca02ea
1643 F20101118_AACOMM jun_j_Page_076.txt
1d388aa6b6d5a7d561e8ffab9da6fabb
a924d3787be36aefdfa5cb8884e7d670dabb0fa4
F20101118_AACOLZ jun_j_Page_048.txt
f496bfd4da58b9c8769c84e34872f7f5
ec020a09871debcd6938678c5fd430b1b9f6bf15
649 F20101118_AACONC jun_j_Page_107.txt
fe2778d2776568ab7442a041d790845b
4273d742e2ab79a9962fdf07786293335613739f
1376 F20101118_AACOMN jun_j_Page_078.txt
19ac2e97f6f8486a9d2311714c9e59f8
bc8e32d534746d9177c53853898521591f6969c7
1912 F20101118_AACOND jun_j_Page_108.txt
9dfcab9db92f39141a1fa2c9c6a6ad1c
7f446167463607be2edd6f35f4914742d13ba7ea
1512 F20101118_AACOMO jun_j_Page_080.txt
ec608a47efe4c0a0bc3a0c4bec217172
b22a0d5f9b9d3a9375169476c4170c32ae94a7b9
1244 F20101118_AACONE jun_j_Page_109.txt
e0d8e4daeec90d6f458d792a35968d11
ae9813a9cde49cd670c57258a77410d81fcc029b
1949 F20101118_AACOMP jun_j_Page_082.txt
f1869419c623f32d04da2bc100786bac
474d52dec8d92f72255ca90b5175717ae3602e79
1294 F20101118_AACONF jun_j_Page_110.txt
5f2207e60e33233a40806d461f4e92c1
b21a7dca2f4afaa80c84621d23c0ab0613da9fbe
1252 F20101118_AACOMQ jun_j_Page_084.txt
08568f756a30f2a30b3263990d089e72
d4292fa268e04a3cd30a17023b5eab1437830cb3
1538 F20101118_AACONG jun_j_Page_111.txt
8f6030a8f8d67ad72d82438974665229
91f72a4cda3fd4a782c30a5f492f49061c645ae2
1567 F20101118_AACOMR jun_j_Page_088.txt
6143955337dce7454f95a90d926fcc79
d985f9c1794795036dabec4a080c0cca4fa83fbb
758 F20101118_AACONH jun_j_Page_112.txt
6cc454255ac4fbbaea5a80a5ff4627cc
8a1a32d4c3a357f8295b1bb27ee67de6eb6430c1
1494 F20101118_AACOMS jun_j_Page_090.txt
c876d896d5dd468085ea38f43c89009c
f95a5aa65a4bdc3e035ade170246a5e723279a46
137 F20101118_AACONI jun_j_Page_113.txt
2d546d062aa5f2f3ddca51a91710e0ca
dfbdf0acb939c7695b34b18989925c5e046b13e0
2024 F20101118_AACOMT jun_j_Page_095.txt
8c6d7f25d5366e29651c20892f843c2f
3e60fb4737503c41e1d2a8851ec562bbbea1aad8
1785 F20101118_AACONJ jun_j_Page_114.txt
4d6f3feb907833e2a3724af8cce393a5
8514405735cf258ed263c06e5b28bd140e0b3481
1973 F20101118_AACOMU jun_j_Page_098.txt
9a7f63a46804331b1bd3b6310dd6e5f8
d499f97c3f5716d4c84f71939044020f87f6e791
1858 F20101118_AACONK jun_j_Page_115.txt
f606ed8c5f906dfffed176c353e8dfda
f9cc3401dcc3db275bba5c9b9f9cf8a14e1929d2
1410 F20101118_AACOMV jun_j_Page_099.txt
e8c2bd6ab8eb994449bae6e125a652c5
687f4c8deb0eb65e4bebe5b5e18708ed5c5f8841
1429 F20101118_AACONL jun_j_Page_116.txt
5b5d497aa37d7dc9689f398d42753ee2
777cce2f12bbc4be1981fcd9f32adbe3633b85eb
1801 F20101118_AACOMW jun_j_Page_101.txt
05be3a6894f886021d4ac56be559c47f
2bd4602b0cf24453ab9ba92a57f6f96fefc521ef
17261 F20101118_AACOOA jun_j_Page_025.QC.jpg
b992c4b15d75bfc63f1facf110b08ab5
4f43757f81a10d3b6c5e047cb0f9b60ec37e1c01
2234 F20101118_AACOMX jun_j_Page_102.txt
142133bd2493ce652e935d072a932fa6
455566694f585ae9ef6f2a9b8f0f9db1f447f25e
19701 F20101118_AACOOB jun_j_Page_018.QC.jpg
e85124fed2f8a8638b219eee1a3d6590
ab61a98afd4a3ca757120f2270976877be104296
193 F20101118_AACONM jun_j_Page_117.txt
1014446790bcf75a9f8673124241ac39
4cffb4e4cd311a7e6467df354693885a827b4b60
1882 F20101118_AACOMY jun_j_Page_103.txt
89ece03a12320cb5d36670779beca25c
471cb95046e7ce15b92e79b6ea7c947a5a5f915f
5609 F20101118_AACOOC jun_j_Page_078thm.jpg
e82b32d2b3f0f1a75731ef9b2c245b72
fa4ab0bd2e155a8328d1f77694f6e8544cd7c1ca
2594 F20101118_AACONN jun_j_Page_120.txt
bce8b4cdf36034668632cb73fb98fdb6
1a7ff88be7925837c15b7a05737882baddad7db6
1580 F20101118_AACOMZ jun_j_Page_104.txt
9359c95f48d16f934ad245276655b49a
d746b78048a6eaae3bfdbfd3a99f9ae9a93de7b4
17123 F20101118_AACOOD jun_j_Page_013.QC.jpg
440c9cc2e95acda2da702ebe32464c34
6d09d685e1ad131f2257eb6b3c73ce08871eed31
2484 F20101118_AACONO jun_j_Page_121.txt
8ec8587fbc40b88dd5c1bda965cd7026
107f9109f420ba8a5aa7255f2595304443b33340
5552 F20101118_AACOOE jun_j_Page_037thm.jpg
d907e319f21d248beaa87554b254e1b1
a79687c8d8efc4f5ed961676366d8d1d52d93140
1820 F20101118_AACONP jun_j_Page_122.txt
10968808b341a9db620a5b1c4dc0f48e
a29fa53fe18ce3e44600776d5f297fecf0e9b202
6520 F20101118_AACOOF jun_j_Page_074thm.jpg
a5ceb6115915c7b8dd2feb5996fdf4bf
1d21312bfbd07a2ea4b99e3ce38c79e17ce6557e
789 F20101118_AACONQ jun_j_Page_123.txt
f2e940f5177cdd68219ee31a3dfc5fb9
5603b0c0b78bb8ced0d299726ef83638c492079e
21616 F20101118_AACOOG jun_j_Page_065.QC.jpg
aea5f63274f3b9986f6b92d4516d255f
6e706aa08f3549776d886504f8f9caaaa61af68a
4618297 F20101118_AACONR jun_j.pdf
e4fa2ab465fa96d9445bc7a19b599f96
20608dd0fc04a0691ff623e95b1f0e86dd8b0b41
9479 F20101118_AACOOH jun_j_Page_107.QC.jpg
310d8378a5d749acc2cae5844c79ae62
3248dcd4821d8133c519875fce692a140f903adc
16838 F20101118_AACONS jun_j_Page_046.QC.jpg
fbdf27c5566ce32874c442c99017ff6e
853fcc3900439de23719cc37576f28cb7cdb95aa
22688 F20101118_AACOOI jun_j_Page_074.QC.jpg
a3f6009f6e1eae8ed43cff856152f558
8be9cb50a1a679cbd8174609eba41443bd760db7
20946 F20101118_AACONT jun_j_Page_108.QC.jpg
3ffbf0da2bc5b0b14cd48c81e6e9c9e5
62737479199dced144bad88ec63127209974a416
16403 F20101118_AACOOJ jun_j_Page_035.QC.jpg
04e0e0749f36769e527c5cbd465a895f
dd064671ae5120af9a7dae9c0b771c222079e253
21100 F20101118_AACONU jun_j_Page_083.QC.jpg
56fe0f3ddabfe05d702f5d075a2a81ad
9dd424814bb986de53d071b52b92bb34ef8c7f7d
18212 F20101118_AACOOK jun_j_Page_104.QC.jpg
44be0a0ca7892fa4528ed1606a732cd1
b9fafbc7ea9b313cabfe7609764146a13beb185e
6164 F20101118_AACONV jun_j_Page_030thm.jpg
4b86eddac6399c59424d023e23310b5b
0fafcc48a7a0b357a30c1ea806e97f1d88ba761a
2592 F20101118_AACOOL jun_j_Page_117thm.jpg
9b749ff7c695d7ef0e9b816db50c5e2f
023c57ad5a7533327015ccba6fbaead84cfb5864
22143 F20101118_AACONW jun_j_Page_097.QC.jpg
6089cc0330e32817f43020426fd8ac0c
be509d61e61367cfbefd15b23e278454af3b824d
5912 F20101118_AACONX jun_j_Page_020thm.jpg
73f1f9350d3dba514fde7035eefe48e3
94ea8867db5d033729312d39baba16f46a790f6f
1382 F20101118_AACOPA jun_j_Page_002thm.jpg
8d92eedef7c7d9dedea27664949e2675
307594a7b69fd2060c68f89717b6a6f81b281c53
21686 F20101118_AACOOM jun_j_Page_030.QC.jpg
8aee7295374950d3201456c99474e1f8
80bc0bdab0a35126fdc98f7d47aaef24c8f1a424
6262 F20101118_AACONY jun_j_Page_115thm.jpg
bbd1db49c3d86cee451e52f687096928
259975c5178fbb37f0062df8e282f4a63e313687
16294 F20101118_AACOPB jun_j_Page_003.QC.jpg
65e21427421c91ec1043b86b64566ec2
a7a113b0aebf639bd8db47cc26059d22734411c1
21087 F20101118_AACONZ jun_j_Page_100.QC.jpg
94e19e71358c272a9c5792fe0836edf6
e6125518e6f11b70d229354dbd1f110ae0f20697
2070 F20101118_AACNMA jun_j_Page_024.txt
87a4631a38118bc682862557d893e7d7
581d4c97899c5c606459659f90f190fc9a4ecafd
17725 F20101118_AACOPC jun_j_Page_004.QC.jpg
e21ab57e307b9ae476277c943904c968
0bd10aee9eb12378cc51e180f7d77e148b6d2a87
19798 F20101118_AACOON jun_j_Page_023.QC.jpg
03d3a9adabcf9b4d073dea77e8cda96b
70d8121e3866672b5753377dfb2833971c6d9df9
F20101118_AACNMB jun_j_Page_062.tif
36be5a96ef06e7992ea4afb17b259b58
26fdcc0d1c303379d9b6427e3125db2b99988f55
24523 F20101118_AACOPD jun_j_Page_005.QC.jpg
b15c8077562f23365f830fad60f1b29a
552f1507787495c89d30f71fbe578357dfedeef2
5029 F20101118_AACOOO jun_j_Page_072thm.jpg
885bba676ae5241341a7967ca21e14be
4435c31c52f21aef74cd2007f01f83eeab8ca633
450 F20101118_AACNMC jun_j_Page_014.txt
571a81dea0920ae45d17444834a09a4a
728ff7ce1ff55b313659a07855271d797f61608a
6015 F20101118_AACOPE jun_j_Page_005thm.jpg
974163551ef57f7dc41e575de96d92c5
66b5048987e9fb9e2d72518c99f05c9a11bf60a0
5182 F20101118_AACOOP jun_j_Page_036thm.jpg
8ca8140ab1439a43c05d2b7b5173e7b1
eb964d202fbd2787555fae78ffbc4e9fcdc4cbd8
6436 F20101118_AACNMD jun_j_Page_033thm.jpg
4ae9bd78cf9f5cc38aacf0cf43dda92a
e29e5434af4a76e9ac6dab6bbf75ae5729d75fe5
20039 F20101118_AACOPF jun_j_Page_007.QC.jpg
d6003c972e5a3196d7915ca3ce951389
890487f63da3cf8b39207eadaae38851fe7ab82b
21573 F20101118_AACOOQ jun_j_Page_076.QC.jpg
0864ac548cbee4c1d3842f737c390320
e1631813997ef1e98113ec040d9f7daa57d7b63e
839277 F20101118_AACNME jun_j_Page_080.jp2
e0c392cb4cada20f8816fa092e5b517d
eb6ed65fa6500c82e8e32062ef84bf1a6511528a
4570 F20101118_AACOPG jun_j_Page_008.QC.jpg
7ba6ac2541b7a2d0bb920d66f626785d
4cd3a3daf4bc48a7c5c50c8f61385a6e7ff64e07
5567 F20101118_AACOOR jun_j_Page_007thm.jpg
f3404d6b41f9d6d3b524ab6ebe8bcf33
c7adda927e12e36b6834beb3840b6f8a8448ca8c
5869 F20101118_AACNMF jun_j_Page_023thm.jpg
63f901354731705ae8c92544e3227dc2
db95f6943635752d6b50ef5988e1af899855051d
5685 F20101118_AACOPH jun_j_Page_009thm.jpg
c69151c943ab2d2697f01b0dbdb0fe58
f928271809619b746d0397462c95b3c27a883210
24406 F20101118_AACOOS jun_j_Page_105.QC.jpg
f0b49b98a941ab478ddc91b125db8a44
62b43edb85e45ce68ce0116af47b49ba050b4f0a
17356 F20101118_AACNMG jun_j_Page_028.QC.jpg
9af32bf03be9c6499dfca7aeaed4217a
01ee538cbaefa90d62430c7eeabde07bae4e5893
24011 F20101118_AACOPI jun_j_Page_011.QC.jpg
c2c7bed47ef5bc0ead0b11725fe40524
f475f602fe1e2ed3605393f7f483ad361a889fb4
7407 F20101118_AACOOT jun_j_Page_119thm.jpg
f6e94f1709cb7bf284cb7d1b4b681b77
ef75d788afa877d6a93646ea1347f7f367442314
69231 F20101118_AACNMH jun_j_Page_039.jpg
1f5c02cfda5ac326cc0a32bb30696424
bc0ea24a3f2565a5584fe0e00e5f0fe89d0b9a08
3493 F20101118_AACOPJ jun_j_Page_012thm.jpg
94300859f55d1fd8c8f688c8e6037170
d99a15f397dbe42726f5bcef58c0619fbb5caee5
22241 F20101118_AACOOU jun_j_Page_090.QC.jpg
215783cc9d580750d20f5b49cc93ac98
407df86ec0195ffa3eab503989b6c572e9ad13ac
5407 F20101118_AACNMI jun_j_Page_071thm.jpg
057500461f362c74dfff3f96c2a6c1c6
2bee7b7fac914c335b94a7e68ab979438a88b178
7193 F20101118_AACOPK jun_j_Page_014.QC.jpg
762c91ab5a618e499dce247e459ec9f8
3fc39267aba1f248d01d965f060ba5aadefd96de
19650 F20101118_AACOOV jun_j_Page_106.QC.jpg
2b77dd99677f01aedef2f2d910406f2d
3cae4b6e6074e206fafb11cf76ba56632c7433cd
22907 F20101118_AACNMJ jun_j_Page_026.QC.jpg
8f5f3c3caf03620832066723885d5477
df43f4af2c5c60c4374d492b9bb6882124a99def
2336 F20101118_AACOPL jun_j_Page_014thm.jpg
fe2171cfe64cf80a18f767623ce220fd
1bb6fcefe1b62f712cd0ff6f11435c211723d05f
5451 F20101118_AACOOW jun_j_Page_088thm.jpg
1f56feaa5592e7bd3f579f57456d2f80
b2b3db65506d597f18776bafe6d77619b1c8f900
18049 F20101118_AACOQA jun_j_Page_036.QC.jpg
91e8dd0a885837911f97c0ec2972f0bb
3e6c684be55446b0a198ffb6a4cea692f567f433
F20101118_AACNMK jun_j_Page_122.tif
d1590fa9ce2a98e77044344c4251d072
b184b931c63f1ad92f5e700c35c7cf5bcd45c046
6034 F20101118_AACOPM jun_j_Page_016thm.jpg
120046dac7ffdbb3aba5c3527ad10420
f8f44865abfed5cfd24dccd3579385c807841981
181935 F20101118_AACOOX UFE0014378_00001.xml FULL
33e69e9124402db0a817d698f0655e61
796f36050ed55dacb478b55d731180f6a8765d93
20125 F20101118_AACOQB jun_j_Page_037.QC.jpg
46b81649368fca842e4c2350e5dc94d8
6fbeea24443e10dc9a5815d4a319aeadfecc8adb
67820 F20101118_AACNML jun_j_Page_084.jpg
1ee85f0412230ae839fdfb817dbe3d0a
5f068e035d6cfdd7a4d909936300bd7cae29fac2
4239 F20101118_AACOPN jun_j_Page_017thm.jpg
80ab8e5245e4c12d2a946e07cff54123
f22a9cfd6683c4c9d7e00930e615dad5102d76d4
7140 F20101118_AACOOY jun_j_Page_001.QC.jpg
bd3078f30645f7273a447ea0db3da968
3c16426bef6aeef5b7f9bfe861c70cfc723c22e2
F20101118_AACNNA jun_j_Page_014.tif
22a950c5d37307397b24a689d37af133
c1fbdbbf58511eedc1f2416c2a2ed8dbb767caee
6579 F20101118_AACOQC jun_j_Page_038thm.jpg
8e1ebf19e168ad7c8b2722320d5bed24
9bf00afd058365a0a9b8ada967fba24615f5a8c5
3326 F20101118_AACOOZ jun_j_Page_002.QC.jpg
9069ea8567942b3fe708fbe5b70aa926
982a80220b4c68bc5c85330436b0718501601dd9
6330 F20101118_AACNNB jun_j_Page_105thm.jpg
1ab78e6e11af352caa3332af80b14ce7
8c28dd348a039d4b0ea5bcfb797ad01740f1c878
16162 F20101118_AACOQD jun_j_Page_040.QC.jpg
f77366aad22e96fedfd5fff2ebef2fe6
17e9167fde731dd3004cf2ae33a4f51396a2a296
5998 F20101118_AACNMM jun_j_Page_066thm.jpg
155783b4bfad9667cc932a81ff9f8415
3c84c98cd3990b4653d4327fcc3836adca00bdb6
21222 F20101118_AACOPO jun_j_Page_019.QC.jpg
a776257015baf3791a4bc792c18d3088
a42ae3de6dae7dfcd71f1d90ba462b7ad22f4d43
F20101118_AACNLY jun_j_Page_005.tif
72530409d4d3e1c31869f7c4b5f34932
18db3a7547e9cf65768d7e66f2457a8ba0354e23
6176 F20101118_AACOQE jun_j_Page_041thm.jpg
f1526fbcf7a91e18155cf6ebbed0c0ef
ca4f4970b392e4cf6e5e3e3cc3cbe7e4afe5e37b
46356 F20101118_AACNNC jun_j_Page_047.pro
e18f855e0b8fd86ba8f57fc0f3aa1fb7
a7684643b2a0c011612fede6c65aa6619d984462
89744 F20101118_AACNMN jun_j_Page_051.jp2
b25e89083b4f18fdf7d650a4efdb2635
ad94c7f9d9340f2d3938c75cfae2acb244805780
6059 F20101118_AACOPP jun_j_Page_019thm.jpg
e2fa344f3e91818c44e3e52c55691d59
cc6306b4dfac0a9ec3bf7e5dc67dcfc6ffc00735
21853 F20101118_AACNLZ jun_j_Page_016.QC.jpg
1f4b1d3c02c2d6d8ca91def8982ac8f7
a66590308d3f9a1f04bba43e18a5d2fc05b6877b
21661 F20101118_AACOQF jun_j_Page_042.QC.jpg
4bed9810c163d157f5ab69b5f1bab660
b4e9b95213c5333c28a53fca0b23d69dd8bd2985
25245 F20101118_AACNND jun_j_Page_102.QC.jpg
c990d28a0ccaa8d6d669b93cda992347
f50ac99c19b110c5da08d2586452ad82d19e1976
6487 F20101118_AACNMO jun_j_Page_082thm.jpg
89b639d9d7ecfe692f2e12825a363c4a
4d96c9c104ccf95e2a35e3b4bc0c78cee62359ab
22699 F20101118_AACOPQ jun_j_Page_022.QC.jpg
35d4bcffed1b2c569fc1319735c7618f
df9a48e248f21f3fb6e3a381e9c10344ae793a75
6325 F20101118_AACOQG jun_j_Page_042thm.jpg
58be834eff9e2b3518d0a8cf38e09c82
83b17ea2c1e91d707e1e8ccb4056d9771d64a9b6
6744 F20101118_AACNNE jun_j_Page_117.QC.jpg
f9e091509cf92b8d62dbf6f79f323a5f
45c82087a6bdce8817a8bfa5ccee4865f2e65967
31850 F20101118_AACNMP jun_j_Page_088.pro
f3e346559f86e27bdb280f9ea849f5f3
d7c978934afeda1bd0c3c294d65ea92e145d9167
24539 F20101118_AACOPR jun_j_Page_024.QC.jpg
aec8ac3bae0d115110799e783123ab1d
8cfa232c6dc1340ae5b6f7d7995913fc6daf4ff5
21000 F20101118_AACOQH jun_j_Page_043.QC.jpg
9c6e248e418704fcebed7bf108cfcc8f
ef4b683d406dca4a116ab6098ceabd2d33533b26
905587 F20101118_AACNNF jun_j_Page_070.jp2
35a362b88c2b6ed3e911260d319d7755
1720d53868b7ec9e7b4dc97df8a5b94d4b8ce788
22071 F20101118_AACNMQ jun_j_Page_103.QC.jpg
837c275abc061cc1909c27dd5ff112d8
3c2a1f88ab729229853d1c4e831706eaa6994496
6736 F20101118_AACOPS jun_j_Page_024thm.jpg
8a68884a8c1894d537ccf31131ab0052
342801e0617ba2d1beaec29178111251f347af21
22235 F20101118_AACOQI jun_j_Page_044.QC.jpg
1875f3883944a3f30136511fb2264b50
9cea6e7a40164f768992c11dce6268713ff37053
21219 F20101118_AACNNG jun_j_Page_081.QC.jpg
2e4f8d08180cb21c404b400fc726f72e
f32d377078f45e407dd4a95a631a0f1125d3b402
5246 F20101118_AACNMR jun_j_Page_040thm.jpg
7309abd280831f8a2490b892704a5057
2a01e8b9cd87229da2bc638facd29b293d58fd53
19845 F20101118_AACOPT jun_j_Page_027.QC.jpg
fa76caaa273d997038770de0782a8985
5d8ee9e4434a2a97c68497cdc9d6c967c20a6cf7
22882 F20101118_AACOQJ jun_j_Page_045.QC.jpg
9e667fb9fb219746c537dc795b983ac3
4433628eca246c6e4882a6a89eed0fb9c2cb343c
974358 F20101118_AACNNH jun_j_Page_084.jp2
995ebd3a651805c157592e7b66537cb1
d406c32df982893f6b8ea3a0a3432dca1929ebc2
1521 F20101118_AACNMS jun_j_Page_036.txt
9c732536e6a6d9a793a0cdc33bd8cb04
783439fa15167ecc3d1c2a75e4da286479712121
5074 F20101118_AACOPU jun_j_Page_028thm.jpg
4703faa3b6351d73202df4ceee5297ac
d6806357d4e8b4f0755b7bb241575259f418d15a
5085 F20101118_AACOQK jun_j_Page_046thm.jpg
7e70a4063e2460ae3135393b8f4654ef
e3a53337f43b6c7ed4aeb93f70b9baa290ecd20d
7644 F20101118_AACNNI jun_j_Page_094thm.jpg
2b24b6cb380489d9814f8935d9260fb9
e7ec1618c17e72bfa262840efe5dd482a05f6df2
35660 F20101118_AACNMT jun_j_Page_116.pro
eb14eb65aef2a0dade6566a85330d26e
1ca842f9c622316777098172258f26cabce36d69
18833 F20101118_AACOPV jun_j_Page_029.QC.jpg
56dd6d4421d581f5148a31e1b63616d3
4f52e2bb78985a915a931a15923caad87038f78e
22543 F20101118_AACOQL jun_j_Page_047.QC.jpg
3975f49a01c8431e6b6d24814671c892
70c51a4c0fb1b315356bf8948887d4e8cb10b6aa
5861 F20101118_AACNNJ jun_j_Page_101thm.jpg
1a0c4ef73edc4646c6ed218d828dbd57
18480bd7bbc5da096fce15ba24ce9cfe6132e0a4
2033 F20101118_AACNMU jun_j_Page_087.txt
060e7b3a3755af0225c9783613ff8ca9
a056912438412f733cd8edd4447adb9fa6be7392
23064 F20101118_AACOPW jun_j_Page_032.QC.jpg
a0d0dc44e6c4ff0d6b317cee493e0062
76aa31063e616c5e0424d87017968cc1e70db3e6
20978 F20101118_AACORA jun_j_Page_064.QC.jpg
31839179d69d92f9917083dccfefcc0b
a18439ac9a3a9867e4de5c2372d671ecd6f80c59
6296 F20101118_AACOQM jun_j_Page_047thm.jpg
4f7c73fe8a4b71378756bb3cc804c37f
fd348233dc9cd8e5c54b2d2a5677f304b1774d01
19264 F20101118_AACNNK jun_j_Page_063.pro
77873b434afd1de0663304103b746a3a
f77002f940316312a9c42250c72094cc8338a99a
84556 F20101118_AACNMV jun_j_Page_054.jp2
e36e46ececd769fb2fb9adb8c0826c64
7772ac95205561084a4fd5fcd89990509cd3443b
22129 F20101118_AACOPX jun_j_Page_033.QC.jpg
0ac7e5afd78129fbce9ec6f521445a92
967487f7d686bcc94eb0467624ad6f7391122587
5981 F20101118_AACORB jun_j_Page_064thm.jpg
31009205581d3d8a8b8b77fe7b7b62a9
f120d8e3d08ee3a4d3a0f5b0fbbbccbe17098130
6090 F20101118_AACOQN jun_j_Page_048thm.jpg
3daad9ac1a580be0e530733e957946f7
73fc722de466db3fdcce34ee75315d4ec57a04b7
46423 F20101118_AACNNL jun_j_Page_097.pro
e516f06b00d4407451de085693e67c70
f75d098a76db359497fd0ea5fafa19af262269f8
F20101118_AACNMW jun_j_Page_077.txt
730f361a3c65490332f56c617af202aa
88e39ef823063672d5b7ac7c6ef737becdf5b45d
6441 F20101118_AACOPY jun_j_Page_034thm.jpg
f52a27db6101d390af2b651f58509a15
c0c585ae47b3436742ba5c1f736b2a9e6807f03e
18939 F20101118_AACORC jun_j_Page_067.QC.jpg
a413df8064015beb8b5efb43f1c89b2d
f5a2a118249ccfe8f13a1554e194962815817985
4782 F20101118_AACOQO jun_j_Page_049thm.jpg
1b07382069f2c22b47d358081b0bd266
bbdd0e1082ba222d77c840d189d6f594b5b74cdb
68248 F20101118_AACNNM jun_j_Page_047.jpg
8fd43f24e5d07d6dc7892d6401bccdc8
117e6924621ffcd86945f3d5d526694d8209aa08
41476 F20101118_AACNMX jun_j_Page_031.pro
00f0f9229fd998a5326f28340cd06f80
152781a768252cd8ee1860235f3e01b1dbc736d9
4449 F20101118_AACOPZ jun_j_Page_035thm.jpg
99c38ab149059c1527abe2d3f9681d48
ea612023cbb568ca10741cc23bf38bd979b70533
898 F20101118_AACNOA jun_j_Page_057.txt
db735c8a7f8ae94fae0d69fec256ee91
06252fd028dea08fcdf5eb00052f158a33bb7f5a
5626 F20101118_AACORD jun_j_Page_067thm.jpg
9ab2ebec214eede4f08abbee8e97c590
91a1773675317f29807ab05f7fe4ad0438601a9f
4753 F20101118_AACNMY jun_j_Page_055thm.jpg
4c11cd63163271fc2f575151233d8be4
4f282776b6e30659220fa38f403f4183b8221d73
52840 F20101118_AACNOB jun_j_Page_112.jpg
5f49e7f518f4ec067755300ae17cbdd7
1acb83c4c86514b7750d888267a6f4db579f251d
22429 F20101118_AACORE jun_j_Page_068.QC.jpg
1a2a54e70e8f3262dab43d27cdca6124
0f5479fe01d03607d27196474442301b17002d3f
5141 F20101118_AACOQP jun_j_Page_050thm.jpg
b499ab3cb3c2bcbd0e2398b1995eafa0
21f4339eff0423b8ba01e66f9104220ad84148f2
595 F20101118_AACNNN jun_j_Page_072.txt
b68499717df4e4d83134fdfff51f0dc7
35d61f36fded64c183461ff73ec83b29bd2a6719
98512 F20101118_AACNMZ jun_j_Page_108.jp2
7b12dbbff9afd99ac7636b2122560fd7
9d7484f45690ee17b39e9b83228a483c5253b0f0
18851 F20101118_AACNOC jun_j_Page_020.QC.jpg
aeabbbcf5511be4746d442214d3768e6
617bab23611b5adbd426091c8ca53ffa15a68baf
6411 F20101118_AACORF jun_j_Page_068thm.jpg
7344f29f61c1542ea9082f959cdae7ac
aad9a2220ff6fb381719783ef95770e5b83f5825
5690 F20101118_AACOQQ jun_j_Page_051thm.jpg
48540cfb414e1fc61176f9e34ba798ed
965d7b4c380fb7d111d1dc2e38b3a5cbff645f94
6636 F20101118_AACNNO jun_j_Page_053thm.jpg
b8d519bd886d8ce13e6cf2b29335faaa
5242b2ed22cfa8c091477416e80a428403e798b9
99194 F20101118_AACNOD jun_j_Page_048.jp2
05b80d8e50682052d61e8f468cd9ab00
88b5de5326886675597b201d72dff6160faa5686
27049 F20101118_AACORG jun_j_Page_069.QC.jpg
c8f9d88ba92bac1f6c347a1d25a3c3b1
0f72dcd2e85f2bd9f9b5ec8ffbe34076ea8e76fc
6283 F20101118_AACOQR jun_j_Page_054thm.jpg
515c30a4774d36e696ef76f150e94b5f
1e02021a6a84612c074d2b808ac87844304ae7ae
2074 F20101118_AACNNP jun_j_Page_053.txt
07b7f8b7ad8bb1a41d05b149acb5e431
5a5ca37d37e364fda11ea27ca70d73657f685608
40836 F20101118_AACNOE jun_j_Page_081.pro
b5c2a76740e5e8e5c7923d1900b674f4
65fd07e5cd26b289bfd4f49ff5e27ed59a590e5d
7040 F20101118_AACORH jun_j_Page_069thm.jpg
b16748978dea5267012a0c1dc17a903a
7cfd19715683fb55fa55a4be4fcd257c993ec52e
6399 F20101118_AACOQS jun_j_Page_056thm.jpg
af100ecfaed23815b061a24d129bcd7c
912ceffb87f7e700e9c79a1d43a636b74f194f56
2007 F20101118_AACNNQ jun_j_Page_022.txt
c59686899d8a93558bedd9f266b7b448
d195b2fcfacdaf0b858fb11be97dd992c325be06
F20101118_AACNOF jun_j_Page_056.tif
6c170bf4c9aecdb7930e7d129310c4a5
07a7f6b0bef47c8643f36757fbb8c05a662216a1
5817 F20101118_AACORI jun_j_Page_070thm.jpg
547d80f61b30e95a4cc0c2bd7a7494ab
5d704119521564d6d972d87edf4a5cc18f2920a5
19077 F20101118_AACOQT jun_j_Page_057.QC.jpg
1d4b9cef5be50491784c52f6dd3f39e6
df517a6549cbc7c37acefec657b9339bff2886d7
30214 F20101118_AACNNR jun_j_Page_109.pro
c48ca4b183fbaeda21a5499df518eb79
7788acb9a2be38fb33be031b113193a7d8a60851
59713 F20101118_AACNOG jun_j_Page_043.pro
ec057176661e5390fb534c681338ece8
b292f2a6fbf85aa6f84e1f2f2a2a54ddcce0b43e
15644 F20101118_AACORJ jun_j_Page_072.QC.jpg
ee257e59c236e2f9658aba4f0caf5461
c37cb84e8bd5de95daac2a433a33d21780ee3220
12907 F20101118_AACOQU jun_j_Page_059.QC.jpg
4292d2910bdc514cc323910e2effb9d4
487033f863a01c63c73715c15745bf5851a7433f
772 F20101118_AACNNS jun_j_Page_063.txt
c2fcb545053bee7bb702cf65c9de77ea
296609b31094ac631dc7e97e9c15f5bd824f9a39
67933 F20101118_AACNOH jun_j_Page_045.jpg
c6847ce7db00dde2d166fc5c5f7197c2
fed56f8346cd07a84923c58e7e5f097d1b204b77
16152 F20101118_AACORK jun_j_Page_075.QC.jpg
5af394d6cc40bb3747b3d217518c3d26
7e2290fe259e0817ad027bce271dd0c7acb80468
19989 F20101118_AACOQV jun_j_Page_060.QC.jpg
b7bc483996de70151e7099832d9e9536
c25f7c5c0a4e39efb5a39f34b4e717453811f4f4
81977 F20101118_AACNNT jun_j_Page_036.jp2
2994baa84ffdc53034778f410da11bde
fe202b4065f7639c5122cf93e89a8d4124e32476
22357 F20101118_AACNOI jun_j_Page_001.jpg
36e8e887b8ce3b9be3ef60daaf286fac
02be2aa1f5437475d7d24db18e90c55f4f587cec
5385 F20101118_AACORL jun_j_Page_075thm.jpg
754427721a9eba5b9d58361d7441a033
426f048973600eb05cd84e9b4ffd7f67e0eb4e43
6734 F20101118_AACOQW jun_j_Page_061thm.jpg
568bd975221ef6028e6bdcfe46928659
689f50acb2c2fc63050ca5e31c8a6f3784203940
55843 F20101118_AACNNU jun_j_Page_036.jpg
36112291e03c921cb34bd9dd1398c104
6f0f246fbf0f478fcad4eb55c733114328fa70e0
17144 F20101118_AACNOJ jun_j_Page_089.QC.jpg
4a925a102d1ce5eb8f30121e9eedb1dc
2cd89fd1f08f53ac0c8b62de9f87acd2231dcf3a
6333 F20101118_AACOSA jun_j_Page_093thm.jpg
42e0a02408d6dca28ad120853f7f8950
27301d543ecf71be43ac72fbf09491186ef2c920
F20101118_AACORM jun_j_Page_076thm.jpg
9f40a69a7866e7f40ba1c63bc2ed3a5a
6f36628656f17b4718cb300922f74756caa9ebd6
18045 F20101118_AACOQX jun_j_Page_062.QC.jpg
f7d62ec2093497a45f08bae1664631f5
6a1c524a7f0f7c6dd26d2fb89d5ee0964cd67985
63267 F20101118_AACNNV jun_j_Page_021.jpg
84e6b46838eb6b9515d98fb65fcd41c6
52ef67e669ce84ab2d7fa0a9f864895f8adef113
2496 F20101118_AACNOK jun_j_Page_119.txt
d9842425ef888a2ff3faf7859403bc0f
9dfa140ffb78e4852dc1d5f0b9f8f430211b424f
19146 F20101118_AACOSB jun_j_Page_095.QC.jpg
fce2917c9b87e99c2b054bf0ca93d51a
108bfcc4a3b813dcdc2e596da1809a632f876292
6511 F20101118_AACORN jun_j_Page_077thm.jpg
e49c284b0af4186f6d3cad60ddb78b39
b166e590d5c1807f96121bfe218ed259fbef12ac
5466 F20101118_AACOQY jun_j_Page_062thm.jpg
c3caf93a476f902c2749460d79d8596a
e6d477dbc95637151052fffa368060ce4e52d4ad
F20101118_AACNNW jun_j_Page_052thm.jpg
8338b9acd295d69811a981d23b76c37c
16d89c45bd7ff37b9f63aec51a0ff08dbdf20dee
69650 F20101118_AACNOL jun_j_Page_007.jpg
5f2d56259d34495974275b5f2d67b579
905b638be5f6409b1ab0de8ee9a037b0d43c95c5
21839 F20101118_AACOSC jun_j_Page_096.QC.jpg
522273cbf6f598172ebe78209d0e363a
0a88601ea6b911c9fc62bcdacaaa5a562f22c845
19067 F20101118_AACORO jun_j_Page_078.QC.jpg
b31c71d23d878e0f87b4331e7b1318b1
f5c9a1f12dd993b91e8c188b7d68391001c4dbba
12938 F20101118_AACOQZ jun_j_Page_063.QC.jpg
35dde4cf237f1723cc70ae2821506dba
d03e090de290e296cbf182b15e1a92a6ec84fec5
20019 F20101118_AACNNX jun_j_Page_088.QC.jpg
d1beb87f04f010f39a2d9e48228937f9
b1185003a46a58cf65d38f79e2fc6137aa89f67a
5719 F20101118_AACNPA jun_j_Page_091thm.jpg
80a5c6a87c3105825c6f4f1a5a1146e2
0169610f8681807cc0a4b8e9597065a9b539a646
7696 F20101118_AACNOM jun_j_Page_001.pro
e24793601c87ada5b251980aee543b0e
be39acfe6e4703c5ee68045ad69f928863239ace
6358 F20101118_AACOSD jun_j_Page_096thm.jpg
c6b6dd25ffae7ad50af15d3335ef008b
9a2bef109c36bcf301943d84aca373add47df4bd
23210 F20101118_AACORP jun_j_Page_079.QC.jpg
a74d743bea4811328a2514f5c068b7ae
222731d823106a644759492ccae03152e777f8b1
5895 F20101118_AACNNY jun_j_Page_043thm.jpg
95e60a8ac01249d71d139caf8c534689
9fd9f38d75e6909c3e111b3dc4c62552c36a587e
F20101118_AACNPB jun_j_Page_110.jp2
41a90c7142197972b2a81cf8d31b2c8c
0eb6f67388d6a1418b605bb55280ad9f67ea75cf
23011 F20101118_AACNON jun_j_Page_118.QC.jpg
7f029dd6b8f6c7a688f416315bfd4ad7
b8fd234b4dcd6e88ffe9c8ff368231c1269f91a8
6269 F20101118_AACOSE jun_j_Page_097thm.jpg
0419c542c3ebff6336fb3dd548f70e8e
74b226176bc0de0d2c74333d91c9329ad779b2cc
66111 F20101118_AACNPC jun_j_Page_064.jpg
ed39a44e514b67155e336567f56997e7
7d542cfd3a562de9979ed8f48452a383c8868d73
944409 F20101118_AACNNZ jun_j_Page_031.jp2
b2c584f79a4c1cab708f09500fdfdafb
4d15bd5e05680a8469e87fba9dfece50533fef0a
6440 F20101118_AACOSF jun_j_Page_098thm.jpg
2fa5aa688d240e16f274cac1ff0e1c99
33ff4fe88e3c705f48462745fe38b1af7d5fdddc
5614 F20101118_AACORQ jun_j_Page_080thm.jpg
826d04dbd0cfc8ff19d0f5876508173f
5cbc8b314b4b9dac529085833cff22bf817af71a
53039 F20101118_AACNPD jun_j_Page_028.jpg
0f5b39b801a6e32d5d4378debd5b53ed
05fc301969f73898b1d08a5d28472c7bd3647b96
F20101118_AACNOO jun_j_Page_102.tif
06a1e0c0b1b411fab5ca8515efca5b01
e293e5b4b90b1e80bf213285a16dbdcfcb85dfde
20595 F20101118_AACOSG jun_j_Page_099.QC.jpg
0a6d74159523ebec189d033d440c33e2
8f147750535a574548a9dbfc87921a43a4aa7620
6123 F20101118_AACORR jun_j_Page_083thm.jpg
2737fb4f22b5c18b54cc39fb4e12d95f
62f762eb74efc5fb686fec3295009906e0b90704
F20101118_AACNPE jun_j_Page_023.tif
1ace3327e185b0a9145c0773aa29ee5d
55c12b2d372b1049e38f48ebc47db6ff618743f4
F20101118_AACNOP jun_j_Page_004.tif
da959c9729358def292ee33140d96187
bc4b4bde787e98332c56c392bd3d12a7e19bea71
6033 F20101118_AACOSH jun_j_Page_099thm.jpg
2db9d610b4237d307924d84eb9e89af5
92b29fa4d50346848cea2d07bd61f19e04a17951
21491 F20101118_AACORS jun_j_Page_084.QC.jpg
405e4967baf6902d0e54b123a6eabbe2
4f4c25cf1b3ac466b88d00496b9b1de49b59d5cf
2344 F20101118_AACNPF jun_j_Page_001thm.jpg
b48cc48d6a3ecd03bd75f7a5f7ba903c
91415b716412bb729a18434a273a92068a5043fe
884759 F20101118_AACNOQ jun_j_Page_019.jp2
0eafd2ec16ba0f8e4c95eae92ebf428b
122b3eb5cea15ccb049a2cec4c084f6c5dd98e08
6082 F20101118_AACOSI jun_j_Page_100thm.jpg
c76225cf528320c74e4499368ad09f18
440e1d7eb2e967124dd3281caa5a04a62141ce81
6238 F20101118_AACORT jun_j_Page_084thm.jpg
fd250c45515e182e636be1f8f1f7a912
64ee8bfac7e2d225a919549168f90d66f407223b
68763 F20101118_AACNPG jun_j_Page_010.pro
b4e7104e398d39156bfac53f470bb944
e8f4d6ffea46475e0bbdff90c0b91f179e7df562
23327 F20101118_AACNOR jun_j_Page_038.QC.jpg
5e3f5ec6fae236aa1de6bde518038246
5420c9ac5f9a4a24fe9b0f999a8a29fecdd95c92
6650 F20101118_AACOSJ jun_j_Page_102thm.jpg
34a14fac71f47fe2a60c2478fa6a37e2
f02ba9e7a4777fbed98ae34dceb240651988a803
23284 F20101118_AACORU jun_j_Page_087.QC.jpg
73f1d9062adb5253da06b4d42b16fa91
5d0accff46c9a8832853b016d7647b859f886828
102494 F20101118_AACNPH jun_j_Page_005.pro
d5bf972896e2e6bb3fb7278bf4bf4ff7
dd3e5c10aec99e06ea70a8bdafcb3f21393aacba
80873 F20101118_AACNOS jun_j_Page_118.jpg
1da1bc74ba124e2e0f5f48a53b846115
62e6be84b97b483eb101c7fec6ce0b4fd040652c
6075 F20101118_AACOSK jun_j_Page_103thm.jpg
e7f8c6763a90b76c8c76c209ff4acce7
47e7bf89eda800d099a982abcffc85d36e165db0
6350 F20101118_AACORV jun_j_Page_087thm.jpg
caef31c7a7459e0c21892cd55e289f89
eab1695cdab947b64c209cc640c45a856a274e6d
1850 F20101118_AACNPI jun_j_Page_064.txt
8b9e79815df7ea149d076d7cd8b11b46
f887089286e1b6afec71b3006aea86639daf091c
F20101118_AACNOT jun_j_Page_029.tif
d2d4519bdcdff28bf58f510e7f315be7
c4488339972d8eb7da2c48b0a33c0f70735a3101
5461 F20101118_AACOSL jun_j_Page_104thm.jpg
6b3fcd78ed27a5373188304fc4775660
9738d2b4b4f943a297449bcb2797329b46dd3933
4998 F20101118_AACORW jun_j_Page_089thm.jpg
2ad81306d70e69516aa97270a230cbcb
e6d0b221c9ea2c419de816af451882a1a456e8d5
5632 F20101118_AACNPJ jun_j_Page_015thm.jpg
9aef542d3fe60447a58f9789f1f1ea34
ff5fa91fdd64438cf77e1ad06014607ca317ca47
16511 F20101118_AACNOU jun_j_Page_055.QC.jpg
8f0aeb3b7b85fe9870dd901530665bf9
8718fa14e2392e3b503e8a51e004e54ff7d43d20
28393 F20101118_AACOTA jun_j_Page_120.QC.jpg
c52a14c56b8e77ab614d9e9239e882dc
13d438791f9d2af8824f7e84364a595a528282cb
5837 F20101118_AACOSM jun_j_Page_106thm.jpg
13143e6d31990189053b3633f5d492cd
efe44e7b4ca6f22c82897e7df98d6817dfbd0201
21747 F20101118_AACORX jun_j_Page_092.QC.jpg
cc338bea55919d58dce24535b813bdbf
15f0c1e77228375f36fd082bc9421d65ba92a295
F20101118_AACNPK jun_j_Page_118.tif
b614c794f8902ed90e12f83139123b27
79c5e2e78fac394f4bc688aeb7ec9b44c0458926
F20101118_AACNOV jun_j_Page_018.txt
b0d4478fe0cde276ec314a265c0c55c8
c36c331ada7d7e2e48df4d5ed1ca566905124f34
7274 F20101118_AACOTB jun_j_Page_120thm.jpg
37839fd09aace88709db19ca2bcda0e3
09934793144cf1580a1bcd1e344657b8f9cfa482
2971 F20101118_AACOSN jun_j_Page_107thm.jpg
d4fb646ffd8a067158abc8ad2b07fa58
954c74d4e17bc1d73c1a8713ecfccf47f117b3ce
6270 F20101118_AACORY jun_j_Page_092thm.jpg
538d046f9235e9a7dbfb0cc7c14f6d0a
9c1961a3a9686c63bb92ecf5007b445ad5bc39bc
68676 F20101118_AACNPL jun_j_Page_004.jpg
5eed68c59946675ae7e69da59dd0e5ed
df54dee3d24702fbb62b66140b8f98b840850395
22167 F20101118_AACNOW jun_j_Page_006.jpg
3c06dc5468f24c1ddd290f92015e91b5
3015454ca75592b33c36bb662a80e0aa6818e612
28522 F20101118_AACOTC jun_j_Page_121.QC.jpg
ff8d6715c6c76a76bb3ccfdaff316e73
285facbbcf297110f28fd387371b3077d7b278b2
5898 F20101118_AACOSO jun_j_Page_108thm.jpg
0c20b5cff5dc8843defb3ac8930d8c30
845ce5610cc2e71f00460802ae9f33476770d3e7
22136 F20101118_AACORZ jun_j_Page_093.QC.jpg
1d4149b967ba16a74bc80e2e67eefd91
9aa906214d2fd8d0dc83ecd902d205a481489fd9
1864 F20101118_AACNPM jun_j_Page_068.txt
d4a5d39ba3b0b6c8f94157019007ab21
b8bfd0c52374f1821a5dbf8dbd02d5c2d099e4b3
92656 F20101118_AACNOX jun_j_Page_018.jp2
9e80c88a316c02a3aaea3c586bceb79c
76be4c1809357c035e67fc1a870174d64487d1df
2453 F20101118_AACNQA jun_j_Page_093.txt
7825bf3cd254c905f347f093cd34b3e6
057a1ac695ef3d14e471e6b273cbfcb75e0e1a9f
7137 F20101118_AACOTD jun_j_Page_121thm.jpg
50d559fcd5258afa339ed393eb7df2de
ad2cde8e27ab57cbb0df3879fa8052ca50e755af
17571 F20101118_AACOSP jun_j_Page_109.QC.jpg
7cd7f0c7ac83f41dd5345b35fbe45f86
d329c401e60763aedc62f2f64f8b511cba77e1d4
626 F20101118_AACNPN jun_j_Page_039.txt
c0eeaa2997785e78c351371ff7280591
557aaf27623c2d80c6702623a0d93af7976f19dc
67043 F20101118_AACNOY jun_j_Page_103.jpg
a5e636061df0b237d1135ac37cc2a3ed
c848ea71d8dcbdb16369d083c862738d9ec93e95
6442 F20101118_AACNQB jun_j_Page_031thm.jpg
cc7a76892583ba23d824fbd6b6e26702
1daec0581f3334990c871dd926284e67e0d9acee
5613 F20101118_AACOTE jun_j_Page_122thm.jpg
0b16895fa475eaf1a184f9bc6cd4cfbd
8d0324649409a994ee12a5528e0ab87e1ffcd0fd
5195 F20101118_AACOSQ jun_j_Page_109thm.jpg
5dcced9d55095de2bcc84b21af7973f4
bad4115023762f913aa829732a8cb01a22b817fc
12458 F20101118_AACNPO jun_j_Page_012.QC.jpg
bb40185bf356ebde761212813feaf8ca
72dfea59d96c3b19f0b02714a7039540a0ab98f9
782796 F20101118_AACNOZ jun_j_Page_088.jp2
8c9697b1c2f7c3cadeb008b427746337
4566271706004032998c37f5979fbe07007d043c
F20101118_AACNQC jun_j_Page_052.tif
08def23e396632bebadc6c4a93f0cc58
34180b6944ecb51dea265b05430dd4a9f373d059
10603 F20101118_AACOTF jun_j_Page_123.QC.jpg
d3609c791f414133693bcf96ace8280e
0f52d5fb5bd1158c42568e77c7ee5d59cd622351
F20101118_AACNQD jun_j_Page_098.tif
baad46bfaf6215b91538410ebb2f8749
948a2d95739727abf73a375fea22dae119fa4492
20824 F20101118_AACOSR jun_j_Page_110.QC.jpg
51c0df6217f1b4b128ba26bff09e0963
3549a9c5ac98551b167fbaad971b66cfa6b9af19
64270 F20101118_AACNPP jun_j_Page_092.jpg
84f3db3512c7c9b9fdebb6d20502cfaa
b1b7a4127004e0cbb724a38b2904f6141bc085df
21091 F20101118_AACNQE jun_j_Page_122.QC.jpg
7d0ad8739259b3525a11cf304b530162
11ef31ea6b3af42acf3b24e9c0b46b686275e1ec
18141 F20101118_AACOSS jun_j_Page_112.QC.jpg
d6029526e3feea8212c395a7b1efe482
893db6cb018021c1001b5319be04831d8c125748
66112 F20101118_AACNPQ jun_j_Page_100.jpg
d9d7acc35eb98a39c417ea1b7a1f22ba
05a9ae3345ccd8a99eeb19f876b7da61b3adfa58
37941 F20101118_AACNQF jun_j_Page_016.pro
ff6f814fc3f994922f7a51ec13c7eb54
8655a1ad08924114a540bc2885eda32aab8c1d04
12312 F20101118_AACOST jun_j_Page_113.QC.jpg
377db22c1bca3202c081c7bca66e741d
850e2aa59fc004843956c0a9458630a41170eabc
57751 F20101118_AACNPR jun_j_Page_029.jpg
76a8eb0f6080de0496afb5a7e9ab7e16
70e4068f503be54e625034990f0a2c15c1e331fb
33695 F20101118_AACNQG jun_j_Page_078.pro
0247b8ea859594347ffc766a908c33bc
df3db4e4adc3a68204895528f518480fc59f32c9
5741 F20101118_AACOSU jun_j_Page_114thm.jpg
bf42def48cefdd7a5398070dfad9e46c
eaee272c24fe9462fe95219708bbdb511cb824be
F20101118_AACNPS jun_j_Page_022thm.jpg
ca018371c84ca8682a3b42092fea05bb
64e1959573f2ce110f0e89edb3e32e2fcfd008cb
4876 F20101118_AACNQH jun_j_Page_013thm.jpg
1ee91cd6e91f8fbd316d8b359cfc6a7e
6b5c351e881429c7b43265bcb69a976085fefd0e
22911 F20101118_AACOSV jun_j_Page_115.QC.jpg
666918a63c3cf41b7f75e44671608c5c
5606cc9ed53f63f777831b16361bd971b30344b0
93197 F20101118_AACNPT jun_j_Page_061.jp2
e2c334eddea834a776c5441187185c8b
15168c30e121aac06a292f97f87032f28b7c7333
F20101118_AACNQI jun_j_Page_112.tif
295e4d524252bc1073876029c422584b
b3dfd933debac3fd6549856c057f3d4eea277038
23750 F20101118_AACOSW jun_j_Page_116.QC.jpg
5db46a3abde7770e86031ee6401fc31d
19b49067586221dff23fa31dab8986174214d51c
47698 F20101118_AACNPU jun_j_Page_102.pro
ef298d3f080768703be7bef41b92b966
3c143da779168cdd4e9b93d905adc1dd74fe25bb
5766 F20101118_AACNQJ jun_j_Page_095thm.jpg
7ead78f5d3906557ca0117ca05b292b5
e32db2835fc4d0b2ba315596ae991d4755faca09
6834 F20101118_AACOSX jun_j_Page_116thm.jpg
fc5ab0a614af3f3dedbf31e140a5f9d7
64020aa88349aed62dbda819bc2c171f731c1396
1062 F20101118_AACNPV jun_j_Page_071.txt
ebdf9d4cc6fedeb398f03d67f36cb65b
675e9d1d9a0bc80421b401307106de620f07cd5d
20219 F20101118_AACNQK jun_j_Page_051.QC.jpg
985180124ec403ac2b22ccd3365790c2
2a64bc950e68d72d559e52e8ad542a8369d01531
6320 F20101118_AACOSY jun_j_Page_118thm.jpg
b47d98a335c5482ab80cb3a1986d03b2
c6552dba5086ec624f4b5a0aa1fb1ef0a5562718
908726 F20101118_AACNPW jun_j_Page_073.jp2
cf3d89e6a3f624b93a1847f5f86e02a1
ed61ecffdbaded222ce2a89684fe395be7843beb
36057 F20101118_AACNQL jun_j_Page_113.jpg
28a1161c3c57b2f14bb5439c9fb11931
6ce8290fa602a7ff3a545196c1ef1ed6cb54945a
29650 F20101118_AACOSZ jun_j_Page_119.QC.jpg
80ca9ec93592655c08f67b99723517d7
90423811f8003f34fa2a26976aebf1eb40869aeb
69954 F20101118_AACNPX jun_j_Page_065.jpg
e13adc71a2fbb694f4a4637c3e7ae93e
e6a470bf0ae813a5a6a9d44494d21501bee665a2
20584 F20101118_AACNRA jun_j_Page_009.QC.jpg
1bb1d9964045479a494c9fe12cab4907
35a1fbae2ab48ca139be7331cdde2ff310e652f6
68465 F20101118_AACNQM jun_j_Page_096.jpg
8f3c26e305821dd03be005a2b8ab2265
ccb573bace54c99d7fea2d8031fbfb185223cce3
70282 F20101118_AACNPY jun_j_Page_038.jpg
f9d3f742a9b8c47d12d4ce4f3cf79298
cff1a3e97156457a1effa3645f13ad120a921a52
16049 F20101118_AACNRB jun_j_Page_085.QC.jpg
4358ce6d123e135902c0aaf2fa526802
a80748dacd852cf14db3f0f1c14b23a9fa91134d
879546 F20101118_AACNQN jun_j_Page_042.jp2
7f8227cec5732856f1be30930222c189
f98a38c360a3fe1ce5d3de240cdacc0b1b92cf40
867695 F20101118_AACNPZ jun_j_Page_099.jp2
2e7c5479a1c96249832940435af48269
58ede6a93cf7cb14c3866aa3098b38a455f0f607
6353 F20101118_AACNRC jun_j_Page_044thm.jpg
26813db6e8723b15649487c1475a4297
3b59a2c3f37e5f8035c95892a645234939b9bb6e
36636 F20101118_AACNQO jun_j_Page_070.pro
a58bf7d89c30d42046b13adc9b30d88d
cf5c10fc1be0770b8d965ce24a4366015e2eb136
21265 F20101118_AACNRD jun_j_Page_101.QC.jpg
4c42c4654683bc3c6976e406c865074c
e6e927eb598f554941d81cfaad04ef39903add3e
31560 F20101118_AACNQP jun_j_Page_083.pro
a973e8f622a8c83e32e70d20c6b2d85c
08e9b901f080c2430fc9986127e28d694272e52d
F20101118_AACNRE jun_j_Page_001.tif
f93b01ed77d94c5aeb21240808b3c7ad
c0e6f32c2903bd9c3c20189ce547c896389daca0
1632 F20101118_AACNRF jun_j_Page_074.txt
bc3a1311c5183e0ffb27346b2607ca5a
58476eb29b04775a4534f7ef16b93feb2a6c6f58
19637 F20101118_AACNQQ jun_j_Page_020.pro
f5f56c1b497556d03f3e99a2d949dd55
3940abb99db3f148da5a93cdc7cd888ce1abed97
1051949 F20101118_AACNRG jun_j_Page_009.jp2
5eec17e282dcdf7dc5c290202d0a953c
a1f945aa36ac4d0b8a27bda3d900ecdebbb3d165
3303 F20101118_AACNQR jun_j_Page_032.txt
237fdab6be2e97c4659ea0d9be232095
5ddea10d1cd3d2d827c0ae54806213bddab7e370
F20101118_AACNRH jun_j_Page_082.tif
60722f44c1cdc96c2b5d36c13f521385
f8de8dcbeabf70868a6ec8e3588459ba27731718
1532 F20101118_AACNQS jun_j_Page_100.txt
cac04453a0c703468dbb9e0974d945ba
84149ff762e12fc99fdf7c2dd6e5c9af886fbc17
2554 F20101118_AACNRI jun_j_Page_043.txt
11b65d6959daa32ff61a48c8f116dae9
cbbcf10668f1b3aa48fb8cacbb935359556577a1
103715 F20101118_AACNQT jun_j_Page_026.jp2
9308925b690cf84a09d102de5b5f19ca
2dd18205ae64763e09aab1d8656134634fd3aae4
F20101118_AACNRJ jun_j_Page_114.tif
4d9db6df086daed225a75dcbdc641f8d
77b61cff7f9f1790569764f708dd29b771a4607c
63313 F20101118_AACNQU jun_j_Page_095.jpg
9d26c510d89ebe7e2a447d52e0a14e5a
d6cfbee0164e890596e46b27dfcbc21b906c829e
31469 F20101118_AACNRK jun_j_Page_003.pro
671a7f72638b0312a1fc38634d52e130
50bd86bac723d3541d7400fca6c5e5be22aa8cd0
14877 F20101118_AACNQV jun_j_Page_050.QC.jpg
5860da0c9fe12a374cbb092a8f4eefae
b62acdd835225b4534c88bd6b3af19a99135aa92
5977 F20101118_AACNRL jun_j_Page_021thm.jpg
3ab3282a6b003d85c7abe57be48066b0
656dd07f912db85fa2f6183e74058d4ae5979895
13416 F20101118_AACNQW jun_j_Page_039.pro
23fbc24475e862fb1a76792b21c2d286
c3b8577575ad734b7368c95156e18f81367b09a9
F20101118_AACNSA jun_j_Page_034.tif
2418b841cad8b818abdbbd0ad5683028
9b644d86230bc142c7410471d2145feabaa2558a
F20101118_AACNRM jun_j_Page_117.tif
7ac0ea85a60e777d91c7a2a3fd8c735a
5cf18f6336a54f6ce9da5ef673d60b5b30716a6d
47214 F20101118_AACNQX jun_j_Page_053.pro
0e19b4a4f7cd50a736727640b98f3a6e
1c6a0922282e77b69d36438cf945351e46a420fe
822 F20101118_AACNSB jun_j_Page_020.txt
2e14c6adf40b8819c7fcee3830ae1068
548dc2d9c1e6b2008a1faa180f684721ee8cbad3
3281 F20101118_AACNRN jun_j_Page_123thm.jpg
8d8a9227205a0fc726e7154f95357f9c
32d958d52c7fc7e06971f9ed4cfc355a24617cb3
1051963 F20101118_AACNQY jun_j_Page_094.jp2
55f2bd4cc7ac42d02e112f8efb537737
fe35903ebb3ba51e081341f294a772c7174b30e7
46600 F20101118_AACNSC jun_j_Page_022.pro
a167a84cbe69f318c9befddcf9deaf96
3c46091963390d25f42b6cceae7586067bb72e6f
F20101118_AACNRO jun_j_Page_064.tif
d0e521951536671ad0976d60e8ba23af
5d08ff858bfca9a46f395741ed8db7cbfd1e8093
21400 F20101118_AACNQZ jun_j_Page_111.QC.jpg
4c3cf44aa2471f98cafbf0d295775b30
0b42693c176cb9326db3ffaee9f7fc90ac40a0d3
1083 F20101118_AACNSD jun_j_Page_075.txt
4371fc7707327b1d44e08c5ef1214c1d
ec12fb9e59d7018494b8323288fed475f1e15fcb
62698 F20101118_AACNRP jun_j_Page_120.pro
9492943f8007b4b6751d9e550d021b0e
97595af0241e7254690a4003fce398f6edd0ef18
101682 F20101118_AACNSE jun_j_Page_115.jp2
404f1f134ffca424d63c51ba971afc0b
61d253882777597c371bf2bbde9557a2ffec9fd7
77309 F20101118_AACNRQ jun_j_Page_028.jp2
8b14cf468c07bc5da77e01b9091332fe
142e52ef3b4e2bd150fb9b0cba9c90594316f24a
5819 F20101118_AACNSF jun_j_Page_057thm.jpg
eac552638c544d144fa50e406d21769c
53d6ea7e5738763db2e2cb118572568217f1ed83
6603 F20101118_AACNSG jun_j_Page_010thm.jpg
9b9f5d36b812d9d73448ece363c2305b
853ac47171c803882210874864981b7ed9fa4d2e
20820 F20101118_AACNRR jun_j_Page_117.jpg
f4dc72ea9adc2d94576ba7a70e32c6c4
62f607f744d0106c1ea4037045ea246a720e25a3
4062 F20101118_AACNSH jun_j_Page_113thm.jpg
4100701ea1b4571e7dc24fe29cd33b2a
3de88c60680ce964e6a6033e71abc02db18e18d9
27616 F20101118_AACNRS jun_j_Page_012.pro
7d18a35594756cebdf9024f0ed8d5024
1d2ed384bf94cfc14baf7f0f4ce0cbc1e92adc0b
30747 F20101118_AACNSI jun_j_Page_085.pro
af55196926f0710bc40fcec3a549e13b
730e18e2c883f66ffe460cd9dfa67b02ebeeca3a
F20101118_AACNRT jun_j_Page_069.tif
a394bd7d0cf11ce2ebab3c8bc737b1a1
d545589d5a77483bee65884d1134f75d62d79537
863193 F20101118_AACNSJ jun_j_Page_089.jp2
bbb99b56a8462ee52343d17fb018fad1
4ae1fdc9d711d8011b0b49d2b2116cee707c330c
61679 F20101118_AACNRU jun_j_Page_027.jpg
4ebc1ba29ae46b39355eb7e034732391
77da0bbc3fac6ce228716881e177e19774e02300
53590 F20101118_AACNSK jun_j_Page_094.pro
b84bec7cc139f6e7b6e24bae7ed68c98
287ad0cb9baa63e68273ea1578ad937c0bb1529d
F20101118_AACNRV jun_j_Page_087.tif
32e7c0b7cb95d2a3df7cc5ec552829c8
9eabf00939f45f426893bb306d733c472ac6e3ef
21067 F20101118_AACNSL jun_j_Page_057.pro
03cf2f17337bb30f82ce9791895b2055
8aadde3892fe8adba1a455249cce42958e0894db
40534 F20101118_AACNRW jun_j_Page_019.pro
09ac61d765420d47ad0c2faa280059dc
358dda332f78a2329c26609c91d2c12c7b140d40
F20101118_AACNSM jun_j_Page_090.tif
ad833692be66e92502e6c643387c89bd
f39d796bca035dcc076b8c3140307ed2f32dcd9e
102561 F20101118_AACNRX jun_j_Page_102.jp2
5a51b70bf091c43d1b256fb243098bba
30bcd356c0c6fe7addf6e85f72f816401e9536b2
60075 F20101118_AACNTA jun_j_Page_121.pro
d07a4089308e60e2985cbec948e443c1
398875db822875b6ba64933539833fac651f174f
43647 F20101118_AACNSN jun_j_Page_101.pro
74d1f733d3ce942ba88875b90ce7ab21
14ff96536f9785e1232d905993c630ee742bd4ab
23035 F20101118_AACNRY jun_j_Page_058.QC.jpg
1d9b5168756f245975a27d978e13e860
3014eb0e6095ca346c33d150c89b0f5da00a5081
99428 F20101118_AACNTB jun_j_Page_005.jpg
43a3a5944f1b604fde0c1b2bd6ade668
d8491d470a107edd48d1d9d69bc79033f3f403b6
F20101118_AACNSO jun_j_Page_028.txt
cdfa6aa26125db7684cfa5bfc3389e53
81b2f623862cd73dc8f1b6369b6691d59741f197
29830 F20101118_AACNRZ jun_j_Page_086.pro
de46acca929e0312563510d3d471c23d
5aa00c2dae2010d8dcce662a36e5671e00f43d85
24575 F20101118_AACNTC jun_j_Page_061.QC.jpg
f22d87504b6d18d541eb4f55cf839330
701c808cf2e22f9aa7dc57f6d0b35b485e5b7eb8
20213 F20101118_AACNSP jun_j_Page_089.pro
2019925970580831a80066d20e4c1bc1
a38378b618e3509b90dcec87449bb57cb838ddfe
102047 F20101118_AACNTD jun_j_Page_105.jp2
f40f358440e7e009cdce940ac5d7c84a
770a0c3e180e72d9c1f0b8942b12881575dabd56
F20101118_AACNSQ jun_j_Page_051.tif
a3ae401291c7b13d62168ae89eeb7a90
8f7a5edbcbee6998b5a8a025fe5b784ac33c659c
F20101118_AACNTE jun_j_Page_019.tif
80533ce42c0b495c4ba87c79ac3b933a
908c6ec4e24a43cda20aae5f501bf632580d0d3a
22497 F20101118_AACNSR jun_j_Page_054.QC.jpg
4d3a74f512e4c790c159de7c766ad0dc
7cce4701382e81ad953bf61e682bbdd2072174cb
871132 F20101118_AACNTF jun_j_Page_041.jp2
0ddab0d15b1c85d9730ecba1c9bef4bd
2ee78e7c9163c30d6134cd37cb800aacf78606ae
F20101118_AACNTG jun_j_Page_049.tif
437dd86c31278a1b92679bdafde65aec
6e6cc268724de14481dda9c5ed2b5fddbac1f42c
1923 F20101118_AACNSS jun_j_Page_117.pro
2967e4b0e66e1a58babd34259a7330ad
d0e462a08fcca23ba0096f9976fdcbd5e789fc64
5653 F20101118_AACNTH jun_j_Page_002.jp2
aabcd9aa105f355403443534343a4952
371a60bc772c40ff5ca379937a0450a527b3678d
F20101118_AACNST jun_j_Page_091.txt
fd51e33738ee705a11e310116c4405a5
646a5789eeedcae251b5df21ab7355a8e4f70ac9
649626 F20101118_AACNTI jun_j_Page_040.jp2
88ee9d198c66ede9dbcfc4c1c040c375
029cd11ff4c6675ed8b967c2937886d97678a74c
34879 F20101118_AACNSU jun_j_Page_100.pro
319ffb12bfc706c3b7a234c3113b2179
9f62c925c4dc53f63be35f7e03b99fc1da51dac7
6630 F20101118_AACNTJ jun_j_Page_032thm.jpg
c904dcf0e141ba337663cfad18b7a920
dd2af1753f62f1e6d9236d09e2ec1f4e2066f246
62517 F20101118_AACNSV jun_j_Page_017.jp2
42a254f189dfe4be355e094537825ff9
f7985f136ca4fa9a6cf6bc5e168096477a17f717
6569 F20101118_AACNTK jun_j_Page_058thm.jpg
9182e5a0da8bcaa1b9576f6d4b710ad4
f38cdc80d3cee7321480c85f30c03c26268e70b9
54369 F20101118_AACNTL jun_j_Page_052.jpg
2b2292bac903d8c41274fdaf86823722
769db85814d1fb1db6c2a0c9c8263a5ae5313091
F20101118_AACNSW jun_j_Page_096.tif
9e5ce1705e8058cecf5756e7bd30cdd4
08784358d7136335e2a3d4d6c0a808fabf7a801b
1473 F20101118_AACNUA jun_j_Page_085.txt
b70d88945a29ff0dba832da375897eec
12ebbc58a5ed9aa7cdb61016c433464ebc957025
62645 F20101118_AACNTM jun_j_Page_060.jpg
0311ff008427128b22316b56cf35b4ab
97e5db8c489a335d21514fba2eebcfb1720a6413
27788 F20101118_AACNSX jun_j_Page_014.jp2
2522226cdd54eb7e594701997a64ba31
596a88f8c90f5e155412659ee3b9bf0d41e4fef1
29221 F20101118_AACNUB jun_j_Page_056.pro
624bede4e6ad22efe1dfff678c8d6650
a40412800f9b11ad244d5a93eea4bfb82e6b48f6
90888 F20101118_AACNTN jun_j_Page_060.jp2
faf0e0c7abc6aec211ffe240e424c14e
bcd7f3a485c5e61878740466bc98c713cce9deba
826827 F20101118_AACNSY jun_j_Page_056.jp2
3ed9261a41e218f496be37e3fd3e1a5f
dd4f4a57e93406640f9c56ab5074f966da3477b3
F20101118_AACNUC jun_j_Page_081thm.jpg
b4fc462ab742ca6c00e228632e14edbe
4e53014703ce1a0f03581a691dc867edc76d399c
35199 F20101118_AACNTO jun_j_Page_080.pro
b5029a7fad336c0b74bb5c392c216360
4f16232ec0655c30b6ed6882b951ba5cfcf6d30d
F20101118_AACNSZ jun_j_Page_105.tif
164b1559f90681363296cad18f27c892
c14fd1b0b1f54125159491569355052c91ac5b0e
F20101118_AACNUD jun_j_Page_010.tif
90f907bee07fac211ec0d31e94fe4103
17ca628db6cd84c670de686a258c7ba26d2a488f
1598 F20101118_AACNTP jun_j_Page_044.txt
e8457e1054396b915637e8d775db9b3e
5a814fae26deb45a30ed9283cde3463d451e1a9a
698955 F20101118_AACNUE jun_j_Page_046.jp2
f9bd88fd275235502098f3c0d1ea27df
7f793131701c5f4cf29bca79d3bacf6df7a20719
6265 F20101118_AACNTQ jun_j_Page_045thm.jpg
03f3b782374ff3afe3cd15f0fa16c252
40c0b361ba1f2be9f33aaaa87d7d2a4c8e443a09
1161 F20101118_AACNUF jun_j_Page_089.txt
f1b19ad8d7cbc6e42fea34779c727d02
6881ff3a5711b1487cb50aca5d4cc3280d55231f
1878 F20101118_AACNTR jun_j_Page_033.txt
46bf6f959ec865ee2eff719dfe62e263
836f0905a4193891b511239eb7c7253013f4a732
35083 F20101118_AACNUG jun_j_Page_062.pro
a72e2a544389587d1cd5e2c1b73515cf
29bece1ff1c4cc99adc091bdcfce04edef248b1f
62557 F20101118_AACNTS jun_j_Page_037.jpg
6c2fb661e2031255c5e6e68e70bf481a
086eef650b5f32f6782f8ededcb9f441561eef76
102285 F20101118_AACOAA jun_j_Page_103.jp2
72e27c6d98a479e85d4483f6375b2ae8
e7e9157be90546f726f535b6f0d8d2dd6d5745fa
49823 F20101118_AACNUH jun_j_Page_118.pro
ff569c8dad3262927b34e6eb14fdb74e
b9ffa1806cbd82897d21392aed468dde317fc3bc
F20101118_AACOAB jun_j_Page_046.tif
bd05bad944ecc67fe7378f3d673343cd
49ed4f39f58e9862d2d1b2c72d87cdfcff9b8ab2
1246 F20101118_AACNUI jun_j_Page_086.txt
e2bdf1ff318d05acb54a754d233007f1
d8b68a12ded46823d111913ad54a7b6475a6c134
20231 F20101118_AACNTT jun_j_Page_039.QC.jpg
b4039615ed9231e8f13e89db2422b782
3355febbc710f0e500b532459774e2aeabd048d0
6256 F20101118_AACOAC jun_j_Page_011thm.jpg
8ba844e4e18d2d365e7df72e58303311
03e34a6775d08121eeeed44343f59c664c7ba011
6173 F20101118_AACNUJ jun_j_Page_065thm.jpg
cf028ffa29298234ee7066aa84d544ef
61e17f8490d6c2c59bbe85091c919bff490a06e6
66553 F20101118_AACNTU jun_j_Page_043.jpg
e9595c308c98e3ec6d2d9d2508a4782b
8243df89ee5cdeed5579ec3a414db733e32d861c
F20101118_AACOAD jun_j_Page_047.tif
fed27c6a10604eca2cc7901ca95015d3
b15345bdd195f6dc5e3a668480daa9b0f2d72008
5697 F20101118_AACNUK jun_j_Page_060thm.jpg
a3cb798fe7cfe4a8ae1039ca90f0d75d
6bb2d9ef1b194d0519dc335339b74020f14e0159
41341 F20101118_AACNTV jun_j_Page_050.jpg
faac2c41a8889326b139da56e36ee9bb
677e69504432e6e7d3229f836e02d95ce9689b47
2193 F20101118_AACOAE jun_j_Page_006thm.jpg
3b064ae43d10c4930bc9777ad305f688
846af7a2237e229be9330782f9edc31baaf93fe4
1863 F20101118_AACNUL jun_j_Page_047.txt
b466e655606e7e71f0d3d15b54117913
8b3dfec2c0e7838e7bd4b7b9429c0528919481d1
62063 F20101118_AACNTW jun_j_Page_030.jpg
4aee660c045c8f52d004151e4887f4af
7cdc82151eee48fe7d5644b783c4c04a568b6304
894274 F20101118_AACOAF jun_j_Page_112.jp2
63626cca5f64e69af4e132a861cc68c0
99a4d288efd880cb8c20c37c070c8d1ec8bd8ae3
35104 F20101118_AACNUM jun_j_Page_013.pro
43117590bcc5337854ac8b66b0b117fa
ec1ec6c9a411c1e07839f0ec48e1e1b36125d96b
16274 F20101118_AACNTX jun_j_Page_107.pro
b7180a261289cbd6e7d9e372dab54c65
77b6301c2d21dcdfa4ab929dc8d343ab19fbc970
21921 F20101118_AACOAG jun_j_Page_048.QC.jpg
24f4882cfda92c149a7a152995dcfd11
6f560d889d7d1cb5529975bc4ff30f7f2d02ddd0
70788 F20101118_AACNVA jun_j_Page_003.jp2
42c10baa54c3c2ecf3e66c6b392e46f4
db0245df736a47f77f70ee50595ed7193ccf3e94
21333 F20101118_AACNUN jun_j_Page_056.QC.jpg
0874c12eb5559f59b71dc4147d63e6c4
702bef300c2245aca7bcbda6e81c321acc0aa44e
50027 F20101118_AACNTY jun_j_Page_098.pro
32a07273b08c3140fe74fea250fe721e
9e9a03c95a8c2a0f3de22e1de2d80af83f6e4967
F20101118_AACOAH jun_j_Page_021.tif
a22f5bb08bb196ae7ebcfb7bb3fe05b6
a66a60be1f677f4c1cf040bc74e78da09cbfd3ec
F20101118_AACNVB jun_j_Page_119.tif
01cb67e49825b2cee4b8d1e2c33c5628
afd5b5287129887500c16c014205b72b31a0c324
89910 F20101118_AACNUO jun_j_Page_114.jp2
2b4149c98acca40ca3205c1ddb18ab84
9e493033a91704dc837df9e95897bd3ea3cb7b89
21166 F20101118_AACNTZ jun_j_Page_041.QC.jpg
4adc2087b4b503728c7421256bc93196
b55561ddb56d4e7b6501d225462b2779b7b7a5d7
1249 F20101118_AACOAI jun_j_Page_002.pro
5fdf87c715b563e853fbf75d3ff60759
5932c455775440ff37f1552fb04ab457c63bb8aa
4850 F20101118_AACNVC jun_j_Page_003thm.jpg
d413e81ca44f63827853ed464fb714e8
807bcaf29d9e6287b1ee8d5297cd423c7ebf74c5
28635 F20101118_AACNUP jun_j_Page_094.QC.jpg
d12e366aee1aaa24926416a575a89474
8f2fe1f4285c88de949af54f75ba90aabaeb639d
140528 F20101118_AACOAJ UFE0014378_00001.mets
14b2595a96562184471a70e78ea0d145
c957da2f4bb9c8bb805582416e4b262c8a18f2a0
F20101118_AACNVD jun_j_Page_001.txt
a53e2d6c535ecab5046c08110942324c
6215e97eb8855d9eb81f41ddd8a9ab3b8e7086e4
68664 F20101118_AACNUQ jun_j_Page_081.jpg
cabfc46a7638d95f2a5d67fe227ca82b
d717137719d5d8645fe9dd17a9f724111f173256
4704 F20101118_AACNVE jun_j_Page_004thm.jpg
4cf651d11e12bd35a73a91f7bb9dcc66
d272cdcfb4e916c9fa9551d2e53f7da431137959



PAGE 1

DESIGN AND TESTING OF A SELF-POWERED WIRELESS HYDROGEN SENSING PLATFORM By JERRY CHUN-PAI JUN A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2006

PAGE 2

Copyright 2006 by Jerry Chun-Pai Jun

PAGE 3

ACKNOWLEDGMENTS This thesis could not have been complete d without the help of many people. First and foremost, I would like to thank my me ntor, Dr. Jenshan Lin, for giving me the opportunity to work on this res earch project, and for providi ng me with guidance, and a RA position for the past two years. I would also like to thank Dr. Khai Ngo, and Dr. Toshik azu Nishida, for taking the time to meet with the group every week to disc uss, brainstorm, and evaluate the status of this project. These weekly discussions were very helpful in keeping the project on track, and for coming up with new ideas. I would like to thank all those who have worked with me on this project. This includes Bruce Chou, Alex Phipps, Anurag Kasyap, Shengwen Xu, Hung-Ta Wang, LiChia Tien, and last, but not least, David Johns on. Additionally, I would like to thank my colleagues, Tien-yu Chang, Lance Covert Ashok Verma, Xiuge Yang, Hyeopgoo Yeo, Yanming Xiao, Changzhi Li, Sang Won Ko, Jae Shin Kim, and Jaeseok Kim, for putting up with me, supporting my research, and crea ting a comfortable a nd productive working environment. Finally, I would like to thank my parent s for their endless love and support.

PAGE 4

iv TABLE OF CONTENTS page ACKNOWLEDGMENTS...................................................................................................3 LIST OF TABLES............................................................................................................vii LIST OF FIGURES...........................................................................................................ix ABSTRACT.....................................................................................................................xiii CHAPTER 1 INTRODUCTION........................................................................................................1 Motivation.....................................................................................................................1 Contributions................................................................................................................2 Thesis Organization......................................................................................................3 2 WIRELESS SENSOR SYSTEM LIMITATIONS.......................................................4 Energy Harvesting and Reclamation............................................................................4 Solar Energy Harvesting........................................................................................5 Vibrational Energy Harvesting..............................................................................6 Design Limitations of Low-Power and Low-Voltage Discrete Components...............8 Dynamic Range.....................................................................................................9 Input Offset Voltage............................................................................................10 Least Significant Bit (LSB).................................................................................11 Limitations of a Wireless System...............................................................................12 Wireless Channel Estimation Techniques...........................................................12 Free space path loss...................................................................................13 Two-ray ground reflection model................................................................14 Shadowing model.........................................................................................15 FCC Part 15 Regulations.....................................................................................17 The Federal Communi cations Commission.................................................18 FCC rules, regulations, and safety...............................................................19 3 SENSOR INTERFACE DEVELOPMENT...............................................................22 ZnO Nano-Rods..........................................................................................................23 ZnO Nano-Rod Fabrication Process....................................................................24

PAGE 5

v Performance of ZnO Nano-Rods.........................................................................25 Uncoated ZnO nano-rods.............................................................................27 Pd coated ZnO nano-rods.............................................................................28 Pt coated ZnO nano-rods..............................................................................29 Detection Interface for Hydrogen Sensitive Devices.................................................30 Wheatstone Resistive Bridge...............................................................................31 Differential Detection Interface...........................................................................34 Difference amplifier.....................................................................................34 Instrumentation amplifier.............................................................................37 Realization and Testing of Diff erential Detection Circuit..........................................38 Selection of operational amplifier................................................................39 Simulation of differentia l detection circuit..................................................39 Fabrication of differen tial detection circuit..................................................43 4 MICROCONTROLLER DEVELOPMENT..............................................................46 Microcontrolle r Selection...........................................................................................46 Modes of Operation....................................................................................................47 Power Requirements of Microcontroller....................................................................48 5 LOW-POWER WIRELESS COMMUNICATION LINK.........................................50 Selection of a Modulation Technique.........................................................................50 Selection of Operating Frequency..............................................................................53 Selection and Performance of a RF Transmitter.........................................................55 Rayming Corporation TX-99 300 MH z AM Transmitter/RE-99 Receiver Pair...................................................................................................................55 Ming Tx-99 transmitter................................................................................56 Ming RE-99 receiver....................................................................................59 Ming distance measurements.......................................................................60 Linx Technologies LR series Transmitter and Receiver.....................................63 Linx Technologies TXM-315-LR................................................................64 Linx Technologies RXM-315-LR................................................................66 Wireless Link Optimization........................................................................................68 Ming TX-99 Power Analysis...............................................................................68 Low Profile Antenna...........................................................................................71 RF Transmitter Optimization..............................................................................77 FCC Part 15.231..................................................................................................80 Central Monitoring Station..................................................................................84 6 MINIMUM REDUNDANCY MINIMUM ENERGY CODING..............................87 Minimum Energy Coding...........................................................................................87 Minimum Redundancy Minimum Energy Coding.....................................................89 7 FULL SYSTEM INTEGRATION TESTING............................................................94 Hydrogen Chamber Equipment Setup........................................................................94

PAGE 6

vi Full System Integration Test.......................................................................................96 8 CONCLUSION AND FUTURE WORK.................................................................100 Conclusion................................................................................................................100 Future Work..............................................................................................................101 LIST OF REFERENCES.................................................................................................104 BIOGRAPHICAL SKETCH...........................................................................................109

PAGE 7

vii LIST OF TABLES Table page 2-1. Typical Values for Path Loss Exponent .................................................................16 2-2. Typical Values of Shadowing Deviation dB.............................................................16 3-1. Various Commercial Operational Amplifiers.............................................................39 3-2. Differential Detection Circuit Component Values.....................................................41 3-3. Initial Measurements of Di fferential Detection Interface...........................................44 4-1. Features of Texas Instruments MSP430F1232IPW..................................................47 5-1. Comparison of Available Commercial ASK/OOK Transmitters...............................55 5-2. Performance of Ming TX-99......................................................................................59 5-3. Maximum Transmission Distances with Varying Antenna Locations.......................62 5-4. Performance of LINX TXM-315-LR.........................................................................65 5-5. Antenna Gain Measurements.....................................................................................73 5-6. Gain Measurements for Matched Antenna.................................................................76 5-7. Component Values of Ming TX-99 Transmitter........................................................78 5-8. Various High-Frequenc y NPN BJT Transistors.........................................................79 5-9. Performance of Various Transistors and Resistors for Ming TX-99 Transmitter......80 5-10. Limitations under FCC Part 15.231 (a -d) **Linear Interpolations........................82 5-11. Limitations under FCC Part 15.231 (e ) **Linear Interpolations............................82 6-1. Minimum Energy Coding Scheme Proposed by Erin and Asada...............................88 6-2. Proposed Source Coding Technique with Comparison to Technique of Erin and Asada........................................................................................................................91

PAGE 8

viii 8-1. Comparison of Performance Between De signed Differential Detection Circuit, and Other Commercially Available Designs..........................................................100

PAGE 9

ix LIST OF FIGURES Figure page 1-1. System Level Block Diagram of Se lf-Powered Wireless Hydrogen Sensor................2 2-1. System Level Implementati on of Energy Harvesting Devices.....................................5 2-2. IXOLAR XOD17-04B Solar Cell................................................................................6 2-3. Pulse Resonant Power Converter.................................................................................6 2-4. Vibration Energy Devices............................................................................................7 2-5. Performance of Vibration Energy Harvesting Devices................................................8 2-6. Dynamic Range of u741 Operational Amplifier..........................................................9 2-7. The Shrinking LSB LSB for Multiple Voltage Supply Spans.................................11 2-8. FCC Organizational Chart..........................................................................................18 2-9. Proposed IEEE RF Safety Guidelines........................................................................21 3-1. ZnO Nano-Rods..........................................................................................................25 3-2. Close-Up of Packaged ZnO Nano-Rod Sensor..........................................................25 3-3. Simple Schematic of Hydrogen Cham ber Used for ZnO Nano-Rod Testing............26 3-4. Schematic of Biasing for ZnO Nano-Rod Hydrogen Sensitivity Testing..................26 3-5. Uncoated ZnO Nano-Rod Relative Re sistance Change for Various Hydrogen Concentrations..........................................................................................................27 3-6. Pd-coated ZnO Nano-Rod Relative Re sistance Change for Various Hydrogen Concentrations..........................................................................................................28 3-7. Pd-coated ZnO Nano-Rod. Absolute Re sistance Change fo r Various Hydrogen Concentrations..........................................................................................................29 3-8. Pt-coated ZnO Nano-Rod Relative Resistance Change for Various Hydrogen Concentrations..........................................................................................................30

PAGE 10

x 3-9. Wheatstone Resistive Bridge......................................................................................32 3-10. Difference Amplifier................................................................................................35 3-11. Difference Amplifier with Non-Inve rting Buffer to Differential Inputs..................36 3-12. Instrumentation Amplifier........................................................................................36 3-13. Full Schematic for Differential Detection Circuit....................................................38 3-14 New Pt-coated ZnO Nano-Rod Grown a nd Packaged for Differential Detection Circuit.......................................................................................................................40 3-15. Agilent ADS 2003 Simulation Setup fo r Differential Detection Circuit..................41 3-16. Agilent ADS 2003 Simulation: Out put Voltage to Swept ZnO Nano-Rod Resistance.................................................................................................................42 3-17. Protel PCB Top and Bottom Layout.........................................................................43 3-18. Fabricated and Assembled Differential Detection Interface Board with Packaged ZnO Nano-Rod Sensor.............................................................................................43 3-19. Measured Output Volta ge vs. ZnO Nano-Rod Resist ance Sweep for Fabricated Differential Detection Circuit...................................................................................45 4-1. Microprocessor State Flow Diagrams........................................................................48 4-2. Initialization Power Required for MSP430F1232IPW...............................................49 5-1. Tradeoffs Between Performance and Architecture Complexity of /4 DQPSK and OOK.........................................................................................................................51 5-2. Signal Constellations..................................................................................................52 5-3. Path Loss Attenuation(dB) with Respect to Carrier Frequency.................................53 5-4. Schematic of Ming TX-99 Taken from Datasheet.....................................................56 5-5. Ming TX-99 Transmitter in OOK Mode. VDD Is Replaced with Data Stream.........57 5-6. Ming TX-99 Transmitter. Red Outline Highlights Onboard Antenna.......................57 5-7. Test Setup for Output Power and Power Consumption of Transmitters....................58 5-8. SMA Connector Soldered to Antenna Tap on Ming TX-99 Transmitter...................58 5-8. Schematic of Ming RE-99 Taken from Datasheet......................................................59

PAGE 11

xi 5-9. Ming RE-99 Receiver.................................................................................................60 5-11. Floorplan of First Floor Atri um of New Engineering Building...............................61 5-12. Experimental Setup for Distance Measurements......................................................61 5-13. Received Power vs. Distance With Reference to Room Shape................................62 5-14. Screen Capture of Received Power Spectrum..........................................................63 5-15. System Level Archite cture of LINX TXM-315-LR.................................................64 5-16. Pin-Out of TXM-315-LR Transmitter......................................................................65 5-17. System Level Architecture for RXM-315-LR..........................................................66 5-18. Pin-Out of RXM-315-LR Transmitter......................................................................67 5-19. Tektronix TDS5104B Digital Phosphor Oscilloscope Screen Capture of Power Analysis Performed for RF Transmission of One Bit..............................................69 5-20. Tektronix TDS5104B Digital Phosphor Oscilloscope Screen Capture of Power Analysis Performed for RF Transmission of Multiple Bits.....................................70 5-21. LINX ANT-315-SP SPLATCH Antenna From Datasheet.......................................71 5-22. Testing Board for SPLATCH Antenna...................................................................72 5-23. S-Parameter for SPLATCH Antenna.......................................................................74 5-24. S-Parameters of Measured (red) and Simulated (blue) in Ansoft Designer.............75 5-25. Matched Antenna (Red) vs Unmatched Antenna (blue).........................................75 5-26. 5 MHz Bandwidth of Matched Antenna...................................................................76 5-27. Microstrip Inductance Measurem ent for Ming TX-99 Onboard Antenna...............77 5-28 Agilent ADS 2003 Simu lation Environment.............................................................79 5-29. Flowchart for FCC Part 15-231 Requirements.........................................................81 5-30. Graphical Representation of Field St rength Limitations for Part 15.231 Section e.............................................................................................................................. ..82 5-31. Moving Average Filter Example..............................................................................85 5-32. Labview Code...........................................................................................................86

PAGE 12

xii 6-1. Power Consumption Reduction per A dditional Redundant Bit Comparison for Multiple Source Coding Schemes............................................................................92 7-1. Redrawn Schematic with Higher De tail of Hydrogen Chamber Components...........95 7-2. Microcontroller with RF Transmitter Attached to Microcontroller USART PORT..96 7-3. Output Data Bit-Stream Pattern of Received Data From Ming RE-99 During System Integration Testing.......................................................................................97 7-4. Solar Cells (left) with Solar Power IC (right).............................................................98 7-5. Vibration Energy Harvesting Components.................................................................98 7-6. System Integration Testing Block Di agram with Screen Caps From Video..............99 8-1. Protel PCB Top and Bottom Lay out For Fully Integrated Board............................102 8-2. Fully Assembled Single Module..............................................................................103

PAGE 13

xiii Abstract of Thesis Presen ted to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science DESIGN AND TESTING OF A SELF-POWERED WIRELESS HYDROGEN SENSING PLATFORM By Jerry Chun-Pai Jun May 2006 Chair: Jenshan Lin Major Department: Electrical and Computer Engineering Within the ongoing interdisciplinary hydr ogen research at the University of Florida, a selfpowered wireless hydrogen se nsor node has been designed and developed. By using multi-source energy harvesting ci rcuitry designed and developed at the University of Florida, scavenged or reclaimed energy from light emitting and vibrational sources serve as the source of power for commercial low power microcontrollers, amplifiers, and RF transm itters. After system power up, the sensor node is capable of conditioning and deci phering the output of hydrogen sensitive ZnO nano-rods sensors also developed at the Univ ersity of Florida. Upon the detection of a discernible amount of hydrogen, the system w ill wake from an idle state to create a wireless data communication link to relay the detection of hy drogen to a central monitoring station. The systems sensitivity is on the order of parts per million, and hydrogen concentrations starting as low as 10 PPM can be detected.

PAGE 14

xiv The thesis will discuss the performan ce of the self-powered wireless hydrogen sensor node, and also focus on the design a nd optimization of the detection circuitry, digital processing considerations, modulati on scheme and wireless communications link to maintain an accurate and reliable syst em, while expending a minimal amount of energy scavenged from ambient light or vibra tion for very long lifetime operation.

PAGE 15

1 CHAPTER 1 INTRODUCTION Motivation Self-powered wireless sensors themse lves are becoming a popular topic for implementation in future systems. The idea of inexpensive sensor devices with very long-lifetime operation involving minimal maintena nce, is poised to make a very strong impact on the engineering community. Because these devices can be deployed in very harsh and dangerous environs, environments can be monitored remotely in real-time to update the end-user of ambient conditions, and report any slight deviations from safe and normal operating conditions. This can help prev ent endangerment to the quality of life for surrounding bodies, without placing expensive mach inery and humans in harms way. As fossil fuel supplies are depleted, alte rnative fuel supplies such as hydrogen which can be quickly replenished, are swif tly growing in popularit y, and with a selfpowered wireless hydrogen sensing platform capab le of sensitivity on orders of parts per million (PPM), the applications of these sensors can include monitoring of hydrogen powered machines, combustion gas detection in spacecrafts, and solid oxide fuel cells with proton-exchange membranes [1,2]. With a self-powered wireless hydrogen sens or, and the current advances in systemon-chip technology and MEMS processes, an integrated circuit capable of harvesting energy supplies through ambient conditions su ch as lighting and vibration, can be realized in a physically small package to c ontrol and report the da ngers involved in the development of hydrogen as a viable fuel source for the future. The overall block

PAGE 16

2 diagram of the self-powered wireless hydrogen sensing platform can be found in Figure 1-1. Figure 1-1. System Level Block Diagram of Self-Powered Wireless Hydrogen Sensor Contributions The objective of this research is to develop a self-powered wireless hydrogen sensor node powered through energy harv esting techniques and capable of PPM hydrogen sensitivity to report and transm it data, via a low-power wireless communications link, to a central monitoring st ation. The specific goals of the research provided in this dissertation, are to: Present limitations involved in the design of a low-power wireless sensing platform given current restrictions placed by government requirements, and those restrictions involved in using commer cial off-the shelf analog components. Develop a low-power sensor interface to convert the react ion of the hydrogen sensitive mechanism into a conditioned si gnal which can be accurately represented in digital form, while consuming minimal scavenged energy. Develop, test, and optimize a wireless communications link for communication between the self-powered wireless sens or node and a central monitoring station, while meeting the regulations im posed by governmental agencies.

PAGE 17

3 Develop a source coding scheme to minimize the required power to transmit a single message. Perform a full system integration and te st system with a live hydrogen source Thesis Organization In chapter 2, the limitations involved in the design of a se lf-powered wireless hydrogen sensor are presented. Chapter 3 in troduces the hydrogen sensing mechanism, and the design and testing of a differential detection circuit to interface the hydrogen sensing mechanism to a digital back-end. Chap ter 4 iterates the microcontroller system, and the operation of the contro ller system with regard to the sensor platform, while Chapter 5 emphasizes the design, testi ng, and optimization of the wireless communications link. A minimu m energy source coding scheme is discussed in chapter 6, and a full system implementation between energy harvesting devices, RF front-end, sensing mechanism and sensor interface, and controller system are described in Chapter 7. Finally, conclusions, current work, and future work are detailed in Chapter 8.

PAGE 18

4 CHAPTER 2 WIRELESS SENSOR SYSTEM LIMITATIONS There cannot be the successful design of a fully wireless sensor platform without the designation of limitations to the system Once a thorough analysis of those limitations which would hinder or stop the progress of the design is completed, design orientated specifications that the system must meet can be set to assist in the realization of the system. To be a truly self-powered wireless hydr ogen sensor node, a study of the potential short-comings of the energy harvesting devi ces serving as the power sources for the system must be completed. By using th e available power provided from energy harvesting techniques, the design of the sensor system must strive for both accuracy and ultra-low power operatio n. With the requirement for a low-power sensor design, also includes those problems associated with available commercial low-power and low voltage analog and digital components. A t horough investigation in to the problems and solutions to the issues expected to arise in the design of the lowpower analog and digital blocks of the system will be required. In addition to a power analysis, an analysis of the limitations that will make a direct impact on the design of a wireless transmitter and receiver will also be discussed. Energy Harvesting and Reclamation This section reviews the solar and vibration harvesting devices and circuits designed, fabricated, and tested at the University of Flor ida. The objective of these designs was for an end product capable of extracting energy from both photovoltaic, and

PAGE 19

5 piezo-electric(PZT) energy harvesting devices This energy would then be primed by efficient power converters for use by th e hydrogen sensor, sensor interface, microcontroller, and transmitter. Interested pa rties should refer to the original references [3,4,5,6,7]. For the purposes of this disserta tion, a system level implementation of the energy harvesting devices can be found in Figure 2-1. Figure 2-1. System Level Implementa tion of Energy Harvesting Devices Solar Energy Harvesting Photovoltaic devices are a mature comme rcial product, and offer the attractive availability of high energy density per area. They are however, limited to real-time lighting and temperature conditions. The IXOLAR XOD17-04B Solar Cell seen in Figure 2-2 was used as the solar harvesting device, and the energy yielded from this device was conditioned by a Pulse-Resonant Po wer Converter designed at the University of Florida for use by the electronics of the wireless sensor system. The Pulse Resonant Power Converter, whos e functional block diagram can be seen in Figure 2-3(a) was designed to be self-powered and self -controlled for maximum power point tracking, low switching lo ss, and to convert an input voltage of 0.8 1.2V to a

PAGE 20

6 steady 2V output voltage to be used by the sensor, sensor interface, microcontroller, and RF transmitter. The bare die photo of this power IC can be seen in Figure 2-3 (b). Figure 2-2. IXOLAR XOD17-04B Solar Cell 1.5mm Converter Controller Charge Pump a) b) 1.5mm Converter Controller Charge Pump 1.5mm Converter Controller Charge Pump a) b) Figure 2-3. Pulse Resonant Power Converter. (a) Functional block di agram (b) Bare die photo of Pulse Resonant Power Converter IC Vibrational Energy Harvesting For the harvesting of vibrat ional energy, piezo-electric devices are attractive as sources in that no light is required, and the collection of energy is proportional to the volume of the devices. The limiting factor however, is the magnitude and frequency of

PAGE 21

7 the vibrations. As MEMS PZT devices are currently being devel oped, for a proof of concept design, commercial PZT bimorph beam s were purchased and used. Four PSI D220-A4-203YB Double Quick Mounted Y-Pole Bender seen in Figure 2-4 (a), were selected as the PZT devices, and were connected to a direct charging circuit (full-bridge rectifier and shunt capacitance) seen in Figure 2-4 (b) constructed at UF. Figure 2-4. Vibration Energy Devices. (a) Four mounted PSI D220-A4-203YB Double Quick Mounted Y-Pole Bender (b) Direct Charging Circuit The vibration energy harvesting system wa s tested under lab conditions and used a mechanical shaker tuned to 1 grms @ 130 Hz (the resonant frequency of the bimorph beam), as the source of vibrations. This pr oof of concept design was able to deliver 250

PAGE 22

8 uW of power. The efficiency and power tran sfer to battery versus the mechanical input power of the PZT bi-morph beams can be seen in Figure 2-5. Power to Battery vs. Mechanical Input Power 0 50 100 150 200 250 300 0.00100.00200.00300.00400.00500.00600.00 Mechanical Input Power [uW]Power to Battery [uW] Direct Charging Efficiency vs. Mechanical Input Power0.00 20.00 40.00 60.00 80.00 0.00200.00400.00600.00 Mechanical Input Power [uW]Efficiency [% ] a)b) Power to Battery vs. Mechanical Input Power 0 50 100 150 200 250 300 0.00100.00200.00300.00400.00500.00600.00 Mechanical Input Power [uW]Power to Battery [uW] Direct Charging Efficiency vs. Mechanical Input Power0.00 20.00 40.00 60.00 80.00 0.00200.00400.00600.00 Mechanical Input Power [uW]Efficiency [% ] a)b) Figure 2-5. Performance of Vibration Energy Ha rvesting Devices (a)Power to Battery vs. Mechanical Input Power (b) Direct Char ging Efficiency vs. Mechanical Input Power Design Limitations of Low-Power and Low-Voltage Discrete Components System designers working with discrete co mmercial integrated circuits (IC) are moving towards single supply, low power, and low voltage designs. Such designs are becoming more popular within design comm unities due to thei r reduction of cost, component count, and power consumption. An innate feature of using low-voltage single-supplies is the reduction of quiescent current, which is necessary for batterypowered systems, or in the case of se lf-powered operation, scavenged energy. Another not-so-obvious reasoni ng for using a single-supply system is the increase of the reliability of the system. Due to the desire for a long-lifetime sensor node requiring minimal maintenance, by designi ng a system with discrete components operating at levels much lower than their ma ximum ratings, this inhe rently increases the lifetime of the device. The trade-offs of designing a system with low power and low voltage operation however, come s the adverse affects associ ated with the slew rate,

PAGE 23

9 bandwidth, and head-room problem s [8]. These trade-offs must be considered in the design of a sensor interface w ith the necessity to obtain an accurate real-world signal. This following section provides insight into the careful design of a sensor interface. Dynamic Range As previously mentioned, Headroom is the associated dynamic range of a lowvoltage system. Since the systems usable resolution is dependent upon the signal to noise ratio, the dynamic range of a system is perhaps one of the most significant tradeoffs in low-voltage single-s upply design[8]. If one considered conventional u741 +/15 V dual supply operational amplifiers, due to the architecture of this device, the input/output has a fixed head room of 2 V. This 2V denotes that the maximum input/output swing of the u741 is between +/13 V--this swing is th e dynamic range of the operational amplifier and can be seen in Figure 2-6 from [8], Figure 2-6. Dynamic Range of u741 Operational Amplifier If one were to then translate this fi xed headroom to a si ngle-supply operational amplifier operating from 0 to 3 V, the input/output maximum swing does not exist because the headroom required is 4 V, and the maximum voltage span provided is only 3 V. The dynamic range of this device is ab solutely unacceptable and will prove to be

PAGE 24

10 disastrous in the final design. Thus, the si ngle-supply design comm unity is constantly increasing the demand for rail-to-rail amplifie rs with very low headroom requirements to increase the dynamic range of the system. [8] Input Offset Voltage Due to the unbalances of the transist ors and resistors within an operational amplifier, it is often the case that an input offset voltage can be found between the input terminals of an operational amplifier. Unfo rtunately, input offset voltages do not scale down with supply voltages, and by reducing the power supply voltage, there may be a disproportional shift in input offset voltages [9]. At h eart, an operational amplifier is simply a differential amplifier which w ill amplify the difference between two input terminals. If one were to consider a syst em with high large-signal gain, along with the signal, the output would also include the amplified error of the input offset voltage. Conventional operational amplifiers such as the u741 usually include offset-null trimming points. Although these points will co mpensate for input offset voltage in a per case basis, the trimming potentiometers, as well as the offset voltage drift of the operational amplifier, are plagued with thei r dependence to temperature. Although most operational amplifiers are factory trimmed, a system designer should also place a heavy emphasis on searching for operational amplifiers with low input offset voltage, and low offset drifts. Often more times than not, the system designer will place more importance on the selection of a device based on errors caused by drift due to temperature or time than due to absolute magnitudes of an error. This is due to microprocessor computing horsepower becoming fairly inexpensive and a llowing system designers to easily linearly correlate data using software such as a cr ude lookup-table approach to compensate for absolute magnitudes of error. Such softwa re techniques however, cannot compensate for

PAGE 25

11 effects due to time and temperature, which ma kes the design of a good sensor system that much more important for a device with a de sired long lifetime of operation with minimal maintenance. Least Significant Bit (LSB) With a wavering voltage span, is the relati ve precision required to acquire a signal. As compared to a +/15 V system, given an ADC of 12-bits, a 3 V span would require up to eight times of precision as compared to that of a +/15 V system. A comparison seen in Figure 2-7 from [9] graphically displays th e increasing precision of the LSB to varying system voltage supplies. Relative Size of a 12-bit LSBSystem Supplies and Voltage Span+/-10 V System +/-5 V System +/-3 V System Single 3 V System Single 5 V System +4.4V +3.5V +2.2V +2.5V -3.5V 0V -2.2V 0V -10V +10V One LSB 4.8mV 1.7mV 1.2mV1.1mV 0.6mV Relative Size of a 12-bit LSBSystem Supplies and Voltage Span+/-10 V System +/-5 V System +/-3 V System Single 3 V System Single 5 V System +4.4V +3.5V +2.2V +2.5V -3.5V 0V -2.2V 0V -10V +10V One LSB 4.8mV 1.7mV 1.2mV1.1mV 0.6mV Figure 2-7. The Shrinking LSB LSB for Multiple Voltage Supply Spans

PAGE 26

12 Limitations of a Wireless System Once a system is successful in overcoming the requirement for low power, and has the ability to accurately obtain the real world signal of interest, if a wireless sensor node is unable to communicate with a central monitoring station, all efforts made in the design of power harvesting and of si gnal processing are wasted. Alike to the design of an accurate analog to digital interf ace, a RF wireless system suffers just as many if not more limitations. In addition to the physical limitatio ns of a reliable wireless front end, include those limitations imposed by the United St ates Federal Communications Commissions (FCC) Code of Federal Regulations (CFR) to further aggravate a system designer. Such limitations will be further discussed in this section. Wireless Channel Estimation Techniques Due to the effects of multi-path, and si gnal power-loss, the accurate modeling of the wireless channel within which a transm itted message will propagate through becomes a fairly difficult and complex parameter to model. An understanding of the wireless channel is essential in the build of a link budge t, to estimate the data range of a wireless system. A link budget can be expressed as [10]; dBdBdBdBdBathossTRTR P LPPGG Where ( PT) is the power of the transmitter, ( GT) and ( GR) are the gains of the transmitter and receiver antennas respectively, ( PR) is the received power, or receiver sensitivity in the case of a link budget, a nd Path loss is the attenuation due to the propagation of a RF signal. Once we can veri fy the output power, gains of antennas, and receiver sensitivity, we can use an appropriate path loss model to extract a theoretical estimate of our distance. Unfortunately, an accurate path loss model is difficult to

PAGE 27

13 realize, however, there are certain technique s one can use to estimate and calculate a theoretical data range for our wireless system. Free space path loss Free-space path loss is the attenuation of the electromagnetic waves of a radio signal traveling from a transmitter to receiver [11] Free-space loss is usually the first unknown parameter to be estimated, and alike to its moniker, free-space path loss is the attenuation based strictly on the ideal propa gation condition for th e spherical expansion of the RF wave front [12] through free-spac e where there is only one clear line-of-sight path between the transmitter a nd receiver, and does not includ e those losses incorporated with reflections from objects, diffraction, refr action, absorption and any other variables. This of course results in a path loss with si gnificantly less attenuation that can be found in real-world practice. [12] The principle behind the free-space path lo ss is-signal attenuation is proportional to both the square of the dist ance as well as the square of the frequency while taking into account the expanding spherical wave-front of the electromagnetic signal. As a electromagnetic signal travels from a transmitter to a receiver, the signal spreads in all three-dimensions creating an expanding spherica l surface. A sphere with radius R has a surface area of: 24()urfacereaphereSASR If R, the radius is doubled, then the surf ace area is increased by a factor of four. This relationship is known as the Inverse Square Law. Given these details, we can derive that the free-space path loss of a wireless channel is:

PAGE 28

14 24reepaceathoss R FSPL Where cc f c is the speed of light (3 x 108), fc is the fundamental frequency of interest, and R is the distance be tween the receiver and transmitter. Using this path loss estimate, H. T. Friis presented his channel model; the Friis Free Space Path Loss Model which is used to calculate the re ceiver signal power ( PR ) with respect to antenna gain of the transmitter ( GT ) and receiver ( GR ) antennas, attenuation due to path loss, and transmitted power ( PT). The Friis Free Space Equation is [13]: 2 2() (4)TTR RPGG PR R The Free-Scale Path loss model makes the assumptions of the most ideal of conditions for the transmission of data. Unfortunately, in the real world, these assumptions are rarely, if ever true. Thus more sophisticated modeling is required. Two-ray ground reflection model A more common approach for the estimati on of propagation signal attenuation is the plane earth propagation mode l which models the average attenuation associated with the distance between a stationary transmitter a nd receiver with direct line-of-sight, while taking into account the ground re flection path. This method is considered to be more accurate of a model and can be used to r oughly estimate the attenuation for fundamental

PAGE 29

15 frequencies within the ultrahigh-frequency bands between 200 MHz and 5 GHz. [10] The path loss for the two-ray ground reflect ion model can be seen in the following equation, where ( R ) is the distance between the transmitter and receiver, and ( ht) and ( hr) are the heights of the transmitter and receiver from the ground, respectively. 4 22 woayroundoss trR TRGL hh From this path loss equation, it is assumed that R is much greater than the heights of either antenna. From this path loss equation, we determine the power received ( PR) given the transmitter power ( PT), the heights ( ht, hr) and gains ( GT, GR) of the transmitter and receivers respectively. 22 4()TTRtr RPGGhh PR R Up to now, both models of propagation have been using the dependence of distance of the propagation path to es timate the received power for the wireless system. This technique of estimation represen ts the communication range as that of a sphere. We have yet to consider the effects due to random multi-paths, which is a more credible model for channel estimation. Shadowing model Both free-space and two-ray ground models neglect the reality that the power received at a given receiver is more of a random variable due to multiple path effects, or fading effects. [13] The use of a shadowi ng model extends the ideal sphere of the previous methods, where the predicted received power is more of a mean of received

PAGE 30

16 signals, and creates a richer more statistica l model based on the environment or terrain of the propagation path. The shadowing model c onsists mainly of two parts. One part, like previous modeling techniques, calculates a mean path loss for the system within the path of propagation, while the second part refl ects variations of power at certain distances [13]. Table 2-1 and Table2-2 show typical values for channel estimation using the shadowing model. Table 2-1. Typical Values for Path Loss Exponent Environment Free Space 2 Outdoor Shadowed Urban Area 2.7 to 5 Line-of-Sight 1.6 to 1.8 In Building Obstructed 4 to 6 Table 2-2. Typical Values of Shadowing Deviation dB Environment dB (dB) Outdoor 4 to 12 Office, hard partition 7 Office, soft partition 9.6 Factory, line-of-sight 3 to 6 Factor, obstructed 6.8 In the path loss portion of the shadow ing model, the mean received power (PR) at distance (R) is found by referenc ing (R) to the power received at a distance closer in, (R0). The mean received power for a given di stance R is computed through the following equation where is a typical path loss exponent found in Table 2-1. 0 0() ()R RPRR PRR This can also be expressed in terms of dB.

PAGE 31

17 00() 10log ()dBR R P RR P RR For the variations of received power at certain distances, it can be modeled as a log-normal random variable, or a Gaussian random variable XdB with a mean of zero, and a standard deviation of dB. Thus, the shadowing model can be fully represented by: 00() 10log ()dBR dB RPRR X PRR The shadowing model gives a more authen tic estimation of a wireless channel. However, a common radio practice for interior wireless channel estimation is to use any path loss models presented in previous secti ons, and to assume an addition 15 to 20 dB of fade margin in the link budget calculation. This margin should account for multi-path phenomena, shadows, reflections, system losse s, and other divergences from an ideal system model. [10] FCC Part 15 Regulations Within the United States of America, th ere are strict impingements placed on the use of the radio spectrum between 9 kHz up to 3 THz. The responsibility of regulating the radio spectrum falls in the hands of th e FCC who administrates the spectrum band for non-federal governmental use, and the Nati onal Telecommunications and Information Administration (NTIA), a unit under the Depa rtment of Commerce, who regulates the spectrum for use by the federal government. Since this design, is for non-Federal Government use, only the details of restricti ons placed from the FCC will be assessed in detail. Interested parties should visit www.fcc.gov for more information.

PAGE 32

18 The Federal Communications Commission The FCC, established by the Communicat ions Act of 1934, is an independent government agency whose duties and responsibi lities are directly d ecided by Congress. The FCC consists of five commissioners appoi nted by the president and confirmed by the senate, and are slated for five year term s except when filling an unexpired term. One commissioner is appointed chairperson by th e president, and only a maximum of three commissioners may be allowed to be of the same political party [14]. Figure 2-8. FCC Organizational Chart The FCC is maintained by a staff organized by function, and has jurisdiction over the entire 50 states, the District of Columbia as well as all U.S. te rritorial possessions. Broken up into six operating Bureaus, and ten Staff Offices, the Bureaus are responsible for processing license applic ations, analyzing complaints conducting investigations, developing and implementing regulatory programs and taking part in hearings while the

PAGE 33

19 Offices provide support services. An organi zational chart of the different operating offices and bureaus of the FCC can be seen in Figure 2-8. FCC rules, regulations, and safety The rules and regulations of devices capable of emitting radio waves within the radio spectrum are located in title 47 of the code of fede ral regulations (CFR). Although the Office of Engineering and Technology (OET) maintains and is responsible for parts 2, 5, 15, and 18 of title 47, the official rules ar e published and maintained in the Federal Register. [15] The OET is an office within the FCC and is responsib le for allocation of the radio spectrum for public, non government use, and provides advice on technical and policy issues governed by title 47 of the CF R. The OET is also responsible for the maintenance of the Table of Frequency Allocations. The Rules of title 47 of the CFR are divided into part 0 through 101, and organized into four sub chapters. Rules applicable to the design of a sensor are outlined as follows: Part 2: Frequency Allocations and radi o Treaty Matters; general rules and regulations Part 5: Experimental Radio Se rvice(other than broadcast) Part 15: Radio Frequency Devices Part 18: Industrial, Scien tific, and Medical equipment The regulations set by the FCC to police the spectrum and mitigate unfair interference between intentional and unintenti onal radiating devices (such as computer monitors), ultimately limits the radiated power from radio systems. Thus, these regulations directly affect both the RF transm itter and any gains associated with antennas. Given the premise mentioned earlier where di stance is proportional to the output power of the transmitter/antenna pair, the FCC creat es further difficulties in the design of a

PAGE 34

20 wireless front end. In addition to those ru les and restrictions set forth by the FCC, additional insight into the hum an-electrical interaction serves as the basis for another regulatory standard. As required by the National Environmenta l Policy Act of 1969, the effect on the quality of the human environment from the emissions of transmitters regulated by the FCC needs to be evaluated by the FCC. Although there are currently no federally mandated standards for RF exposure, several non-governmental organizations such as the American National Standards Institute (ANSI), the Institute of Elect rical and Electronics Engineers, Inc. (IEEE) and the Natio nal Council on Radiation Protection and Measurements (NCRP) have recommended limitations for human and RF electromagnetic field exposure. Several of these recommended limitations are [15] Limitations set for the span of 3 kHz to 300 GHz. Controlled environments (where energy le vels can be accurately determined and every person on premise is aware of the presence of EM fields) allow for higher power than that of uncontrolled enviro nments ( where energy levels are unknown and where personnel on facilities may be unaware of presence of EM fields) Lowest E-field exposure limits occur at frequencies between 30 and 300 MHz.( 1 mW/cm2 (61.4 V/m) controlled, 0.2 mW/cm2 (27.5 V/m) uncontrolled) Lowest H-field exposure limits occur at 100-300 MHz. (1 mW/cm2 (0.163 A/m)-controlled, 0.2 mW/cm2 (0.0728 A/m) uncontrolled Above 100 MHz, safety limits for E and H fields remain the same. Below 100 MHz, E-field radiation has lowe r power density limits than do H-field. The reasoning behind setting more stringent limitation on power densities within the frequency band of 30 MHz to 300 MHz, is th at the natural resonant frequencies of the human body occur between 30 to 300 MHz, a nd at frequencies above and below, the

PAGE 35

21 human body should absorb less en ergy. [16] A more graphic representation of IEEE RF Safety Guidelines can be seen in Figure 2-9. Figure 2-9. Proposed IEEE RF Safety Guidelines The guidelines set by the FCC and the safety standards recommended by IEEE, along with the uncertainty of the wireless ch annel between the tran smitter and receiver, show that the design of a reliable wireless front end is not a menial task; instead, it requires an intensive selection of proper components to optimize an already limited system.

PAGE 36

22 CHAPTER 3 SENSOR INTERFACE DEVELOPMENT The proper development of the hydrogen sensing mechanism and the analog to digital interface for this device is essentia l to the success of the design of a robust hydrogen sensing wireless sensor. Extensive effo rts must be made to fully understand the hydrogen sensing mechanism before a prope r design can be made for the interface between the sensing mechanism and an An alog to Digital converter (ADC). The hydrogen sensing mechanisms used were ZnO na no-rods developed and fabricated at the University of Florida. These nano-sensors pr oved themselves to be a robust solution for the missing role of a hydroge n sensing mechanism. The nano-rods were interfaced to an an alog to digital converter with careful considerations as prev iously described in Chapter 1. Af ter careful planning, a differential detection circuit consisting of an instrumenta tion amplifier topology is used to ease in the transition from a real world signal to a di gitized representation. This topology showed strong immunity to error sources as detailed from before, and served as a successful interface between the reactions due to hydrogen on the ZnO nano-rod and the analog-todigital converter of th e digital realm. The background, analysis, design, and im plementation of the ZnO nano-rod and instrumentation amplifier are detailed within this chapter.

PAGE 37

23 ZnO Nano-Rods The Zinc-Oxide (ZnO) nano-rods developed at the University of Florida between the collaboration of students and faculty of the Department of Materials Science and Engineering and the Department of Chemi cal Engineering, were used as the hydrogen sensing mechanisms of the system. These lightweight hydrogen sensors were designed with the goal of achieving high sensitivity, ra pid response to stimuli, reversibility, and low power consumption, all within a phys ically small and light package. The unique characteristics of ZnO nano-tubes and nanorods make them fundamentally appropriate candidates for th e sensing of hydrogen. ZnO is a material currently used in the detecti on of humidity, UV light, and gas, and has shown to change resistance with respect to both temperature and hydrogen exposure. Because of its wide bandgap of 3.2eV, the ease of synthesizi ng nanostructures, th e availability of heterostructures, and the bio-safe characteristics of this material, ZnO is a most attractive material for the specific sens ing application at hand [1,2]. With ZnO nano-rods placed in an array, as a gas sensor, they are able to create a large chemically sensitive surface-tovolume ratio which is needed for high sens itivity in hydrogen sensing. ZnO nano-rods can also be produced cheaply, and are highly compatible with other microelectronic devices. To enhance the devices sensitivity to hydr ogen, catalytic coatings or dopings of platinum (Pt) or palladium (Pd) were us ed to further increase the ZnO nano-rods hydrogen sensing mechanisms [1,2].

PAGE 38

24 ZnO Nano-Rod Fabrication Process The processes of growing and packaging the ZnO Nano-Rods for use as the hydrogen sensing mechanisms in the sensor sy stem are presented. Interested parties should read [1,2]. ZnO Nano-Rod site selective growth was accomplished by nucleating the NanoRods on discontinuous Au islands of nominal th ickness 20 coated with a substrate. It was previously shown that synthesis of ZnO Nano-Rods on wide assortment of substrates is a fairly uncomplicated task, increasing th e ease of Nano-rod synthesis [1] Deposition of ZnO Nano-Rods were achieved using molecu lar-beam epitaxy with a base pressure of 5 10-8 mbar using high purity (99.999%) Zn metal and O3/O2 plasma discharge as the source chemical. Growth time was approxima tely 2 h at 600 C. Resultant ZnO NanoRods grew to a typical length of 2 15 um, with diameters in the range of 30-150 nm. A schematic of a multiple ZnO Nano-Rod Sensor can be seen Figure 3-1 (a), while Figure 3-1(b) shows a scanning electron microgr aph of the home-grown ZnO Nano-Rods. Selected area diffraction patterns showed the ZnO Nano-Rods to be of single crystalline form. Additional coatings of Pt or Pd were deposited by sputtering in some cases, forming Pd thin films coatings of approximate thickness of 80 or Pt thin film coating of approximate thickness of 10 Electrodes on multiple ZnO Nano-Rods were created by using a shadow mask to pattern sputtered Al/Ti/Au electrodes contacting both ends of multiple ZnO Nano-Rods on Al2O3 substrates. Electrodes were separated through a spacing of approximately 400 um. Au wires we re then bonded to the contact pads of the package for transportation a nd current-voltage measurements. The packaged hydrogen sensing mechanism can be seen in Figure 3-2.

PAGE 39

25 S DZnO M-NRs Al2O3Substrate Al/Pt/Au a)b) S DZnO M-NRs Al2O3Substrate Al/Pt/Au a)b) Figure 3-1. ZnO Nano-Rods. (a) Schematic of Multiple ZnO Nano-Rodss (b) Scanning Electron Micrograph of ZnO Nano-Rods. Figure 3-2. Close-Up of Pack aged ZnO Nano-Rod Sensor Performance of ZnO Nano-Rods The fabricated and packaged ZnO NanoRods were tested under laboratory conditions where their reactions to hydrogen can be monitored under a controlled hydrogen environment. This controlled environment is within a hydrogen chamber

PAGE 40

26 located at the University of Florida, and allows for an accurate and quick assessment on the performance of the fabricated ZnO Nano-Rod sensors. Figure 3-3 shows the schematic of the hydrogen chamber, and a more detailed description on the operation of the chamber will be included in a later chapter. Figure 3-3. Simple Schematic of Hydrogen Chamber Used for ZnO Nano-Rod Testing + ZnONano-Rod 0.5 V + + ZnONano-Rod 0.5 V Figure 3-4. Schematic of Biasing for Zn O Nano-Rod Hydrogen Sensitivity Testing For the purpose of this section, the only information pertinent to understanding the performance of the ZnO Nano-Rods are the ch ambers functions of injecting controllable amounts of N2 and H2 to create different concentrati ons of hydrogen on the orders of parts per million (PPM) within the chamber. All measurements were performed using a

PAGE 41

27 HP 4156B Semiconductor Parameter Analyzer w ith the packaged ZnO Nano-Rods biased with a 0.5 V supply on one terminal, and ground on the other as seen in the schematic of Figure 3-4. Uncoated ZnO nano-rods The relative transient respons e of uncoated ZnO Nano-Rods as the gas ambient is switched from pure N2 to concentrations of hydrogen in nitrogen ranging from 10 to 500 PPM can be seen in Figure 3-5. Figure 3-5. Uncoated ZnO Nano-Rod Rela tive Resistance Change for Various Hydrogen Concentrations This shows a relative resistance change of approximately 0.70% for 500 PPM of H2 in N2 after 10 minutes of exposure, with incons istent results for lower concentrations. The gas-sensing mechanisms include the de sorption of adsorbed surface hydrogen and grain boundaries in poly-ZnO, exchange of ch arges between adsorbed gas species and the ZnO surface, leading to changes in depleti on depth and changes in surface or grain boundary conduction by gas adsorp tion/desorption [2]. Thus, it is shown the ZnO Nano-Rods are a suitable candidate for the sensing of hydrogen. The performance of un-

PAGE 42

28 coated ZnO Nano-Rods serves as a starti ng point for comparison to the increased hydrogen sensitivities of Pt or Pd coated ZnO Nano-Rods. Pd coated ZnO nano-rods Pd coated ZnO Nano-Rods have a relative transient response shown in Figure 3-6. Once again, this shows the reaction of the Pd coated ZnO Nano-Rods as the gas ambient is switched from N2 to increasing concentrations of H2 in N2 starting from 10 PPM to 500 PPM. Figure 3-6. Pd-coated ZnO Nano-Rod Rela tive Resistance Change for Various Hydrogen Concentrations By comparison, there is an approximate fi ve-fold increase in response to hydrogen as compared to that of uncoated devices. This shows that the addition of Pd appears to be effective in catalytic dissociation of the H2 to atomic hydrogen. The relative response of Pd-coated Nano-Rods can be seen as a function of H2 concentration in N2, where Pdcoated ZnO Nano-Rods were capable of de tecting hydrogen down to less than 10 PPM with a relative response of great er than 2.6% at 10 PPM. They have also shown to have greater than 4.2% relative resi stance change to 500 PPM afte r 10 minutes of exposure.

PAGE 43

29 Additionally, as seen in Figure 3-7, ZnO NanoRods show no response to the presence of O2 at room temperature. Figure 3-7. Pd-coated ZnO Nano-Rod. Abso lute Resistance Change for Various Hydrogen Concentrations The reversible chemi-sorption of reactive gases at the surface of metal oxides such as ZnO can provide a large and reversible vari ation in conductance of the material. Thus, the recovery of the initia l resistance upon removal from the hydrogen ambient is quick (less than 20 seconds). The sputtered Pd-ZnO Nano-Rod also shows to have an increased effective conductivity due to the presence of meta l. Palladium has proved to be a catalyst which can enhance the sensitivity of the hydrogen sensing mechanisms. Pt coated ZnO nano-rods Similar fabrication and tes ting procedures were done with Pt Coated ZnO NanoRods, except for the replacement of Pd with Pt as the sputtered metal. Figure 3-8 shows the relative transient response of the Pt-ZnO Nano-Rods as the ambient gas is switched from N2 to various concentrations of H2 in N2 ranging from 10 to 500 PPM.

PAGE 44

30 Figure 3-8. Pt-coated ZnO Nano-Rod Rela tive Resistance Change for Various Hydrogen Concentrations Similar transient characteristics can be seen between Pd-ZnO Nano-Rods and PtZnO Nano-Rods such as rapid recovery of ini tial resistance, with a 90% recovery within 20 seconds upon removal of the hydrogen from the ambient with the replacement of O2 or air, and no response to the presence of ambient O2 or N2 at room temperatures. However, the change from initial resistance after 10 minutes of exposure to H2 has an almost two fold increase as compared to Pd coated ZnO Nano-Rods. This shows that as a hydrogen sensing mechanism, Pt coated ZnO Nano-Rods have increased sensitivity to hydrogen, and would prove as a better sensi ng device than both Pd and uncoated ZnO Nano-Rods. Detection Interface for Hydrogen Sensitive Devices The challenge in designing the interface between a sensor and the Analog-toDigital (A/D) converter of a system is found in the necessity to obtain an accurate real world signal with the limitati ons of low power and reduced voltage swings. These limitations were previously discussed in chapter 1, and will be reiterat ed as the design of

PAGE 45

31 the differential detection ci rcuit is described. Furthermore, for accomplishing a longlifetime mode of operation, this sensor syst em is faced with the limitations of analog development, as well as the requ irement for low-power operation. One of the most important objectives of the sensor interface design is for the compatibility between the sensing mechanis m to the detection circuitry and ADC. Thankfully, ZnO Nano-Rods were chosen due to their high compatibility with microelectronic devices. The design will focus on meeting the demands of the ZnO Nano-Rods, power requirements, and meeting those requirements set by the resolution of the ADC input of the microcontroller. Add itionally, the differential detection interface should only detect reactions due to hydr ogen, and void all changes caused by other variables such as temperature. Given that the ZnO Nano-Rods in itial response to any exposure of hydrogen is distinct and immediate, this intrinsic characteristic will serve as an ally for the successful detection of hydrogen within the sensor system. This section will detail in depth the design of a differential detection sensor interfacefrom theoretical concep t, to considerations of co mponent selection, and finally an evaluation on the performance of the fabr icated and packaged differential detection circuit. Wheatstone Resistive Bridge The intrinsic characteristic of ZnO Na no-Rods that makes them suitable as hydrogen sensing mechanisms is their change in resistance with resp ect to how much and how long the device has been exposed to hydroge n. To accurately detect the presence of hydrogen in the ambient, since the nomina l resistance of the ZnO Nano-Rods is a function of hydrogen concentrations, the precise measurement of the change in resistance

PAGE 46

32 of the ZnO Nano-Rods can be used to corr elate the change in resistance to the concentration of hydrogen in the ambient. Bridge circuits, although old and primitive, continue to commonly serve as the best solution for the measurement of resistance, capacitance, and inducta nce. The resistive bridge instrument used for the measurement of an unknown electrical resistance is known as a Wheatstone bridge. The Wheatstone resist ive bridge as seen in Figure 3-9, invented by Samuel Hunter Christie in 1833 and popularized and improved by Sir Charles Wheatstone in 1843 [17], illustrates the concep t of using a difference measurement for obtaining the value of an unknow n electrical component. V2V1R3R1R4R2VgVs V2V1R3R1R4R2VgVs Figure 3-9. Wheatstone Resistive Bridge A difference measurement is taken across Vg seen in Figure 3-9, where Vg is dependent upon the nominal resistance values of R1, R2, R3 and R4. With R1 and R3 resistance values set and non-changing, th e voltage divisor combinations of R3/R4 and R1/R2 respectively, form the voltage Vg across V2 and V1. Vg can be equated as: 21 gVVV So,

PAGE 47

33 42 3412 g sRR VV RRRR Where Vs is the supply voltage Thus, it can be seen that if R4 and R2 are of equal values, the voltage across Vg should be 0 V, and any variation of R2 with respect to R4 will create a voltage drop across Vg. By measuring this voltage drop, assuming the reference resistor R4 has remained constant, we can decipher the change in resistance of R2. Because the ZnO Nano-Rods exhibit a ch ange to both temperature and hydrogen concentrations, the re lationship between R4 and R2 serves to be advantageous in providing a way for making the sy stem impervious to the variable of temperature. By using a passivated ZnO Nano-Rod device encased in glass as the reference resistor R4 and an un-covered and exposed ZnO Nano-Rod device as R2, additional changes in resistance for the exposed ZnO Nano-Rod due to effect s of temperature will be compensated by R4, and so the only changes seen at Vg are those caused by the exposed ZnO Nano-Rod reacting to the presence of hydrogen gas in the ambient. For the mode of low power operation, by setting resistors R1 and R3 as bias resistors much larger magnitude in comparison to that of R4 and R2, these resistors can be used to limit the power consumed by the bridge circuit Using the relationship of Voltage = Current Resistance, and Power = Voltage Current, using very la rge resistance values for R1 and R3 should limit the current passing through the two legs of the bridge circuit, and since power is directly proportional to current, limit th e overall power consumption of the Wheatstone Bridge. This limiting of current however, lowers the dynamic voltage swings of V2 and V1, which in turn makes the output voltage Vg very small in comparison to the supply

PAGE 48

34 voltage. To successfully detect the change in resistance, the voltage Vg must be amplified to meet the resolution requirements of the ADC. This requ ires the addition of an amplifier stage to buffer and amplify th e signal before processing by the ADC of the sensor system. Since the Wh eatstone resistive bridge uses the concept of differential measurement, the amplification stage must adhe re to the design of a differential detection interface. Differential Detection Interface The design of a differential detection in terface must meet the requirements of several issues noted in the pr evious section. Firstly, the de sign of the interface must remain steadfast to the original concept of using differential measurements to determine the detection of hydrogen, and la stly, the interface must have high large-signal gain to amplify the output of the current-limited Wheatstone bridge. Difference amplifier To remain a differential measurement instru ment, a difference amplifier as seen in Figure 3-10 from [18] is employed. The archit ecture of an operati onal amplifier with no feedback by itself is already a differential amp lifier with an uncontro llable gain. Typical inverting or non-inverting amplif iers are restrictive in that there is the practical loss of one of the two inputs; however, by using a difference amplifier topology, a nominal gain can be set, with both inputs remaining intact. By keeping both input s intact, the internal differential architecture of an operation amplifier can be exploited to be used as the differential measurement inputs. In a difference amplifier configuration, the an alysis of the circuit is essentially the same as that of an inverting amplifier. The only difference is th e non-inverting input of

PAGE 49

35 the operational amplifier being set to a voltage that is a fraction of V2 rather than ground, where V2 is dependent upon the resistance values of R3 and R2. R3R3R2R2 R3R3R2R2 Figure 3-10. Difference Amplifier The output of the operational amplifier, VOUT, can be found using the following equation: 3 21 2 OUTR VVV R Unfortunately, as compared to a non-invert ing amplifier configuration, the input to the difference amplifier has fairly low impedance. Because V2 and V1 of the Wheatstone resistive bridge is to be conn ected to the differencing inputs of the interface, if the inputs of the interface are of low i nput impedance, this will pose a serious problem to the accuracy of Wheatstone bridge due to the presen ce of a separate path for current to cause an inaccurate representation of output voltage at the output nodes of the Wheatstone

PAGE 50

36 bridge. Fortunately there is a simple solu tion. All that is needed is a non -inverting buffer, or a voltage follower to be added as seen in Figure 3-11. R3R3R2R2V1V2V3V4VOUT R3R3R2R2 R3R3R2R2V1V2V3V4VOUT Figure 3-11. Difference Amplifier with N on-Inverting Buffer to Differential Inputs An improvement to a simple voltage fo llower can be seen in Figure 3-12. This topology is known as an instrumentation amplifier. R3R3R2R2R1R1 RgV1V2V3V4VOUT R3R3R2R2R1R1 RgV1V2V3V4VOUT Figure 3-12. Instrumentation Amplifier

PAGE 51

37 Instrumentation amplifier Alike to the addition of a voltage follo wer as seen in Figure 3-11, Figure 3-12 shows an improved version of Figure 3-11 through the addition of three resistors connecting the two input buffer circuits to the difference amplifier. Using this topology, there can be the establishment of a gain stage before the large-signal gain of the difference amplifier, while maintaining a hi gh input impedance to isolate the Wheatstone resistive bridge from the f eedback resistor network of the difference amplifier. With the introduction of the three resist ors connecting the hi gh impedance input buffers to the difference amplifier, additional gain can be provided before the gain of the difference amplifier. This gain is ac hieved by creating a voltage drop across Rg from the isolated input voltages at V1 and V2. This voltage drop induces a current through Rg, and since the feedback loops of the input buffers draws little to no current, the same amount of current is drawn through the two resistors labeled R1. This produces a voltage drop across nodes 3 and 4 equal to: 1 34212 ()1gR VVV R With the combination of the input buffer stages, three connecting resistors, and difference amplifier, the total large-signal ga in of the instrumentat ion amplifier is found to be equal to: 13 21 22 ()1OUT gRR VVV RR The schematic for the full differential detec tion circuit can be seen in Figure 3-13. Here we see the Wheatstone resistive bridge serving as the input to the instrumentation

PAGE 52

38 amplifier, with the inclusion of the e xposed ZnO Nano-Rod, passivated ZnO Nano-Rod, Wheatstone current limiting bias resistors, and the feedback network of resistors, Rg, R1, R2, and R3 of the instrumentation amplifier. At this point, the design of the differential detection interface is completed, and the se lection of the proper components for the fabrication of the differential detection interface can start. + + + VDD GND GNDExposed ZnO PassivatedZnO R Bias R Bias R1R1RGR2R2R3R3VOUT + + + VDD GND GNDExposed ZnO PassivatedZnO R Bias R Bias R1R1RGR2R2R3R3VOUT Figure 3-13. Full Schematic for Differential Detection Circuit Realization and Testing of Differential Detection Circuit Thorough considerations from the limitati ons of commercial discrete components noted in Chapter 2 remained in mind for the selection of components. Instrumentation amplifiers intrinsically include very low DC offset, low drift, lo w noise, high open-loop gain, high common-mode rejection, and hi gh input impedances, making them highly accurate, and stable circuits for long and shor t term use [19]. From the discussions of

PAGE 53

39 chapter 2, heavy emphasis should be put on the selection of an opera tional amplifier with requirements of low supply voltage, low supply current, low DC input offset, low drift, high Common-mode rejection, and rail to rail voltage swi ng capabilities. Selection of operational amplifier A thorough internet search of availabl e low power operational amplifiers was conducted, and the results of this search can be summarized in Table 3-1. Table 3-1. Various Commercial Operational Amplifiers NAME Vsupply(MIN) Isupply(uA) VOS(MAX)(uV) VOS(TYP)(uV) Rail to Rail MAX4289 1 9 2000 200 Output MAX406 2.5 1.2 500 250 Output MAX478 2 17 70 30 no INA321 2.7 40 500 200 Output OPA336 2.3 20 125 60 Output TLV2401 2.5 0.88 1200 390 Input Output From Table 3-1, Maxim ICs MAX4289 was c hosen as the operational amplifier of choice. At the time, the MAX4289 showed to ha ve the best input offset as compared to power consumption. The most attractive pa rt of the MAX4289, is the minimum voltage range it was rated for. Out of all the othe r operational amplifiers studied, none were able to operate from a single-supply voltage as low as 1 V, while draining only 9uA and maintaining an input offset voltage of 200 uV. Simulation of differential detection circuit The assembly of all components involved both the components required for the instrumentation amplifier, as well as the re sistive bridge. For us e as the exposed ZnO Nano-Rods, new Pt-Coated ZnO Nano-Rods were grown and packaged for use. The nominal resistance change to th e injection of 500 PPM of H2 in N2 into the ambient can be seen in Figure 3-14.

PAGE 54

40 ZnO with increase Pt catalyst1400 1420 1440 1460 1480 1500 1520 1540 1560 15800. 00 1.75 3. 50 5.25 7. 00 8.75 10.5 0 12. 2 5 14.0 0 15. 7 5 17.5 0 19. 3 3 21.00 22 .8 3 24.50time(min)Resistance(ohms) Figure 3-14 New Pt-coated ZnO Nano-Rod Gr own and Packaged for Differential Detection Circuit Because current testing only involves testi ng at room temperature, the requirement for a passivated ZnO Nano-Rod to compensate fo r temperature deviations is not required. This simplifies the design in allowing the passivated ZnO Nano-R od to be replaced by a resistor. In order to tune the gains of the amplifier, the system was simulated using Agilent ADS, with the operational amplifiers pa rameters entered with the values seen in Figure 3-15, which are the parameters found on the datasheet for the MAX4289. After gain tuning, the Differential Detec tion circuit seen in Figure 3-13, was simulated with the component values found in Table 3-2. In the simulation, the exposed ZnO Nano-Rods resistance value was sw ept from 1565 ohms to 1461 ohms, which correlates to the resistance span of th e Pt-ZnO Nano-Rods with 500PPM of H2 in the ambient seen in Figure 3-14. The simulation set up of the Differen tial Detection circuit including parameters of the MAX4289 can be seen in Figure 3-15. The output of the

PAGE 55

41 instrumentation amplifier was plotted against the swept resi stance span, and can be seen in Figure 3-16. Table 3-2. Differential Detection Circuit Component Values PART VALUE Rg 470 kohms R1 2 Mohms R2 39 kohms R3 2 Mohms Rbias 270 kohms Passivated ZnO 1565 ohms Operational Amplifier MAX4289 Vout R R11 R=270k R R8 R=270k R R10 R=R OpAmp AMP3 BW=17 kHz VOS=.2 mV IOS=0.5 nA SlewRate=6e+3 CCom=0 F RCom=1 MOhm CDiff=0 F RDiff=50 MOhm Rout=100 Ohm CMR=85 dB Gain=90 dB OpAmp AMP2 BW=17 kHz VOS=.2 mV IOS=0.5 nA SlewRate=6e+3 CCom=0 F RCom=1 MOhm CDiff=0 F RDiff=50 MOhm Rout=100 Ohm CMR=85 dB Gain=90 dB OpAmp AMP1 BW=17 kHz VOS=.2 mV IOS=0.5 nA SlewRate=6e+3 CCom=0 F RCom=1 MOhm CDiff=0 F RDiff=50 MOhm Rout=100 Ohm CMR=85 dB Gain=90 dB ParamSweep Sweep1 Step=1 Stop=1461 Start=1565 SimInstanceName[6]= SimInstanceName[5]= SimInstanceName[4]= SimInstanceName[3]= SimInstanceName[2]= SimInstanceName[1]="DC1" SweepVar="R" PARAMETER SWEEP DC DC1 DC VAR VAR1 R=1565Eqn Var R R7 R=2 MOhm R R5 R=2 MOhm R R4 R=39 kOhm R R6 R=39 kOhm R R3 R=2 MOhm R R2 R=2 MOhm R R1 R=470 kOhm R R9 R=1565 Ohm V_DC SRC1 Vdc=2.0 V Figure 3-15. Agilent ADS 2003 Simulation Se tup for Differential Detection Circuit

PAGE 56

42 Figure 3-16. Agilent ADS 2003 Simulation: Output Voltage to Swept ZnO Nano-Rod Resistance Given the supply voltage of 2 V, the desi gned output power of the power IC power converters designed for the sensor system, and assuming a 10-bit ADC, the A/D has a resolution of about 2mV, with 1024 voltage le vels between 0 and 2V, and The output of the interface must be able to meet this re quirement and provide at least a 2mV per ohm (output voltage to ZnO resistance change) outpu t. Fortunately, from the graph of Figure 3-16, we can see that this requirement of reso lution can be achieved, with an approximate 4mV/ohm falling slope. The simulation also re veals the effects of the input offset voltages of the operational amplif iers. From Figure 3-16, it can be seen that for the case where both the exposed ZnO Nano-Rod and pa ssivated ZnO Nano-Rod is matched, there exists an approximate 200 uV DC offset. Th is is unfortunate, but can be compensated

PAGE 57

43 through software coding. The next step is fo r the design of a printe d circuit board (PCB) for use of fabricating and testing of the differential detection system. Fabrication of differential detection circuit The next step was to layout the schema tic of the instrumentation amplifier onto PCB, and assemble the device for testing. The designed PCB layout can be seen in Figure 3-17. Same component list as seen in Table 3-2 were used in the assembly, and the final assembled device with a packaged ZnO Nano-Rod device can be seen in Figure 3-18. Initial measurements for the fabricated se nsor interface can be seen in Table 3-3. Figure 3-17. Protel PCB Top and Bottom Layout Figure 3-18. Fabricated and Assembled Differential Detection Interface Board with Packaged ZnO Nano-Rod Sensor

PAGE 58

44 Table 3-3. Initial Measurements of Differential Detection Interface Supply Voltage ZnO Resistance H2 PPM Supply Current Power Sensor Output 2V 1565 0 42uA 84uW 30uV 2V 1522 10 44.2uA 88.4uW 152mV 2V 1500 500 44.3uA 88.6uW 210mV Given the A/D resolution of 2mV, from Ta ble 3-3, we can see that the sensor and sensor interface is capable of meeting the resolution requirements of the system while consuming minimal power, and remaining suff iciently impervious to the limiting effects described in chapter 2. A study was then conducted to measure th e linearity of the whole assembled differential detection device. For this experiment, the exposed ZnO Nano-Rod was replaced by discrete chip resistors ranging from 1 460 to 1562, while the passivated ZnO Nano-Rod was replaced by a chip resistor with nominal resistance of 1562. This procedure is similar to the simulation detailed earlier in th e chapter. 1562 ohms was used instead of 1565 due to the available value of re sistors. A plot of the performance for the linearity of the instrumentati on amplifier with respect to th e swept value of the exposed ZnO Nano-rod bridge resistance is seen in Fi gure 3-19. From these results, it is shown that the performance of the differential detect ion circuit is actually better than simulated results. The detection circuit shows to have good large-signal lin earity with an approximate -4mV/ohm slope, which matches the slope of the simulated system. In addition, for the case when both exposed ZnO Nano-Rods a nd passivated ZnO Nano-Rods are matched, the DC offset is only a mere 30 uV. This shows that the instrumentation amplifier is

PAGE 59

45 indeed impervious to the effects of input o ffset voltage, and the architecture has managed to compensate for the absolute input offs ets of the MAX4289 operational amplifiers. Software compensation of input offset volta ges mentioned earlier is unnecessary for the differential detection circuit. The next proces s would be to integrat e the sensor interface to that of the digital signal processing portions of the wireless hydrogen sensor node. Output voltage vs sweep of exposed Pt-ZnO Nominal Resistance 0 100 200 300 400 146014801500152015401560 Nominal Resistance (Ohms)Output Voltage (mV) Figure 3-19. Measured Output Voltage vs. ZnO Nano-Rod Resistance Sweep for Fabricated Differential Detection Circuit

PAGE 60

46 CHAPTER 4 MICROCONTROLLER DEVELOPMENT Microcontroller Selection The proper selection of a microcontroller was essential to the succ ess of the design. The system required the microcontroller to include an onboard ADC with enough resolution to track the changes of the se nsors and be capable of conditioning and processing the data received from the sensor interface. Enough onboard memory is needed to retain both the runtime code as well as store the data sampled by the ADC. There is also the requirement for our microc ontroller to have the ability to encode and send this data to the transmitter via a serial output. The system would be optimized if the microcontroller included a serial output port capable of s ourcing enough power to drive and power the transmitter. There also exists the requirement for the microcontroller to be easily reprogrammable and consume a minima l amount of power. Interested parties should reference to [20]. Because the initial goal is for a truly se lf-powered system, an emphasis must be placed on the assumption that the bulk of power will be consumed in the active states of the microcontroller and transmitter. Howeve r, because the time required for the ZnO Nano-rods to saturate is on th e order of minutes, the system can operate with a very low duty cycle, and so, idle or standby current also becomes a significant factor in the decision of which microcontroller to select. Eventually, Texas Instruments MSP430F 1232IPW was selected. This specific microcontroller was chosen because of its many features, large searchable

PAGE 61

47 knowledgebase, and the quality of assistan ce and samples given by TI. Table 4-1 highlights all the pertinent features of our microcontroller selection. Table 4-1. Features of Texa s Instruments MSP430F1232IPW Type of Program Memory Flash Program Memory 8 kB RAM 256 Bytes I/O Pins 22 pins ADC 10-bit SAR Interface 1 Hardware SPI or UART, Timer UART Supply Voltage Range 1.8 V 3.6 V Active Mode 200uA @ 1 MHz, 2.2 Vsupply Standby Mode 0.7 uA # of Power Saving Modes 5 Modes of Operation Currently the microcontroller is programmed to run as a state machine, and has two different reprogrammable modes of operation. In each mode of operation, the microcontroller operates within the following stat es: initialize, collect data, transmit data, and sleep. The first mode of operation is for the level monitoring of hydrogen. This mode runs through each state until a discerna ble threshold of hydrogen is detected. This threshold is set so that although hydrogen is present, the level of hydrogen is not enough to pose any serious danger. Once hydrogen is detected, the microcontroller forces the RF front-end to transmit an emergency pulse to the central monitoring station, and returns back to an idle mode. The second mode of operation is of da ta transmission. In this mode, the microcontroller collects data from the sensor interface, and queues this data to the RF front-end to be transmitted to the central mon itoring station. This mode is for a constant tracking of hydrogen levels, while the level monitoring mode is to alert the end user that

PAGE 62

48 hydrogen has indeed been detected. The st ate flow diagram for the Level Monitoring Mode and Data Transmission Mode can be seen in Figure 4-1. SLEEP Timer Set? NO Collect Data YES Transmit Pulse Initialization Analyze Data Threshold Detected? NO YES (a) SLEEP Timer Set? NO Collect Data YES Store Data Timer Set? SLEEP NO Transmit Data YES Initialization Analyze Data Threshold Detected?NO YES (b) SLEEP Timer Set? NO Collect Data YES Transmit Pulse Initialization Analyze Data Threshold Detected? NO YES SLEEP Timer Set? NO Collect Data YES Transmit Pulse Initialization Analyze Data Threshold Detected? NO YES (a) SLEEP Timer Set? NO Collect Data YES Store Data Timer Set? SLEEP NO Transmit Data YES Initialization Analyze Data Threshold Detected?NO YES (b) Figure 4-1. Microprocessor State Flow Diag rams. (a) Level Monitoring State Flow Diagram. (b) Data Transmission State Flow Diagram Power Requirements of Microcontroller To analyze the power consumption of the microcontroller during various stages of operation, a 383 ohm resistor wa s connected in series be tween the power supply, and microcontroller. Differential probes were used to measure the voltage across the 383 ohm resistor. Calculation for the aver age power consumption is as follows: Total Area (TA) under the measured curv e is calculated (units of Vsec) Peak power is calculated as: 383MAX eakPowerSUPPLYV PV Where VMAX is the maximum point of the measured curve Average power is calculated as 383otalrea SUPPLY durationTA AvgPowerV Where duration is the duration of the measured curve

PAGE 63

49 A power analysis was done by David J ohnson at Cisco to examine the power requirements of the controller system. From this analysis, it is observed that the power consumption for the microcontroller to remain idle, output data via serial power (either high or low bit), and to scan the ADCs input is a constant 2.5 uW. The most power consumed by the microcontroller at any time, is in the microcontrollers initialization state, which occurs only once during initia l power up of the microcontroller. The initialization time for the microcontroller is only for 12.5ms, where average power consumption is 3.07mW with a peak power of 7.3mW (as seen in Figure 4-2). Figure 4-2. Initialization Powe r Required for MSP430F1232IPW

PAGE 64

50 CHAPTER 5 LOW-POWER WIRELESS COMMUNICATION LINK In the design of wireless sensors, the most power consuming component is often found in the wireless front-end. To make matters worse, components within a RF transceiver/transmitter/receivers such as power amplifiers or oscillators, at best, only have efficiencies slightly better than 50% [ 10]. This means at best performance, to transmit 100mW of power, the device will require 200mW of power. However, the effects of low efficiency can be mitigated through several techniques. Because typical sensor nodes remain in idle states much l onger than in active states, the sensor nodes themselves are of very low duty-cycle. By using low duty-cycle and low data rates, components within the transmitter for the wirele ss sensor can be turned off when no data is present to be transmitted, and the entire tr ansceiver/transmitter/receiver can be placed into a low-power sleep mode. The consider ations for the design and implementation of the wireless communications link ar e detailed in this chapter. Selection of a Modulation Technique From the previously mentioned reality of both the nature of the sensor system, as well as the limitations of the components within a RF transceiver, transmitter, and receiver, there exists a m odulation technique which can take advantage of all the mentioned limitations and requirements of the sy stem. Firstly, the system must be able to obtain power from scavenged sources, a nd allow for long life-time operation with minimal maintenance, which reduces the comple xity of the wireless front end from that of a transceiver to that of a lone transmitter. Additionally, because the system itself is of

PAGE 65

51 low duty-cycle and lower data rate, and because of the low efficiencies of RF discrete components, a modulation scheme is requi red which can exploit the concept of consuming power only when transmitting data, and requires a simple transmitter architecture, with few discrete RF parts c ounts. By using a modul ation scheme of low complexity, the depth of modulation can be realized through a tran smitter architecture with fewer discrete power consuming components. Because these components intrinsically show poor efficiency, a lowe r component count will further reduce the power consumption for the RF front end. -DQPSK OOK 4 a) b) -DQPSK OOK -DQPSK OOK 4 a) b) Figure 5-1. Tradeoffs Between Performa nce and Architecture Complexity of /4 DQPSK and OOK. (a) BER performance (b) Architecture Complexity The simplest modulation scheme available is that of a carrier present, carrier absent technique, also k nown as On-Off Keying ( OOK). What made OOK an appealing modulation scheme was the intrin sic premise that an OOK transmitter would only be on and consuming power when the RF front end was transmitting a high or a , and that the transmitter has the advantage of going into an idle state so that little to no power would be consumed on the transmissi on of a low or . A comparison of the transmitter architecture and performance for /4 DQPSK as a reference to OOK [21]

PAGE 66

52 can be seen in Figure 5-1. It is shown that the /4 DQPSK transmitter architecture is significantly more complex than that of an OO K transmitter. This complexity is a trade off between performance and complexity (correlating to power consumption) for the selection of a modulation scheme. The disadvantage of OOK modulation howev er, is found in the error caused by the presence of unwanted or undesirable si gnals. Typically, OOK is an unappealing modulation scheme for networks of heavy traffi c, but because a sensor system rarely will transmit with a duty cycle of more than 25% of the time, OOK modulation is suitable for use as the modulation scheme of the wirele ss sensor node. OOK differs from Amplitude Shift Keying (ASK) in that there is no carrier present during the transmission of a zero. This allows for additional power reduction on the transmitter side, however, it allows OOK to be more susceptible to an interfering signal maki ng detection by the receiver more difficult as compared to ASK. This is a performance trade-off between ASK and OOK. a)b)c) a)b)c) Figure 5-2. Signal Constellations. (a) OOK( b) ASK(c) and as a reference, FSK The signal constellations for ASK and OOK can be seen in Figure 5-2 from [22], with the signal constellation for FSK as a re ference for comparison. For the design of the

PAGE 67

53 wireless sensor node, because power requi rements out-weigh any other requirements, OOK is selected as the choice of modulati on due to the characteristics of OOK which make it a highly power conserving form of modulation. Selection of Operating Frequency Once a modulation scheme is selected, the selection of which frequency band to operate within is needed. The considerations for the selection of which frequency band are heavily dependent on the limitations stat ed in Chapter 2 considering impairments of the wireless channel, FCC regulations, a nd the modulation scheme derived in the previous section. Because OOK is highly sens itive to interference signals, the selection of the operating frequency is very important for the success of the wireless system. Path Loss (dB) vs Frequency (Hz) 1 1051 1061 1071 1081 1091 1010 50 0 50 100 Frequency(Hz)Path Loss (dB)PLf cf c Figure 5-3. Path Loss Attenuation(dB) w ith Respect to Carrier Frequency. As seen in Figure 5-3, regarding free space path loss, it can be seen that as operating frequency is increased, for a given distance, the attenuation of the propagation path increases. Thus, with a lower frequency, longer transmission distances can be

PAGE 68

54 achieved with lower power. Since lower power is required to attain a particular distance, less output power is needed, which in turn reduces the power consumption requirements of the transmitter. This however comes at a cost. From the discussion of Chapter 2, it was also mentioned that the most stringent output pow ers were placed on the frequency band of 30 to 300 MHz. To reiterate, this was done to limit RF power absorption by the human body, because the human body is naturally resona nt between those frequencies. Thus, a trade-off exists in that although limited in output power, using a lower frequency increases the transmission distance, but decreas es the required output power to transmit a certain distance. Another factor to consider is the tra ffic involved in each fr equency band. Common operating frequencies such as the 902-928 MH z band, are constantly being used with continuous transmissions of voice, data, vi deo, and offer high level interference from microwave ovens and spread spectrum devi ces. Because an OOK m odulation scheme is highly susceptible to interference, the overc rowding of frequency bands within 900 MHz, 2.4 GHz, and 5 GHz can prove to be treacherous to a low complexity modulation scheme such as OOK. Other lower frequency bands such as the 260 to 470 MHz bands are much more open and less crowded. Typical frequenc ies within these bands such as 315 MHz, 418 MHz, and 433.92 MHz only compete with ga rage door/keyless entry systems, or interference from amateu r radio users [23]. Due to these factors, and the availability of commercial products, the operating frequency of 300 MHz to 315 MHz within the 260-470 MHz FCC frequency band was selected. This operating frequency offers a fairly interference free band, but is shown to

PAGE 69

55 have some unusual restrictions specific to the frequency band of 260 to 470 MHz set by the FCC under part 15.231 which will be detailed later on in the chapter. Selection and Performance of a RF Transmitter To simplify the design of the wireless sensor node, commercial RF transmitters were chosen to be used as the RF front e nd of the sensor system. After selection of modulation scheme and operating frequency, an internet search was performed to find available transmitter/receiver pair pack ages capable of OOK within the 315 MHz operating frequency band. A brief listing de tailing available transmitters with their corresponding performance specifications can be seen in Table 5-1. Table 5-1. Comparison of Available Commercial ASK/OOK Transmitters MODEL VMIN VMAX ISUPPLYMAX ISUPPLYMIN Output Power(min) Output Power(max) Freq TYPE LINX TXM315-LR 2.1 3.6 5.1mA 1.8mA -4 dBm 8 dBm 315 MHz SAW MAX1472 2.1 3.6 9.1mA 1.5mA 3.3 dBm 6 dBm 315 MHz Crystal MAX1479 2.1 3.6 6.7mA 2.9mA 2.7 dBm 5.3 dBm 315 MHz Crystal Ming TX-99 ? 5 1.6mA ? ? ? 300 MHz LC Atmel U2741B 2 5.5 12.5mA? 1.5 dBm 5 dBm 315 MHz Crystal The list found in Table 5-1 was whittled dow n to two choices for a RF front-end the TX-99 manufactured by Rayming Corpor ations, and the TXM-315-LR manufactured by Linx Technologies. These specific transmitte rs were selected due to their low power consumption, low component count, and low comp lexity for the ease of rapid prototyping and development while meeting the requirements for a low-power OOK transmitter operating within the 260 to 470 MHz frequency band. Rayming Corporation TX-99 300 MHz AM Transmitter/RE-99 Receiver Pair From a previous project, th ere was a Ming TX-99 transmitte r/ RE-99 receiver pair available for use for this project. The di fficulties of the Ming TX -99 were that little

PAGE 70

56 documentation was provided for both the company, and the device. The datasheet for the Ming TX-99 offered very few maximum and minimum operating conditions. However, what made up for the difficulties of the Ming TX-99 was the simplicity of the design. Ming Tx-99 transmitter The architecture of this transmitter is base d on a colpitts oscillator design seen in Figure 5-4 and consists of a single high frequency NPN BJT tr ansistor and a LC tank to tune the transmitter to oscillate at a specific frequency. Figure 5-4. Schematic of Ming TX-99 Taken from Datasheet Initial performance tests show that when biased at 0.6 V, the transmitter drains 850uA, which translates a power requirement of 510uW to transmit a constant 50% duty cycle 580mV peak to peak square pulse train of 1 kHz. The LC tank included a variable capacitance with a range of 27 pF to allow for the tuning of the operating frequency. Another advantage of the Ming TX-99 tran smitter was the onboard antenna. The onboard antenna served as the inductor for the LC tank. With the Supply Voltage and Data nodes tied together as seen in the schematic of Figure 5.5, the transmitter can be used as an OOK transmitter. The whole transmitter module can be seen in Figure 5.6 with the onboard antenna hi ghlighted for detail.

PAGE 71

57 VDD GND VDD GND Figure 5-5. Ming TX-99 Transmitter in OOK Mo de. VDD Is Replaced with Data Stream Figure 5-6. Ming TX-99 Transmitter. Red Outline Highlights Onboard Antenna Additionally, the printed micr o-strip inductor which serv es as the onboard antenna can be tapped for attachment of an extern al antenna. For output power measurements, the signal pin of a SMA connector was sold ered and tapped to th e micro-strip inductor where an external antenna w ould be attached to, with gr ound of the SMA connector tied to ground of the Ming TX-99. The SMA was then attached to one end of a 1ft SMA cable (FLX402#1), and the other end to a DC bloc ker before finally attached to the input of a HP 8563E 9kHz to 26.5 GHz Spectrum An alyzer. The DC blocker was used to prevent damage to the spectrum analyzer by bl ocking DC current from directly entering the input of the spectrum analyzer. Both th e supply voltage and data node of the Ming

PAGE 72

58 TX-99 were connected to a power supply (Agilent E3631A) set to 2 V to send a 100% duty cycle signal (transmitter co ntinuously on). Test setup can be seen in Figure 5-7, and the transmitter with attached SMA c onnector can be seen in Figure 5-8. Figure 5-7. Test Setup for Output Power and Power Consumption of Transmitters Figure 5-8. SMA Connector Soldered to Antenna Tap on Ming TX-99 Transmitter

PAGE 73

59 The result was an output power of approximately -4.67 dBm while draining 1.95mA continuously from a 2V supply. To test for minimum vo ltage operation, for 100% duty cycle, the transmitter was capable of operating at 1.2 V while draining 290 uA at an output power of -21.17 dBm. Assu ming efficiency can be calculated as: OutputPower SupplySupplyP Efficiency VI The data found in Table 5-2 can be tabulated. Table 5-2. Performance of Ming TX-99 VSUPPLY ISUPPLY(mA) Output Power(dBm) Power(mW) Efficiency(%) 1.2 0.29 -21.17 0.007638358 2.194930413 2 1.95 -4.5 0.354813389 9.097779211 Ming RE-99 receiver Since a central monitoring station can be assumed to provide as much power as needed, the power requirements are not as st ringent on the receiver side, making the RE99, the receiver complement of the TX-99 a su itable receiving unit. Alike to the TX-99 transmitter, the RE-99 also lacks in docume ntation. From the schematic found in the datasheet as seen in Figure 5-8, it is seen th at the RE-99 is an enve lope detection circuit which is typical for AM/ASK/OOK receivers. Figure 5-8. Schematic of Ming RE-99 Taken from Datasheet

PAGE 74

60 Alike to the TX-99 the RE-99 provides an onboard antenna which also serves as the LC tank resonating at 300 MHz as the input for the receiver. It differs from the TX99 in that it is no longer a micro strip antenna, but a loop antenna of 2 turns, and diameter of 5mm. It also offers a tap for an extern al antenna, and as recommended, a quarter wave monopole antenna was created by cutting a 22 gauge copper wire down to a length of 9.36 inches, which is approximately the wa velength for 300 MHz, and soldered to the antenna tap. Unfortunately, the sensitivity of the receiver was unable to be measured, but instead, a distance measurement was performed. A picture of the RE-99 receiver without the external antenna can be seen in Figure 5-9. Figure 5-9. Ming RE-99 Receiver Ming distance measurements Because the sensitivity of the receiver is unknown, it was decided to perform an experiment to find the maximum transmissi on distance. The experiment was conducted in the atrium on the first floor of the New Engineering Building at the University of Florida. The floor plan of the atrium can be seen in Figure 5-11. The setup is detailed as follows: Transmitter was tied to serial output (U SART) of a microcontroller (MSP430) outputting a constant data stream. The seri al output of the mi croprocessor was tied to both the supply voltage and data n ode of the transmitter, forcing an OOK modulation scheme, while providing power to the transmitter. The microprocessor was powered with a 2V supply.

PAGE 75

61 The transmitter remained stationary and the receiver was attached to a cart for mobility. The height of the transmitt er and receiver were 0.45m and 0.55m respectively. The receiver was powered via another power supply. A diagram of this setup can be seen in Figure 5-12. The output of the receiver was tied to the input of a Tektronix TDS210 Two Channel Digital Real-Time Oscilloscope For received power measurements, a 22 gauge copper quarter wave monopole antenna soldered to a SMA connector wa s connected to an Agilent E4448A PSA Series Spectrum Analyzer AtriumHallway Hallway Atrium3.5 m 10 m 20 m 0 mTransmitter Figure 5-11. Floorplan of First Floor Atrium of New Engineering Building 0.45 m 0.55 mDistance (m)Transmitter Receiver 0.45 m 0.55 mDistance (m)Transmitter Receiver Figure 5-12. Experimental Set up for Distance Measurements Within the testing, the placement of a 22 gauge copper quarter wave monopole antenna was placed on the transmitter, receive r, or both. The maximum transmission

PAGE 76

62 distances for these test cases can be seen in Table 5-3. From this data, it shows that with a quarter wave monopole antenna on both th e transmitter and receiver, the maximum distance for the successful dete ction of the original data st ream serially outputted by the microprocessor, was found to be 19.4 m. Table 5-3. Maximum Transmission Distan ces with Varying Antenna Locations Antenna Location Maximum Distance Receiver Only 14.5 m Transmitter Only 16.8 m Transmitter & Receiver 19.4 m In addition to this maximum transmissi on distance experiment, measurements for received power were taken as well. These measurements were taken from an Agilent E4448A PSA Series Spectrum Analyzer connect ed to a quarter wave monopole antenna. For all measurements, the transmitter also had a matched quarter wave monopole antenna. Figure 5-13 shows the received power versus distance, with reference to the layout of the testing environment, the at rium of the New Engineering Building. Atrium -75 -70 -65 -60 -55 -50 -45 -40 -35 05101520Distance (m)Received Power (dBm) Atrium -75 -70 -65 -60 -55 -50 -45 -40 -35 05101520Distance (m)Received Power (dBm) Figure 5-13. Received Power vs. Distance With Reference to Room Shape From this Figure, it can be seen that at around 10m, the hallways of the floor plan began to act as a sort of waveguide. This caused the received power to increase after 10

PAGE 77

63 meters was reached. From the maximum di stance experiment, it is shown that the maximum distance achieved was set to be around 19.4m. At that distance, the received power was approximately -70 dBm. From this, it can be concluded th at the sensitivity of the receiver is approximatel y -70 dBm. Figure 5-14 shows the power spectrum taken from the spectrum analyzer at 1m, and 8m respectively. a)b) a)b) Figure 5-14. Screen Capture of Received Power Spectrum. (a) at 1m (b) and 8m Linx Technologies LR series Transmitter and Receiver Unfortunately, Rayming Corporation no l onger exists. As time goes by, locating the Ming TX-99 or RE-99 for purchase becomes increasingly difficult as suppliers have depleted their stocks with no new shipment s coming in to replenish their supply. This prompted the selection of the Linx LR series transmitter and receiver, which is an update from their LC series line of transmitters and receivers. With these transmitters and receivers, comes a plethora of applica tion notes and documentation to aide in the design of a wireless communications link. Linx Technologies also provides prefabricated low-profile antennas for use with the transmitters and receivers. The outstanding product provided by Linx Technol ogies, as well as the plethora of

PAGE 78

64 documentation notes, makes rapid prototyping and development with the LR series transmitter and receiver fairly painless. Linx Technologies TXM-315-LR The LR series transmitter from Linx Technologies is a high performance synthesized ASK/OOK transmitter which has the ab ility to reach a serial data rate of 10 kbps. The transmitter consists of a PLL s ynthesized architecture offering low-power consumption, accurate operating frequency, a nd power-down functions with an antenna serving as the only external part needed. The system level architecture of the transmitter can be seen in Figure 5-15 from the TXM-315-LR datasheet. Figure 5-15. System Level Arch itecture of LINX TXM-315-LR The components of the transmitter consist of a Voltage Controlled Oscillator (VCO) locked through a phase locked loop (PLL) which is referenced to a high precision crystal. The output of the VCO is then amplified and buffered by a power amplifier before the carrier is filtered to attenuate and suppress harmonics and spurious emissions to within legal limits. The carrier is then output to free space via the 50 ohm antenna port of the transmitter. The pin-out of the TX M-315-LR transmitter can be seen in Figure 516.

PAGE 79

65 Figure 5-16. Pin-Out of TXM-315-LR Transmitter Several unique features such as a Power Down line (PDN) which is used to power down the transmitters power amplifier when no data is present for transmission, and a Level Adjust line (LADJ) which is used to limit the output power for the transmitter, can be used in unison to even further reduce the power consumption of the transmitter. The LADJ line can prove to be even more usef ul during FCC testing and verification to compensate for antenna gains. By tying th e PDN, supply voltage, and Data input node together, and driving this node with the se rial output of the microprocessor, the transmitter, like the Ming TX-99, can be used as an OOK transmitter. Initial test setup where the antenna port is soldered to a 50 ohm micro-strip, and connected to a spectrum analyzer (Agilent E4448A PSA series Spectrum Analyzer) via a 1 ft long SMA cable show the minimum bi as of 1.6V to send a 100% duty cycle signal (transmitter continuously on) requires 6 mA to output 0.34 dBm of output power. Additionally, at 2V bias, th e transmitter can output approximately 3.08 dBm while draining 8 mA of current. A list of power specifi cations, with efficiency calculated in the same fashion as for the Ming TX-99 analysis detailed previously, is seen in Table 5-4. Table 5-4. Performance of LINX TXM-315-LR VSUPPLY ISUPPLY(mA) Output Power(dBm) Power(mW) Efficiency(%) 1.6 6 0.34 1.081433951 11.26493699 2 8 3.08 2.032357011 12.70223132

PAGE 80

66 As compared to the Ming TX-99, alt hough output power and efficiency are both better than the TX-99, the minimum power required to turn the transmitter on is significantly higher, and may pose a serious problem when trying to obtain power from scavenged energy. This shows that although Li nx Technologies TXM-315-LR is a more robust and mature commercial product, th e most important requirement of minimal power expenditure is not met, and may not serve as a suitable RF transmitter for the wireless hydrogen sensor node. Linx Technologies RXM-315-LR Alike to the assumption made for the Ming RE-99, the same assumption that a central monitoring station can provide as much power as deemed necessary is also made for the RXM-315-LR. Unlike the products by Ming, and alike to the TXM-315-LR, there exists a plethora of documentation for the RXM-315-LR, as well as several application notes. The system level architecture of th e RXM-315-LR is seen in Figure 5.17, and unlike the envelope detecti on circuit of the RE-99, the RXM-315-LR receiver modules employs a single-conversion super-heterodyn e architecture to demodulate the received signal. Figure 5-17. System Level Architecture for RXM-315-LR

PAGE 81

67 The RF signal entering from the 50 ohm matched antenna is band-pass filtered before being amplified by an NMOS cas cade, Low Noise Amplifier (LNA). The amplified signal is then down-converted to a 10.7 MHz Intermediate Frequency (IF) which is done by mixing the amplified signal with a VCO controlled by a PLL, referenced to a high precision crystal. The mixer stage, which down-converts the signal, consists of a pair of double balanced mixers, and includes an image rejection circuit. Once down-converted to an IF frequency, the signal is further amplified, filtered and demodulated to recover the original base-ba nd data bit-stream. This baseband signal is squared by a data slicer and output to the DATA pin of the receiver. This architecture, along with the high IF frequency and ceramic IF filters, helps reduce the susceptibility to interference, which is a problem associated with OOK modulation. The pin-out of the receiver can be seen in Figure 5.18. Due to the architecture and components of this receiver, it is able to achieve a very high sensitivity of -112 dBm, while remaining unsusceptible to interfering sign als which plague OOK communication links. Figure 5-18. Pin-Out of RXM-315-LR Transmitter Compared to the Ming RE-99 receiver, Linx Technologies RXM-315-LR receiver shows to have the best performance out of both receivers, while the Ming TX-99 shows

PAGE 82

68 to have the most favorable performance between the two transmitters. To optimize the wireless communication link, the combina tion of the Ming TX-99 and Linx RXM-315LR should be used together. Unfortunately both components are set to operate at different frequencies. As stated before, one of the attractive charac teristics of the Ming TX-99 is the tunable variable capacitor whic h can change the resonant frequency of the LC tank. If the LC tank were to be tuned to 315 MHz, the Ming TX-99 can be used in conjunction with the RXM-315-LR to create an optimized wireless link for the wireless hydrogen sensor node. Wireless Link Optimization The optimization of the wireless link includes not only the combination of the MingTX-99 transmitter with the Linx Technol ogies RXM-315-LR receiver, but also the minimization of power consumption by the Ming TX-99 transmitter, development of a low-profile antenna to reduce the size a nd increase the compactness of the overall wireless sensor package, and work on the receiver side to gather data from the transmitter for use as a central monitoring station. Add itionally, studies should be done to develop a wireless node operating within the restrictions set by F CC part 15.231, which regulates operation within the 260-470 MHz range. Prio r to all the additional work, a starting ground should be set by performing an in itial power analysis of the system. Ming TX-99 Power Analysis A power analysis on the Ming TX-99 wa s performed by David Johnson, at Cisco Systems in Bradenton, FL. Similar to the power analysis performed for the microcontroller in Chapter 4. To reiterat e the test setup, a 383 ohm resistor was connected in series between the power suppl y, and microcontroller/RF transmitter. The USART output of the microcontroller was used to power and drive the Ming TX-99

PAGE 83

69 transmitter. P6248 Differential probes att ached to a Tektronix TDS5104B Digital Phosphor Oscilloscope were used to measur e the voltage across the 383 ohm resistor. Calculation for the average power consumption is as follows: Total Area (TA) under the measured curv e is calculated (units of Vsec) Peak power is calculated as: 383MAX eakPowerSUPPLYV PV Where VMAX is the maximum point of the measured curve Average power is calculated as 383otalrea SUPPLY durationTA AvgPowerV Where duration is the duration of the measured curve It was previously shown in Chapter 4, that for any task other than for initialization, the MSP430 microcontroller only consumed a lo w 2.5uW of power. To test the average power consumption for the transmission of a 500 uS bit length pulse the procedure above is applied, and the measured curve take n from a screen dump of the Tektronix TDS5104B Digital Phosphor Oscilloscope can be seen in Figure 5-19. Figure 5-19. Tektronix TDS5104B Digital Phos phor Oscilloscope Screen Capture of Power Analysis Performed for RF Transmission of One Bit

PAGE 84

70 From this curve, it can be calculated that for the transmission of a 500 uS pulse, the transmitter consumes an average power of 261 uW with a peak of 522 uW. As expected, the RF transmitter does not consume power for the RF transmission of a logical low or . The charging and discharging characteri stics of Figure 5-19 may be due to the LC resonant tank which is used to set the operating frequency of the RF transmitter to 300 MHz. Another experiment was performed to an alyze the power consumption for the RF transmission of multiple bits. The screen capture from the Tektronix TDS5104B Digital Phosphor Oscilloscope can be seen in Figure 5-20. From Fi gure 5-20, it can be seen that all rising slopes and falling slopes are equal. Figure 5-20. Tektronix TDS5104B Digital Phos phor Oscilloscope Screen Capture of Power Analysis Performed for RF Transmission of Multiple Bits By grouping all rising slopes together at the front, a nd grouping all falling slopes together in the back to arrange a simple triangle, and assuming that the power is not completely discharged between high bits, it can be gathered that the worst case

PAGE 85

71 average power consumption (maximum average power consumption) to send N bits, can be equated as: 1 2eakPowerbitsWorstCasePowerPN Where PeakPower is the peak power required to transmit a single bit (261 uW), and the worst case data input to the transmitter is a continuous train of or high pulses, one after the other with no low or bits in between. Low Profile Antenna To increase the compactness of the wirele ss sensor, Linx Tec hnologies provides the ANT-315-SP, or SPLATCH antenna. The featur es of this antenna are an ultra-compact package and good resistance to proximity effects. The SPLATCH uses a grounded-line technique to achieve a quarte r wave type antenna centered at 315 MHz, with a bandwidth of 5 MHz. The SPLATCH antenna with di mensions can be seen in Figure 5-21. Figure 5-21. LINX ANT-315-SP SPLAT CH Antenna From Datasheet

PAGE 86

72 Unfortunately, unlike the TXM-315-LR tr ansmitter or RXM-315-LR receiver, the SPLATCH antenna does not provide much in documentation. The only details given are the antennas requirements for a 1.5 x 3.0 ground plan, and a 50 Oh m micro-strip line between RFIN of the antenna, and the RF out put node of the transmitter. An assembled FR4 testing board for the SPLATCH ante nna can be seen in Figure 5-22. Figure 5-22. Testing Board for SPLAT CH Antenna. (a) Front(b) and Back Initial tests to obtain the gain of the SPLATCH antenna were performed with an Agilent E8316A 10 Mhz to 6 GHz PNA seri es Network Analyzer, and an Agilent E4448A PSA Series Spectrum Analyzer. An antenna test structure was tied to the Agilent E8254A 250 kHz to 40 GHz PSG-A Series Signal Generator via a FLX402#1 SMA cable to serve as the transmitting ante nna, and an identical antenna structure was tied to the Agilent E4448A PSA Series Spec trum Analyzer outputti ng a RF power of 10 dBm at 315 MHz, via a FLX402#1 SMA cable to serve as a receiving antenna. With respect to the free space pa th loss equation, where

PAGE 87

73 2 2() (4)TTR RPGG PR R Assuming frequency, distance, and transmitted and received power are all controlled parameters, since the transmitter and receivers antennas are identical, their respective gains can be calculated. For these calculations, cable losses were also taken into account. Initial gain measurements from different distances showed the data found in Table 5-5. As a comparison, the gains of the 22 gauge copper quarter-wave monopole antennas were also tested to serve as a comparison to the SPLATCH antennas. Table 5-5. Antenna Gain Measurements Distance (m) Received Power (dBm) Transmitted Power (dBm) Path Loss (dB) Cable Loss (dB) Gain (dB) Antenna 4 -58.11 10 34.449183090.41 -16.6254 SPLATCH 5 -60 10 36.387383350.41 -16.6013 SPLATCH 4 -26 10 34.449183090.41 -0.57041 Monopole 5 -28 10 36.387383350.41 -0.60131 Monopole From these measurements, it can be seen that the SPLATCH antennas exhibited poor performance with an approximate gain of -16.6 dB, as compared to the -0.6 dB gain of the monopole antennas. Clearly, there is a distinct performan ce loss of the SPLATCH antenna over the monopole antennas, and a more comprehensive analysis of the SPLATCH antenna should be conducted. To measure the resonant frequency of th e SPLATCH antenna, a 1 Port S-parameter measurement was taken using an antenna, a nd an Agilent E8316A 10 Mhz to 6 GHz PNA series Network Analyzer. Additionally, the antenna test structure was simulated in Ansoft Designer. From the S-Parameter measurements taken by Agilent E8316A 10 Mhz to 6 GHz PNA series Network Analyzer shown in Figure 5-23, the SPLATCH antennas

PAGE 88

74 showed to have a resonant frequency located at 330 MHz with a bandwidth of 5 MHz, rather than 315 MHz as shown in th e specifications of the datasheet. Figure 5-23. S-Parameter for SPLATCH Antenna Since the fundamental operating frequency is 315 MHz, this may be the reason for the poor performance of the antenna. A plan ar EM simulation in Ansoft Designer was performed, and showed similar results compar ed to the measurements taken from the E8316A 10 Mhz to 6 GHz PNA series Network An alyzer as seen in Figure 5.24(a). Also seen in Figure 5.24(b) is the corresponding an tenna test structure model within Ansoft Designer. To shift the resonant frequency of th e SPLATCH antenna down to 315 MHz, a matching circuit consisting of a shunt and se ries capacitor were simulated in Ansoft Designer, and then realized and tuned on th e SPLATCH antenna test structure. The

PAGE 89

75 frequency shifted matched antenna S parame ters can be seen in Figure 5-25 with comparison to the original un-matched ante nna. Figure 5-26 shows that the matched antenna retains an approximate -10 dB bandwidth of 5 MHz. Port1 Port2 1034mil 428mil 1 4 9 3 0 0 5 4 m i l 3005mil a)b) Port1 Port2 1034mil 428mil 1 4 9 3 0 0 5 4 m i l 3005mil a)b) Figure 5-24. S-Parameters of Measured (red) and Simulated (blue) in Ansoft Designer. (a) S11 (b) Simulation Setup Figure 5-25. Matched Antenna (Red) vs. Unmatched Antenna (blue)

PAGE 90

76 Figure 5-26. 5 MHz Bandwidth of Matched Antenna Once matched, the SPLATCH antenna was agai n tested for gain. The experiment setup is identical to the previously mentione d set up, and the data for this experiment can be found in Table 5-6. Table 5-6. Gain Measurements for Matched Antenna Distance (m) Received Power (dBm) Transmitted Power (dBm) Path Loss (dB) Cable Loss (dB) Gain (dB) 2 -49 4 28.42858318 0.11 -12.2307 3 -50 4 31.95040836 0.11 -10.9698 4 -48 4 34.44918309 0.11 -8.72041 These results show an approximate 6 dB in crease in performance, but the resulting gain is still lower than that of the quarter wave 22 gauge copper monopole antenna. This may be due to the poor radiation effici ency stemming from electrically small characteristics of the antenna as compared to the 22 gauge copper monopole antennas. A trade-off thus exists in the reduction of ga in for a compact surface mount antenna. For

PAGE 91

77 optimization purposes, since size is an un limited parameter at the central monitoring station, the quarter wave 22-ga uge monopole antenna can be used as the antenna for the receiver, while the SPLATCH antenna can be us ed as the antenna for the wireless sensor node. RF Transmitter Optimization As previously mentioned, an attractive feature of the Ming TX-99 transmitter is the simplicity of the design, and the exposed disc rete components that can be changed and replaced to tune both freque ncy, and power consumption. A ll components were extracted and measured, with the micro-strip inducta nce measured by an Agilent E8316A 10 Mhz to 6 GHz PNA series Network Analyzer. The impedance measurement of the inductor can be seen in Figure 5-27. Figure 5-27. Microstrip I nductance Measurement for Ming TX-99 Onboard Antenna

PAGE 92

78 By placing the value of these componen ts into Agilent ADS 2003, the discrete components can be varied to reduce the minimum power requirements of the RF transmitter. A list of all values relating to the discrete components of the Ming TX-99 can be found in Table 5-7. Table 5-7. Component Values of Ming TX-99 Transmitter Component Value C1 2-7 pF C2 12 pF C3 3.3 pF L1 28.22 nH L2 1 uH R1 47 kOhms R2 100 Ohms Transistor MMBTH10 By simulating the circuit found previously in Figure 5-5 in Agilent ADS 2003, the discrete components directly correlating to output power a nd power consumption can be found. The simulation setup in Agilent ADS can be found in Figure 5-28. From these simulations, it is seen that the major com ponents that control output power and power consumption were related to the resistors found at the base and emitter of the transmitter. The emitter resistance increases the stability of the transmitter, so the component with the most effect on the output power and power consum ption is the resistor in the base. Also discovered, is the output power can be increas ed by the removal of the diode also found at the base of the transistor. Another component that had an influe nce on the power characteristics of the transmitter was the high-frequency NPN BJT transistor. By choosing a different transistor capable of high frequency operation, and had lo wer power requirements, the

PAGE 93

79 overall power of the transmitter can be further reduced. Table 5-8 shows the results of an internet search for high-frequency NPN BJT transistors. Vc Vout Tran Tran1 MaxTimeStep=10 nsec StopTime=300 usec TRANSIENT HarmonicBalance HB1 Order[1]=7 Freq[1]=300 MHz HARMONIC BALANCE DC DC1 DC PinDiodeModel NLPINM1 AllParams= Ffe= Af= Kf= Imelt= Imax= Fc= M= Vj= Cjo= Rs= Tau= Cmin= Rr= Wi= Un= Vi= Is= PinDiode NLPIN1 Mode=nonlinear Temp= Region= Area= Model=NLPINM1 R R2 R=300 Ohm R R1 R=200 kOhm V_DC SRC1 Vdc=2 V I_Probe I_Probe2 C C1 C=2 pF C C3 C=3.3 pF L L1 R= L=28.22 nH I_Probe I_Probe1 OscPort Osc1 MaxLoopGainStep= FundIndex=1 Steps=10 NumOctaves=2 Z=1.1 Ohm V= L L2 R= L=1.0 uH C C2 C=12 pF BJT_NPN BJT1 Mode=nonlinear Trise= Temp= Region= Area= Model=MMBTH10LT1 BJT_Model MMBTH10LT1 AllParams= Xti=1 Xtb=0.1 Eg=1.05 Trise= Tnom= Approxqb=yes RbModel=MDS Lateral=no Ffe= Nk= Ns= Iss= Rbnoi= Fb= Ab= Kb= Af=1 Kf=0 Tr=.1 usec Ptf=0 Itf=1 Vtf=1 Tf=.4 nsec Xtf=1.5 Fc=0.533333 Mjs=0.5 Vjs=0.75 Cjs=0 Xcjc=0.8 Mjc=0.23 Vjc=0.95 Cjc=1.65813 pF Mje=0.35 Vje=0.65 Cje=11.2827 pF Imelt= Imax= Cco= Cex= Dope= Rcm= Rcv= Rc=9.5364 Re=0.0001 Rbm=0.1 Irb=2.37229 Rb=0.1 Vbo= Gbo= Cbo= Nc=1.05387 C4= Isc=.1 fA Kc= Ke= Ikr=2.62558 Var=100 Nr=1.5 Br=0.162338 Ne=1.14992 C2= Ise=.1 fA Ikf=0.262558 Vaf=10 Nf=1.17427 Bf=75.328 Is=32.868 fA PNP=no NPN=yes Figure 5-28 Agilent ADS 2003 Simulation Environment Table 5-8. Various High-Fre quency NPN BJT Transistors NPN Transistor Ic(mA) PTOT(mW) fT Noise Figure MMBTH10 50 1250 650 MHz N/A BFS19 25 500 260 MHz N/A BFT25A 6.5 32 5 GHz 1.8 @ 1GHz BFW92A 25 375 3.2 GHz 2.5 @ 800 MHz BFS17A 25 300 3 GHz 2.5 @ 800 MHz Agilent ADS was used to find the minimal operating point for the transmitter to turn on and transmit present a carrier. The simulation setup within ADS 2003 was seen previously in Figure 5-28. To actually test the performance changes, a manual sweep of these values was performed by de-soldering and re-soldering diffe rent components onto the transmitter board, and measuring the out put power and current consumption given a 2V supply, in a similar setup as previously described. However, instead of using an

PAGE 94

80 Agilent E4448A PSA Series Spectrum An alyzer, a HP 8563E spectrum analyzer was used. Table 5-9 shows the results of this trial and error experimentation with the transmitter. Table 5-9. Performance of Various Transist ors and Resistors for Ming TX-99 Transmitter Transistor Base Resistanc e (kOhms) Emitter Resistance (Ohms) Supply Voltage Supply Current Output Power (dBm) Output Power (mW) Efficiency (%) MMBTH10 47 100 2 2.75 -3 0.5012 9.1124952 MMBTH10 47 300 2 2 -13 0.0501 1.2529681 MMBTH10 100 100 2 1.77 -5 0.3162 8.9329877 MMBTH10 100 200 2 0.911 -11 0.0794 4.35965 MMBTH10 200 100 2 0.996 -9 0.1259 6.3199067 MMBTH10 200 200 2 0.923 -11.83 0.0656 3.5544164 MMBTH10 200 300 2 0.86 -13 0.0501 2.9138793 BFT25A 47 100 2 1.73 -1.33 0.7362 21.277662 BFT25A 47 200 2 1.52 -3.33 0.4645 15.280108 BFT25A 100 100 2 1.095 -5.5 0.2818 12.869328 BFT25A 150 100 2 0.8 -8.5 0.1413 8.8283597 BFT25A 200 100 2 0.65 -10.93 0.0826 6.3541381 BFS17A 47 100 2 2.1 -2.67 0.5408 12.875103 In Table 5-9, the resistance combination which produces the highest efficiencies for each transistor are boxed in red. It is f ound that the Philips BFT25A high-frequency NPN BJT transistor served as the best transistor for use in the RF transmitter. By using the same nominal resistances from the origin al Ming Tx-99 transmitter, it was capable of increasing the efficiency by increasing th e output power, while lowering the overall power consumption for the transmission of a constant 315 MHz carrier, as compared to the original Ming TX-99 tuned to 315 MHz. FCC Part 15.231 The operating frequency of 315 MHz lies in the FCC frequency operating band of 260 MHz to 470 MHz, which is regulated by Part 15.231 of CFR47. The specific

PAGE 95

81 regulations of part 15.231 are rather unusual in that for many bands, the FCC specifies only fundamental power, harmonic levels, and allowed bandwidth, while for the frequency band of 260 to 470 MHz, the FCC regulates this spectrum based on the intended function and form of the transmitted data. Part 15.231 is broken into paragraphs A through D, while paragraph E applies only if the rules specific to paragraph A are broken. Due to the complexity and applic ation specific regulations of part15.231, the limitations of part 15.231 are best illustrated in a flowchart form as seen in Figure 5-29 taken from [23]. Interested part ies should visit the FCC website at http://wireless.fcc.gove/rules.html Figure 5-29. Flowchart for FCC Part 15-231 Requirements

PAGE 96

82 The limitations of part 15.231 section A are expressed in sections B through D, and can be found in Table 5-10, while the limitations of section E can be found in Table 5-11 and illustrated in Figure 5.30 from [24] Table 5-10. Limitations under FCC Part 15.231 (a-d) **Linear Interpolations Fundamental Frequency (MHz) Field Strength of Fundamental (uV/m) Field Strength of Spurious Emission (uV/m) 40.66 40.70 2250 225 70 130 1250 125 130 174 1250 to 3750** 125 to 375** 174 260 3750 375 260 470 3750 to 12500** 375 to 1250** Above 470 12500 1250 Table 5-11. Limitations under FCC Part 15.231 (e) **Linear Interpolations Fundamental Frequency (MHz) Field Strength of Fundamental (uV/m) Field Strength of Spurious Emission (uV/m) 40.66 40.70 1000 100 70 130 500 50 130 174 500 to 1500** 50 to 150** 174 260 1500 150 260 470 1500 to 5000** 150 to 500** Above 470 5000 500 Figure 5-30. Graphical Representation of Field Strength Limitations for Part 15.231 Section e.

PAGE 97

83 For the frequency of 315 MHz, the field st rength of the fundamental in section A can be calculated as uV/m at 3 meters = 41.6667(F) 7083. 3333, while for section E, the field strength can be calculated as uV/m at 3 meters = 16.6667(F) = 2833.3333, where F is the operating frequency in MHz. This show s that the requirements for section E of part 15.231 are lowered by almost 40 % as compared to the limitations for part 15.231 section A. Since multiple accessing schemes are currently being developed, and an ID tag will be assigned to every transmitted message, Th e RF transmitter should be designed to meet the requirements of Part 15.231 s ection a found in Table 5-10. For an operating frequency of 315 MHz, the field strength fundamental is calculated to be 6.041mV/m referenced to 3 me ters. Assuming an isotropic antenna of gain 1, the allowed transmitted power can be calculated as: 221 30TransmittedPEd Where E is the field strength and d is th e reference distance of 3m. From this equation, it can be found that the transmitted power, assuming an isotropic antenna of gain 1, is approximately 10 uW, or -19.6 dB m. Given the best case gain from the SPLATCH antenna of -8.72 dB, an output power less than -10.88 dBm, or approximately -11 dBm is required from the transmitter to meet the FCC part 15.231 requirements for radiated power. From Table 5-9, it is show n that the Ming TX-99, w ith a base resistance of 200k ohms, emitter resistance of 100 ohms, a nd the NPN BJT transistor replaced with the Philips BFT25A, fulfills the .88 dBm output power requirement with an output power equal to -10.93 dBm. Thus, by using this specific transmitter configuration, the requirements of FCC part 15.231 can be fulfilled.

PAGE 98

84 Given a receiver sensitivity of -112 dBm, receiver antenna of approximately -1 dB, transmitter antenna of -9 dB, output powe r of -10.93 dBm, and the two-ray ground reflection model, assuming the height of bot h the transmitting and receiving antenna is 1 meter, the theoretical distan ce of the wireless sensor node can be calculated to be approximately 189 meters. If we include an es timated 15 dB path loss associated with the loss due to random variables such as multi-path fading effects, this distance is lowered to approximately 79.7 meters. Central Monitoring Station The transmitters of the wireless hydrogen sensor eventually will be required to communicate to a central monitoring station. For the design of the central monitoring station which consists of a RF receiver m odule, a Data Acquisition (DAQ) device, and a laptop or desktop computer, it is assumed that power consumption is of little worry for the development of the receiving unit. The central monitoring station was originally designed for use with the Ming RE-99. Since the RE-99 is of an envelope detection topology, the RE-99 is especially sensitive to interference and noise when a carrier is absent. To remedy this problem, consider ations into software decoding should be emphasized in the design of the central monitoring station to filter and remove the effects of noise and interference. The central monitoring station currently c onsists of the Ming RE-99 receiver, and a National Instrument USB-6008 DA Q device. The Ming receiver is powered and outputs data through an analog output and input node on the USB-6008 DAQ device. The USB6008 DAQ device itself is powered via an US B port from a laptop running LabVIEW 7.1. The USB-6008 provides basic data acquisitio n functionality for the purpose of data logging, and making portable measurements, while being fairly affordable, but powerful

PAGE 99

85 enough for sophisticated measurement applicat ions. The USB-6008 consists of eight 12bit analog input channels, 12 digital I/O lines 2 analog outputs, and 1 counter, with a maximum analog sampling rate of 10 k Sample s/sec and an analog output range of 0 to +5 V which is sufficient for powering and ga thering received data from the Ming RE-99 receiver. When a carrier is absent, due to the e nvelope detection topol ogy of the Ming RE99, the receiver becomes susceptible to noise a nd interference. To m itigate the effects of the noise and interference, software coding wa s developed to distingu ish data pulses from noise and interference. The proposed soluti on is to use the maximum sampling rate of 10kHz for the DAQ, and continuously sample, taking one analog sample every 100 uS. Then these samples are buffered, and a moving average filter is used to differentiate pulses from noise and interference. An exampl e of a moving average filter can be seen in Figure 5.31. Noise Noise Pulse Noise Noise Pulse Figure 5-31. Moving Average Filter Example From Figure 5.31, it shows that the middle f our samples are the pulse, and the rest is noise. If the samples are continuously f iltered over a 4 sample window, and if the average of those 4 samples is equal to, or greater than the expect ed amplitude, then a

PAGE 100

86 pulse has been detected. This moving aver age filter was implemented in LabVIEW code, and can be seen in Figure 5-32(a). To pr ovide a graphical interface, the LabVIEW front panel user gui can be found in Figure 5-32(b). a) b) a) b) Figure 5-32. Labview Code. (a)Block Diag ram Code (b) Labview Front Panel Gui The considerations, design, and optimiza tion of a wireless communications link have been described in this chapter. A modulation scheme of OOK was selected, and a carrier frequency of 315 MHz was picked as the operating frequency of the wireless communications link. Additionally, a low-prof ile antenna was tested and matched, and the transmitter set to operate within the confines of FCC part 15.231. The development of a central monitoring station was also descri bed. To reiterate the selection of an OOK modulation scheme, OOK is attractive in that power is only consumed for the transmission of a or hi gh bit, and consumes little to no power for the transmission of a or low. By exploiting these ch aracteristics of OOK, perhaps a source coding scheme can be used to further reduce the power consumption required to send a fixed length message. A source coding scheme for minimum energy expenditure is described in the following chapter.

PAGE 101

87 CHAPTER 6 MINIMUM REDUNDANCY MINIMUM ENERGY CODING Within the system level design of the wi reless sensor, are the trade-offs involved with system optimization. Current devel opment in wireless technologies involve new coding and modulations schemes which can bring about better performance for power, bandwidth, data-rate, and error. With these new coding and modulation schemes however, come the complexities of the correl ating circuits and system topologies. The trade-off exists in that a system like th e wireless hydrogen sensor can reduce data-rate and bandwidth, reduce circuit and system complexity, and opt for a modulation scheme that consists of a more simple design for th e sake of lower power consumption. This is why OOK was selected as the modulation sc heme of choice. But why should system level power optimization just stop at the circ uit level implementation? Perhaps a coding technique can be used to s acrifice data rate, but lower the power consumption of the system, which is the most important factor in the design of a wire less hydrogen sensor node. The development of a source coding t echnique to reduce the power required to send a message is reported in this chapter. Minimum Energy Coding Currently, an On-Off Keying (OOK) modula tion approach is used for RF data transmission. OOK was selected as the m odulation technique due to the intrinsic property of OOK, where power is consumed at a or high, and no power is consumed for a . It is this intrinsic property that can be exploited to further minimize the current power consumption for data transmission. If the number of or high bits in the

PAGE 102

88 transmitted message can be reduced, then the overall power needed for the successful relay of a message will also be reduce d. Thus, if a Minimum Energy (ME) coded message has less high bits as compared to the original source message, the result is a reduction of power consumption. A ME coding scheme is reported in detail by Erin and Asada [25] where the same properties of an OOK modulation scheme are exploited for their Minimum Energy Coding scheme. The Minimum Energy Coding scheme deve loped by [25] uses a source coding method to convert a source code of a specifi ed length, to a codeword sequence of fixed length with a maximum of one high bit in th e entire codeword sequence. A minimum energy code table for the coding scheme propos ed by [25] can be seen in Table 6-1. Table 6-1. Minimum Energy Coding Sc heme Proposed by Erin and Asada Source Bits Codeword Source Bits Codeword ME(3,2) ME(3,2) ME(7,3) ME(7,3) 00 000 000 0000000 01 001 001 0000001 10 010 010 0000010 11 100 011 0000100 100 0001000 101 0010000 110 0100000 111 1000000 Although the concept behind this source coding scheme is concrete, and will definitely would lower the power required fo r the transmission of a message with fixed length, this coding scheme lowers the power consumption at an expense to codeword length. To represent all the symbols possibl e for a source message of bit of length n within a coded message of only one high bit, a coded message length of 2n 1 is required. So, to transmit a source code of length 4 bits, a coded message of length 15 is required. This adds 9 redundant bits to th e original source message, and is not an

PAGE 103

89 efficient coding scheme given an OOK m odulation scheme. Because OOK modulation consists of data transmission over a single ca rrier, the longer a single transmitter occupies the fundamental carrier, the higher the pr obability of a collision caused by other transmitters attempting to transmit at the same time. This collision would result in the corruption or total loss of transmitted data from all sensor nodes attempting to transmit at the same time. Minimum Redundancy Minimum Energy Coding A technique proposed for use with the wi reless hydrogen sensor, is one of both minimum energy, and minimum redundancy. This technique differs from previous work on this matter due to the consideration of minimizing energy w ithout sacrificing the codeword length, redundant bits, and amount of time required for a transmitter to send a single message. Previous work in ME Codi ng [25] has suffered from the concatenation of many redundant bits--for example, requiring a coded message of 32 bits in length to send a 5 bit source message. Because of the use of an OOK modulation t echnique on a single frequency carrier, as mentioned before, when two sensors are tr ying to transmit data on the channel at the same time, this collision results in a corrupti on or complete loss of data. By reducing the need for redundant bits, and lowering the messa ge length, there is a re duction in the time one transmitter is occupying the channel, and reduces the chances for collision from multiple transmissions. There are two different schemes that are currently being considering. One scheme is that of a delay based method with only one high bit per message, and another is the mapping of (n) source bits to a message with a maximum of two or three high bits per coded message.

PAGE 104

90 For the delay based method, consider a (n ) bit source message and assume that; Each message has equal probability of occurrence All bit lengths are of equal durations Power is only expended on the tran smission of a high bit. For the delay based source coding, the code d message consists of a single high bit with a specific pre-determined number of zer os appended before the high bit. These predetermined zero bits will look like a delay at the receiver si de, and since all bit lengths are equal, there can be a mapping of this delay to the original source code as seen in Table 6-1. Another source coding scheme involves the coding of a message with a pre-defined number greater than one, which translates to the maximum number of high bits in the coded message. This coding scheme assumes the same assumptions previously presented, and maps the source message to a coded messa ge with a maximum of only two or three high bits per message. For example: Consider a source message of 6 bits long. To find a coded message of minimum length (m) that can allow for 26 or 64 symbols, while using only a maximum of three high bits, solve for, 6 2264 3210 n CCCC mmmm So, the minimum (m) needed to satisfy the above equation is 7. This shows that a coded message of 7 bits lo ng is required to be able to express all 64 symbols of the source message of 6 bits long. Thus, by the addition of 1 redundant bit,

PAGE 105

91 64 symbols can be achieved while only using a maximum of three high bits as seen in Table 6-1. Table 6-2. Proposed Source Coding Technique with Comparison to Technique of Erin and Asada. Source CODED 1 high (Previous Work) CODED 1 high-delay CODED -2 high 0000 000000000000000 000000000000000 00000 0001 000000000000001 000000000000001 00001 0010 000000000000010 00000000000001 00010 0011 000000000000100 0000000000001 01000 0100 000000000001000 000000000001 00100 0101 000000000010000 00000000001 00011 0110 000000000100000 0000000001 00101 0111 000000001000000 000000001 01010 1000 000000010000000 00000001 01000 1001 000000100000000 0000001 01001 1010 000001000000000 000001 10001 1011 000010000000000 00001 10010 1100 000100000000000 0001 00110 1101 001000000000000 001 01100 1110 010000000000000 01 10100 1111 100000000000000 1 11000 If the same assumptions made in the expl anation of the delay coded scheme are considered again, the power consumption reduction of the coding technique can be calculated. Since power is only consumed on a high bit, the power reduction of the coding scheme can be calculated as: ## #|| %100avgsourcehighbitsavgcodedhighbits Reduced avgsourcehighbitsPower Now, there can be a comparison between the power consumption reductions to the number of redundant bits, so that the sour ce code can be optimized for both power consumption and data redundancy. The follo wing graph seen in Figure 6-1, shows a

PAGE 106

92 comparison of different coding schemes for di fferent source bit lengths, to see where the power reduced per additional redundant bit is maximized. Power Consumption Reduction per Additional Redundant Bit0 10 20 30 40 50 60 70 80 90 345678910Original Source Bit Length Percentage of Power Reduced per additional Redundant Bit 3 high 2 high 1 high 1 delay Figure 6-1. Power Consumption Reduction per Additional Redundant Bit Comparison for Multiple Source Coding Schemes From this graph, the maximum power redu ction per redundant b it can be found in the delay based scheme with source bit length of 3 bits. Also, it can be seen that for the case of 3 source bits, with a coded maximum of two high bits, and the case of 6 source bits with the coding scheme of a maximum of three high bits are maximum points for each coding technique. Another consideration for the source coding scheme, is the multiple accessing scheme required to reduce collisions caused by transmitters attempting to transmit on the same carrier at the same time. A stop a nd random wait interval scheme, taken from

PAGE 107

93 current RFID technologies, is being consider ed for the multiple accessing scheme. Every sensor node, will be assigned its own random nu mber. This random number will serve as both the stop and wait interval, as well as the ID of the sensor, which will be appended before or after the data bits of the transmitted message. By setting the stop and wait interval to be the random number times the maximum amount of time required to transmit a single message, this will help reduce the number of transmitters transmitting at the same time, since all stop and wait in tervals are randomly generated, and preassigned.

PAGE 108

94 CHAPTER 7 FULL SYSTEM INTEGRATION TESTING To combine all efforts made for the de velopment of a self-powered wireless hydrogen sensor, a full system integration test of the energy harvesting devices, sensor mechanism and sensor interface, microcontroll er, RF transmitter, and RF receiver was conducted at the University of Florida on October 20, 2005. The set up of the system integration, including procedure for the intr oduction of a controlled level of hydrogen into the ambient, is detailed in this chapter. Most figures in this chapter were taken from screen captures of a video made to document this experiment. Hydrogen Chamber Equipment Setup A full system integration test was successful in detecting and transmitting via a wireless link, the presence of a controlle d amount of hydrogen in the ambient, while obtaining power through scavenged ener gy using energy harvesting techniques mentioned in previous chapters. This full syst em integration and test was performed with a hydrogen source provided by the hydrogen chamber in the basement of the New Physics Building. The schematic of the hydr ogen chamber is re-drawn in Figure 7.1. The procedure for the inje ction of 500 PPM of hydrogen is as follows: There are two large gas tanks where, one tank is 500 PPM of compressed hydrogen, and other is 99.99% compressed nitr ogen. By using different flow rates for hydrogen and nitrogen, different concentr ations of hydrogen in nitrogen can be achieved. This is how 10 PPM, 100 PPM, and 200 PPM of hydrogen in nitrogen were tested. The furnace seen in the figure is used to run tests at high temperatures. For the purposes of this experiment, the Pt ZnO Na no-Rods remained at room temperature.

PAGE 109

95 The Pt-ZnO Nano-Rod was is attached to the glass support beam using two metal leads which can be accessed from outside the hydrogen chambe r for connection to the sensor interface via alligator style clips. The Pt-ZnO is first measured with air in the ambient by the HP4156B Semiconductor Parameter Analyzer. Sample is loaded into hydrogen chamber, and opening is sealed tight by 3 screws which clamps a rubber o-ring at the base of the glass support beam to the opening of the hydrogen chamber. Mechanical pump is turned on, and Valv e 1 and Valve 2 are opened. Valve 3 is slowly opened to vacuum the chamber until pressure reaches less than 0 atm, or .010 torr. Valve 3 is closed, and Nitrogen is pumped at maximum flow rate until pressure gauge reaches 760 torr, or 1 atm. Once pre ssure has reached 1 atm, or 760 torr, cut nitrogen flow Set hydrogen flow to maximu m flow rate, and flow H2 into system. Open valve 5 so that there can be a constant flow of hydrogen, and pressure within hydrogen chamber remains at 1 atm or 760 torr. Figure 7-1. Redrawn Schematic with Higher Detail of Hydrogen Chamber Components

PAGE 110

96 Full System Integration Test As mentioned in the procedure, the Zn O Nano-Rod sensor was connected to the sensor interface via alligator clips. The out put of the sensor interface was connected via alligator clips and a BNC cab le to the ADC input of th e microcontroller. The RF transmitter is connected to the USART output of the microcontroller as seen in Figure 72. Figure 7-2. Microcontroller with RF Transm itter Attached to Microcontroller USART PORT Power supply nodes were tied together via BNC cables, and were attached to the output of the energy harvesting devices. A dditionally, to see that the Pt-ZnO Nano-Rod was indeed changing to the introduction of hydrogen, and to see that the ADC of the micro controller was capable of detecting th ese changes, the microcontroller was set to run in level monitoring mode, wh ere there is a constant transm ission of data. Since this mode necessitates for a higher power requirement due the increase in duty cycle for data transmission, as well as more high bits pe r message as compared to threshold detection, if the energy scavenging techniques are succ essful for this mode of operation, then both

PAGE 111

97 modes of operation can operate through the use of energy harvesting. On the central monitoring / RF receiver side, the receiver was powered by an analog output node of the USB-6008 DAQ device connected to the USB por t of a Hewlett Packard laptop running LabVIEW 7.1. The Output of the Ming RE99 Receiver was connected to a Tektronix TDS210 Two Channel Digital Real Time Osc illoscope to clearly see the data pattern from the ADC of the microcontroller receiv ed by the Ming RE-99 receiver, showing the Pt-ZnO Nano-Rods reaction to the intr oduction of 500 PPM of hydrogen into the ambient. Figure 7-3 shows a data bit-stream pattern received by the Ming RE-99 receiver during the full system integration test. Figure 7-3. Output Data Bit-St ream Pattern of Received Da ta From Ming RE-99 During System Integration Testing This same setup was used with the sola r energy harvesting circ uitry, and vibration harvesting circuitry separately, and was su ccessful in powering the system for each scenario. The solar and vibra tion harvesting circuitry were de tailed earlier in chapter 2, and can be seen in the experimental set up in Figures 7-4 and 7-5, respectively. For the solar harvesting technique, since th e basement of the New Physics Building provides little light in the ambient for solar harvesting, in the case of the solar power reclamation system, a lamp was used to pr ovide ambient light for the solar cells and

PAGE 112

98 power IC. For the case of the vibration ener gy reclamation circuit, a shaker tuned to the resonant frequency of the four PSI D 220-A4-203YB Double Quick Mounted Y-Pole Bender PZT devices connected to a direct char ging circuit, was used as the vibration source for the vibration harvesting circuitry. Figure 7-4. Solar Cells (left) with Solar Power IC (right) a)b) a)b) Figure 7-5. Vibration Energy Harvesting Co mponents (a)PSI PZT Beams (b) Direct Charging Circuit A block diagram of the system level inte gration testing with corresponding screen caps from the video made to document th is test can be found in Figure 7-6.

PAGE 113

99 Figure 7-6. System Integration Testing Bloc k Diagram with Screen Caps From Video

PAGE 114

100 CHAPTER 8 CONCLUSION AND FUTURE WORK Conclusion In conclusion, a fully self-powered wireless hydrogen sensor was successfully designed and tested at the Univer sity of Florida. Three contri butions made to this project, and detailed in this thesis, ar e the design of a low-power sens or interface for the Pt-ZnO Nano-Rod hydrogen sensing mechanism, the design and optimization of the RF communications platform for the Self-Pow ered Wireless Hydrogen Sensor and Central Monitoring Station, and a proposed mi nimum redundancy, minimum energy coding scheme to both provide multiple accessing sc hemes for future deployment of a network of self-powered wireless hydr ogen sensors, and to minimize energy without expense to transmitted message length to lower possibility of data transmission collisions, and lower required energy to transmit a given message. For the case of the differen tial detection circuit, as a comparison to current technologies, is MAXIM-ICs new for 2006, MAX4208/4209 instrumentation amplifier package claiming to be the worlds best instru mentation amplifier. A comparison of this amplifier to the instrumentation amplifier of the differential detection circuit designed for this project can be seen in Table 8.1. Table 8-1. Comparison of Performance Between Designed Differential Detection Circuit, and Other Commercially Available Designs. PART VMIN VMAX ISUPPLY VOS This design 1 V 5.5 V 27uA 30uV MAX4208/4209 2.85 V 5.5 V 750uA 15uV

PAGE 115

101 It is seen that for almost simila r performance, the MAX4208 and MAX4209 require a higher supply voltage, and mu ch higher supply current to power the instrumentation amplifier. Future Work Although a fully self-powered wireless hydr ogen sensor was successfully designed and tested, the work doesnt end there. Currently, a single modul e, integrating all components (energy harvesting, microcontro ller, sensing mechanism and sensor interface, and RF transmitter) of the original self-powered wireless hydrogen sensor, is currently being designed and tested. Additionally, this new module will contain the minimum redundancy, minimum energy coding scheme and multiple accessing schemes as detailed in a previous chapter, so that multiple sensor nodes can be realized. Also, rather than using the Linx technologies ANT-315-SP surface mount SPLATCH antenna, and a separate MING Tx -99 transmitter board, an onboard microstrip antenna is being currently designed, and a transmitter similar to the MING TX-99 has already been designed and integrated on to the single board module. The PCB trace of the newly designed single module board can be seen in Figure 8-1, with the assembled board shown in Figure 8-2. The assembled board seen in Figure 8-2, has a SMA connector, which will eventually be repla ced by a LINX low profile SPLATCH antenna, or the antenna currently being designed. Additionally, new operational amplifiers were made available after the full system integration test, which require even lower power without cost to performance. These operational amplifiers are currently being tested and fitted into the differential detection circuit design to replace the MAX4289. One of these operation amplifiers, the

PAGE 116

102 MAX4039 operational amplifie r outputs a buffered 1.232 V reference voltage, which can be used to power the resistive bridge, a nd isolate the bridge from the supply voltage, making it even more insusceptible to any voltage shifts of the supply power. The MAX4039 itself, also only requires a mi nimum of 1.8 supply voltage, and 2uA supply current to achieve an internal 200uV input offset voltage. Another operational amplifier currently be ing considered is the MAX991X line. These operational amplifiers al so exhibit similar performanc e of input offset voltages on the order of 200uV, with a low bias requirement of 1.8 V, and 4uA supply per amplifier. Additionally, these opera tional amplifiers have an availa ble shutdown/ idle state which drains a mere 1 nA during shutdown state. W ith this amplifier, even more power can be saved by strobing this amplifier so that it will only be turned on when the collection of data by the ADC is needed. For the receiver side, LabVIEW code is also currently being re-coded for the replacement of the Ming RE-99 with the Linx Technologies RXM-315LR. Eventually, another full system de monstration, including MEMS scale PZT vibration energy harvesting devices currently being fabricated, w ill need to be done. TopBottom TopBottom Figure 8-1. Protel PCB Top and Bottom Layout For Fully Integrated Board

PAGE 117

103 Figure 8-2. Fully Assembled Single Module

PAGE 118

104 LIST OF REFERENCES 1. H. T. Wang, B. S. Kang, F. Ren, L. C. Tien, P. W. Sadik, D. P. Norton, S. J. Pearton, J. Jun and J. Lin, Hydrogen Sensi ng at Room Temperature with Pt-coated ZnO Thin Films and Nano-Rods. Appl. Physics. Letters,. vol 87,no. 222106, Nov 28, 2005. 2. H. T. Wang, B. S. Kang, F. Ren, L. C. Tien, P. W. Sadik, D. P. Norton, S. J. Pearton, and J. Lin, Hydrogen-Selective Se nsing at Room Temperature with ZnO Nanorods, Appl. Physics. Letters. vol. 86,no. 243503, Jun 13, 2005 3. Xu, Shengwen, Power for Wireless H ydrogen Sensor Network, November 2005 review: NASA, Cocoa Beach, FL, Nov. 2005. 4. X. Shengwen, T. Ngo, K.-D., T. Nishida, C. Gyo-Bum, and A. Sharma. Converter and Controller for Micro-Power Energy Harvesting,. Applied Power Electronics Conference and Exposition, 2005. APEC 2005. Twentieth Annual IEEE Vol. 1 (2005), pp. 226-230 Vol. 1. 5. X. Shengwen, T. Ngo, K.-D., T. Nishida, C. Gyo-Bum, and A. Sharma. Low Frequency Pulsed Resonant Converter for Energy, unpublished University of Florida, Gainesville, FL 2005 6. Xu, Shengwen, Power for Wireless H ydrogen Sensor Network, November 2005 review: NASA, Cocoa Beach, FL, Nov. 2005. 7. A. Phipps, Direct Charging Component s, November 2005 review: NASA, Cocoa Beach, FL, 2005. 8. S. Guinta, Considerations In Designing Single Supply, Low-Power Systems Part I: Designs Using Ac Line Power. Analog Dialogue [online], vol. 29 no. 3, Oct. 2004 [accessed Jun 10, 2005 ], http://www.analog.com/library/analogD ialogue/archives/29-3/consider.html 9. Maxim-IC, Making Low Power and Low Voltages Work in Analog Design Application Note 747. 10. C. Falcon, Calculating Data Rate a nd Data Range for Wireless Medical Applications. Medical Electronics Manufacturing [online], Fall 2004 [accessed Feb. 14, 2005],http://www.devicelink.com/mem/archive/04/10/007.html

PAGE 119

105 11. Wikipedia.org, Free Space Loss San Diego, CA, March, 23 2006 [accessed Mar. 30 2006], http://en.wikipedia.org/wiki/Free_space_loss 12. Connect802.com An On-Line Calculator For Tower Height, Link Budget, and Wireless Network Design Suitability Analysis. [online], San Ramon, CA.: Connect802.com, Mar, 2005,[accessed Mar. 14, 2005], http://www.connect802.com/antenna_c_main.php 13. Y. Wei Chapter 17: Ra dio Propagation Models, ns(Network Simulator)-2 Notes and Documentation. UC Berkeley, CA, March 06, 2005. 14. FCC, FCC Homepage[online] April 1, 2005 [accessed Mar. 30 2006], http://www.fcc.gov 15. Radio Frequency Circuits and Syst ems, class notes for EEL6374, Dept. of Electrical and Computer Engineering, University of Florida, Fall 2005. 16. American Radio Relay League, Newington, CT, RF Radiation and Electromagnetic Field Safety, 1996, [accessed Jun. 30 2005] http://www.arrl.org/new s/rfsafety/hbkrf.html 17. Wikipedia.org, Wheatstone Bridge San Diego, CA, March 23 2006, [accessed Mar. 30 2006], http://en.wikipedia.org/ wiki/Wheatstone_bridge 18. T. Kuphaldt, Bellingham, WA, Lessons In Electric Circuits Vol III, Chapter 8: The Instrumentation Amplifier, March 06, 2006 [accessed Mar. 30 2006], http://www.ibiblio.org/obp/electric Circuits/Semi/SEMI_8.html#xtocid182869 19. Wikipedia.org, Instrumentation Amplif ier San Diego, CA, March, 23 2006 [accessed Mar 30, 2006], http://en.wikipedia.org/wik i/Instrumentation_amplifier 20. D. Johnson, Design of a Controller System for a Self-Powered Wireless Hydrogen Sensor. MS thesis, Dept. of Electrical and Computer Engineering, University of Florida, Gainesville, FL, May 2006 21. En-Yi Lin, Berkeley Wireless Research Center, UC Berkeley, CA, Why OOK?, Nov 11, 2005. 22. J. Anthes, RF Monolithics, OOK, ASK and FSK Modulation in the Presence of an Interfering Signal. Dallas, TX July 27 2004. 23. Linx-Technologies, Considerations for Operation Within the 260-470 MHz Band. Application NoteAN-00125. 24. Radiometrix Ltd, Using Radiometrix Modules Operating on 418 or 433.92 MHz in North America Under FCC Regulation Part 15, section 231, paragraph (3), Harrow Middlesex, England, December 20, 2003 [accessed Feb 13, 2006], http://www.radiometrix.co.uk/apps/apnt102.htm

PAGE 120

106 25. Y. Prakash, S. Gupta, Energy Effi cient Source Coding and Modulation for Wireless Applications. IEEE Wireless Communications and Networking Conference, 2003. WCNC 2003 Volume: 1, 16-20 pp. 212-217 New Orleans LA, March 2003. 26. Jerry Jun, B. Chou, J Lin, X Shengwen, A Phipps, K Ngo, D Johnson, A Kasyap, T Nishida, H. T. Wang, B. S. Kang, F. Ren, L. C. Tien, P. W. Sadik, D. P. Norton, L. F. Voss, S. J. Pearton, Low-Power Dete ction of Hydrogen Leakage Using a SelfPowered Wireless Hydrogen Sensor Node presented at AIChe 2006 Spring National Meeting, Orlando, FL, April 24, 2006. 27. Maxim-IC, Datasheet: MAX4289 1.0V Micropower, SOT23, Operational Amplifier, Sunnyvale, CA, Feb 14, 2002 [accessed July 10, 2005], http://www.maxim-ic.com/quick_view2.cfm/qv_pk/3254 28. Maxim-IC, Datasheet: MAX406 Single, Dual, Quad, 1.2uA Max, Single-Supply Op Amps, Sunnyvale, CA, January 17, 2004 [accessed July 10, 2005], http://www.maxim-ic.com/quick_view2.cfm/qv_pk/1468 29. Maxim-IC, Datasheet: MAX478 17 A Max, Dual Quad, Single-Supply, Precision Op Amps, Sunnyvale, CA, June 20, 2005 [accessed July 10, 2005], http://www.maxim-ic.com/quick_view2.cfm/qv_pk/1110 30. Texas Instruments ,Datasheet: INA 321 MicroPower Single-Supply CMOS Instrumentation Amplifier, Dallas, TX, Nov 30, 2004 [accessed July 10, 2005], http://focus.ti.com/docs/pr od/folders/print/ina321.html 31. Texas Instruments, Datasheet: OPA 336 Single-Supply, MicroPower CMOS Operational Amplifiers MicroAmplifie r(TM) Series, Dallas, TX, Oct 27, 2004 [accessed July 10, 2005], http://focus.ti.com/docs/pr od/folders/print/opa336.html 32. Texas Instruments, Datasheet: TLV2401 Single MicroPower, RRIO Operational Amplifier with wide supply voltage ra nge and high CMRR, Dallas, TX, Jun 5, 2004 [accessed July 10, 2005], http://focus.ti.com/docs/pr od/folders/print/tlv2401.html 33. Texas Instruments, Datasheet: M SP430F1232IPW 16-bit Ultra-Low-Power Microcontroller, 8kB Flash, 256B RAM, 10 bit ADC, 1 USART, Dallas, TX, Oct 10, 2004 [accessed Feb 3, 2006], http://focus.ti.com/docs/prod/folders/print/msp430f1232.html 34. Linx-Technologies Datasheet: Linx TXM-315-LR LR Series Transmitter, Grant Pass, OR, Aug 31, 2005 [accessed Nov 14, 2005], http://www.linxtechnologies.com/index.php? section=products&category=rf_modul es&subcategory=lr_series

PAGE 121

107 35. Maxim-IC, Datasheet: MAX1472 300M Hz-to-450MHz Low-Power, CrystalBased ASK Transmitter, Sunnyvale, OR July 5, 2005 [accessed Aug. 4, 2005], http://www.maxim-ic.com/quick_view2.cfm/qv_pk/3820 36. Maxim-IC, Datasheet: MAX1479 300MHz to 450MHz Low-Power, CrystalBased +10dBm ASK/FSK Transmitter, Sunnyvale, OR, July 7, 2005 [accessed Aug. 4, 2005], http://www.maxim-ic.com/quick_view2.cfm/qv_pk/4420 37. Rayming Corp. Datasheet: Ming TX-99 300 MHz AM Transmitter, San Jose, CA, January 29, 2001 [accessed May 4, 2005], http://roboflag.carleton.ca/resource s/technical/datasheets/ming_tx_tx99v3.pdf 38. Atmel, Datasheet: Atmel U2741B UHF Remote Control Transmitter, San Jose, CA, March 6, 2005 [accessed Aug. 4, 2005], http://www.atmel.com/dyn/products/product_card.asp?part_id=2410 39. Rayming Corp., Datasheet: Ming RE-99 300 MHz AM RF Receiver Board, San Jose, CA, January 29, 2001 [accessed May 4, 2005], http://www.rentron.com/Micro-Bot/re99v3a.pdf 40. Linx-Technologies, Datasheet: Linx RX M-315-LR LR Series Receiver, Grant Pass, OR, Aug 30, 2005 [accessed Nov 14, 2005], http://www.linxtechnologies.com/index.php? section=products&category=rf_modul es&subcategory=lr_series 41. Linx-Technologies, Datasheet: Linx ANT -315-SP SPLATCH Series Antenna, Grant Pass, OR, July 28, 2005 [accessed Feb 10, 2006], http://www.linxtechnologies.com/index.php? section=products&category=antennas &subcategory=embeddable&series=sp_series 42. Fairchild Semiconductor, Datasheet: MMBTH10 NPN RF Transistor, South Portland, ME, Sept 10, 1998 [accessed Aug 9, 2005], http://www.fairchildsemi.com/pf/MM/MMBTH10.html 43. Philips Semiconductors, Datasheet: BFS19 NPN Medium Frequency Transistor, Koninklijke Philips Electroni cs N.V., Jan 5, 2004 [accessed Feb 12, 2006], http://www.semiconductors.philips.com/pip/BFS19.html 44. Philips Semiconductors, Datasheet: BFT25A NPN 5 GHz Wideband Transistor, Koninklijke Philips Electr onics N.V., Jul 6, 2004 [accessed Feb 12, 2006], http://www.semiconductors.philips.com/pip/BFT25A.html 45. Vishay Semiconductors, Datasheet: BFW92A Silicon NPN Planar RF Transistor, Malvern, PA, Apr. 29, 2005 [accessed Feb 12, 2006], http://www.vishay.com/transist ors-rf-af/lis t/product-85041/

PAGE 122

108 46. Philips Semiconductors, Datashee t: BFS17A NPN 3 GHz Wideband Transistor, Koninklijke Philips Electroni cs N.V., September 1, 1995 [accessed Feb 12, 2006], http://www.semiconductors.philips.com/pip/BFS17A.html 47. FCC, FCC Rules and Regulations [online], Columbia, MD, Feb 5, 2005 [accessed Jan. 25, 2006], http://wireless.fcc.gov/rules.html 48. Microindustrie Inc, Transmitted Power for FCC Part 15 radio systems, San Diego, CA, Dec 19, 2002 [accessed Jan. 25, 2006], http://www.microindustrie.com/fcc/trans.htm 49. H. Kinley Conversions for RF exposure measurements, Chicago, IL, Mar 1, 1998 [accessed Jan. 25, 2006], http://mrtmag.com/mag/radio_c onversions_rf_exposure/index.html 50. National Instruments, Datasheet: NI US B-6008 12-Bit, 10 kS/s Multifunction Data Acquisition for USB, Austin, TX, July 2005 [accessed Oct. 25, 2005], http://sine.ni.com/nips/cds/view/p/lang/en/nid/14604 51. H. Asada, K. Siu, C. Erin, Massachus etts Institute of Technology, Minimum Energy Coding with Application to RF Transmission. Phase 2 Progress Report: Home Automation and Healthcare C onsortium, Boston MA, October 1, 1998. 52. Maxim-IC, Datasheet: MAX4208 World s Best Instrumentation Amplifier Sunnyvale, CA, Feb 2006 53. Maxim-IC, Datasheet: MAX4039 Low IBIAS, +1.4V/800nA, Rail-to-Rail Op Amps with +1.2V Buffered Reference, Sunnyvale, CA, Mar 3, 2006 [accessed Feb. 25, 2006], http://www.maxim-ic.com/quick_view2.cfm/qv_pk/4194 54. Maxim-IC, Datasheet: MAX9910 200kHz, 4 A, Rail-to-Rail I/O Op Amps with Shutdown, Sunnyvale, CA, Nov 11, 2005 [accessed Feb. 25, 2006], http://www.maxim-ic.com/quick_view2.cfm/qv_pk/4911

PAGE 123

109 BIOGRAPHICAL SKETCH Jerry Chun-Pai Jun finished his undergraduate coursework in elec trical engineering at the University of Florida, and entered gr aduate school in electri cal engineering at the University of Florida in Fall of 2004. Jerry is interested in the desi gn and applications of wireless sensors, specifically in the field of RFID systems, implantable bio-sensors, and ambient monitoring sensors. As a masters graduate student, Jerry worked under Dr. Jenshan Lin as a research assistant in the RF System on Chip (RFSO C) laboratory at the University of Florida. After graduation, Jerry will be working at Motorola in Plantation, FL, and hopes to continue his career in RF/microwave technologies.


Permanent Link: http://ufdc.ufl.edu/UFE0014378/00001

Material Information

Title: Design and Testing of a Self-Powered Wireless Hydrogen Sensing Platform
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0014378:00001

Permanent Link: http://ufdc.ufl.edu/UFE0014378/00001

Material Information

Title: Design and Testing of a Self-Powered Wireless Hydrogen Sensing Platform
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0014378:00001


This item has the following downloads:


Full Text












DESIGN AND TESTING OF A SELF-POWERED WIRELESS HYDROGEN
SENSING PLATFORM















By

JERRY CHUN-PAI JUN


A THESIS PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
MASTER OF SCIENCE

UNIVERSITY OF FLORIDA


2006

































Copyright 2006

by

Jerry Chun-Pai Jun















ACKNOWLEDGMENTS

This thesis could not have been completed without the help of many people. First

and foremost, I would like to thank my mentor, Dr. Jenshan Lin, for giving me the

opportunity to work on this research project, and for providing me with guidance, and a

RA position for the past two years.

I would also like to thank Dr. Khai Ngo, and Dr. Toshikazu Nishida, for taking the

time to meet with the group every week to discuss, brainstorm, and evaluate the status of

this project. These weekly discussions were very helpful in keeping the project on track,

and for coming up with new ideas.

I would like to thank all those who have worked with me on this project. This

includes Bruce Chou, Alex Phipps, Anurag Kasyap, Shengwen Xu, Hung-Ta Wang, Li-

Chia Tien, and last, but not least, David Johnson. Additionally, I would like to thank my

colleagues, Tien-yu Chang, Lance Covert, Ashok Verma, Xiuge Yang, Hyeopgoo Yeo,

Yanming Xiao, Changzhi Li, Sang Won Ko, Jae Shin Kim, and Jaeseok Kim, for putting

up with me, supporting my research, and creating a comfortable and productive working

environment.

Finally, I would like to thank my parents for their endless love and support.
















TABLE OF CONTENTS

page

A C K N O W L E D G M E N T S ........................................................................ .....................3

LIST OF TABLES .............. ................. ........... ................ .......... vii

LIST OF FIGURES ......... ......................... ...... ........ ............ ix

A B S T R A C T .............................................. ..........................................x iii

CHAPTER

1 IN TR OD U CTION ............................................... .. ......................... ..

M o tiv atio n ....................................................... ................. .
C o n trib u tio n s ................................................................................................................2
Thesis O organization .................. .............................. .. ....... ... ........ ..

2 WIRELESS SENSOR SYSTEM LIMITATIONS............................. ...4

Energy H arvesting and Reclam ation ........................................ ........................ 4
Solar E energy H arvesting ....................................... .......................................5
V ibrational E energy H arvesting ......................................................... ..................6
Design Limitations of Low-Power and Low-Voltage Discrete Components ...........8
D ynam ic R ange .............................................. ........................ 9
Input O ffset V oltage .................. .............................................. 10
Least Significant Bit (LSB ) ............................................................ .... ............ 11
L im stations of a W wireless System ................................................................... ..... 12
Wireless Channel Estimation Techniques ....................................................12
Free space path loss ................................................. .. ........ ....13
Tw o-ray ground reflection m odel ..................................... ............... ..14
Shadowing model .......................................... ..........................15
F C C P art 15 R regulations ........................................................ ..................... 17
The Federal Communications Commission...............................................18
FCC rules, regulations, and safety .................................... ............... 19

3 SENSOR INTERFACE DEVELOPMENT .................................... ...............22

ZnO N ano-R ods .................. .............. ................. ........... ............ 23
ZnO Nano-Rod Fabrication Process......................................... ............... 24









Performance of ZnO Nano-Rods....... ..................... ...............25
U ncoated ZnO nano-rods ........................................ ......... ............... 27
Pd coated ZnO nano-rods............................ ........................ ...... ......... 28
Pt coated ZnO nano-rods................................ ...................................29
Detection Interface for Hydrogen Sensitive Devices ...........................................30
W heatstone R esistive B ridge...................................... .......................... ......... 31
D ifferential D election Interface................................................. .................. 34
D difference am plifier ......................................................... ............. 34
Instrumentation amplifier............................................... 37
Realization and Testing of Differential Detection Circuit............... .................. 38
Selection of operational amplifier ......................................................39
Simulation of differential detection circuit ...............................................39
Fabrication of differential detection circuit...............................................43

4 MICROCONTROLLER DEVELOPMENT ................................... .................46

M icrocontroller Selection ................................................................ .....................46
M odes of O operation ............... .......... .. .................................................. 47
Power Requirements of Microcontroller ......... ...................................48

5 LOW-POWER WIRELESS COMMUNICATION LINK .............. ....................50

Selection of a M odulation Technique ....................... ..............................................50
Selection of O operating Frequency ........................................ ......................... 53
Selection and Performance of a RF Transmitter........................................................55
Rayming Corporation TX-99 300 MHz AM Transmitter/RE-99 Receiver
P air ............ ................ ............................. ...........................5 5
Ming Tx-99 transmitter ....................................................56
Ming RE-99 receiver ......... ................ ................... 59
M ing distance m easurem ents ........................................... ............... .... 60
Linx Technologies LR series Transmitter and Receiver ................................63
Linx Technologies TXM-315-LR.....................................................64
Linx Technologies RXM-315-LR.....................................................66
W wireless Link O ptim ization............................................... ............................. 68
M ing TX -99 Pow er A nalysis......................................... .......................... 68
L ow P rofile A ntenn a ........................................ ............................................7 1
RF Transm itter O ptim ization ........................................ ......................... 77
F C C P art 15.23 1 .............................. .............. ............................ 80
Central M monitoring Station........................................................... .................. 84

6 MINIMUM REDUNDANCY MINIMUM ENERGY CODING ..............................87

M inim um E energy C oding ................................................................ ..................... 87
Minimum Redundancy Minimum Energy Coding ................. ............................89

7 FULL SYSTEM INTEGRATION TESTING...................................94

Hydrogen Cham ber Equipm ent Setup.............................................. .................. 94


v









F ull Sy stem Integration T est........................................ ..........................................96

8 CONCLUSION AND FUTURE WORK .................................100

C o n c lu sio n ........................................................................................................... 1 0 0
Future W ork ............................................................................. 101

L IST O F R E FE R E N C E S ......................................................................... ................... 104

BIOGRAPHICAL SKETCH ............................................................. ..................109
















LIST OF TABLES


Table p

2-1. Typical Values for Path Loss Exponent ...................................... ............... 16

2-2. Typical Values of Shadowing Deviation GdB ...................................... ........ 16

3-1. Various Commercial Operational Amplifiers.................................. ............... 39

3-2. Differential Detection Circuit Component Values....... ......... ..............................41

3-3. Initial Measurements of Differential Detection Interface ..................................44

4-1. Features of Texas Instruments' MSP430F1232IPW...............................................47

5-1. Comparison of Available Commercial ASK/OOK Transmitters............. ...............55

5-2. Perform ance of M ing TX -99 ......................................... .......... .............................59

5-3. Maximum Transmission Distances with Varying Antenna Locations....................62

5-4. Performance of LINX TXM-315-LR .......... ....................... ...................65

5-5. A ntenna G ain M easurem ents.......................................................... ............... 73

5-6. Gain Measurements for Matched Antenna.............................................. ...........76

5-7. Component Values of Ming TX-99 Transmitter...................................................78

5-8. Various High-Frequency NPN BJT Transistors .....................................................79

5-9. Performance of Various Transistors and Resistors for Ming TX-99 Transmitter......80

5-10. Limitations under FCC Part 15.231 (a-d) **Linear Interpolations.....................82

5-11. Limitations under FCC Part 15.231 (e) **Linear Interpolations ............................82

6-1. Minimum Energy Coding Scheme Proposed by Erin and Asada .............................88

6-2. Proposed Source Coding Technique with Comparison to Technique of Erin and
A sa d a ............................................................................ 9 1









8-1. Comparison of Performance Between Designed Differential Detection Circuit,
and Other Comm ercially Available D esigns ........................................................ 100
















LIST OF FIGURES


Figure pge

1-1. System Level Block Diagram of Self-Powered Wireless Hydrogen Sensor ............2

2-1. System Level Implementation of Energy Harvesting Devices.................................

2-2. IXOLAR XOD 17-04B Solar Cell ........................................................... ............... 6

2-3. Pulse Resonant Power Converter. ............................... ......................6

2-4. V ibration Energy D evices. ................................................ ............................... 7

2-5. Performance of Vibration Energy Harvesting Devices .............. ..........................8

2-6. Dynamic Range ofu741 Operational Amplifier ........................................ ...............9

2-7. The Shrinking LSB LSB for Multiple Voltage Supply Spans ..............................11

2-8. FCC O organizational Chart........................ .. .................... ................. ............... 18

2-9. Proposed IEEE RF Safety Guidelines ............................................. ............... 21

3-1. ZnO N ano-Rods...................................... .......................... ...............25

3-2. Close-Up of Packaged ZnO Nano-Rod Sensor .....................................................25

3-3. Simple Schematic of Hydrogen Chamber Used for ZnO Nano-Rod Testing ............26

3-4. Schematic of Biasing for ZnO Nano-Rod Hydrogen Sensitivity Testing ..................26

3-5. Uncoated ZnO Nano-Rod Relative Resistance Change for Various Hydrogen
C concentrations ................................................................... ......... 27

3-6. Pd-coated ZnO Nano-Rod Relative Resistance Change for Various Hydrogen
C concentrations ................................................................... ......... 28

3-7. Pd-coated ZnO Nano-Rod. Absolute Resistance Change for Various Hydrogen
C concentrations ................................................................... ......... 29

3-8. Pt-coated ZnO Nano-Rod Relative Resistance Change for Various Hydrogen
C concentrations ................................................................... ......... 30









3-9. W heatstone Resistive Bridge ................................................... ... ............32

3-10. D difference A m plifier ........................................................................ .................. 35

3-11. Difference Amplifier with Non-Inverting Buffer to Differential Inputs ..................36

3-12. Instrumentation Amplifier .................. ....... ......... ............................... 36

3-13. Full Schematic for Differential Detection Circuit..............................................38

3-14 New Pt-coated ZnO Nano-Rod Grown and Packaged for Differential Detection
C ircu it ......... .... ...... ........... ........... ........................................... 4 0

3-15. Agilent ADS 2003 Simulation Setup for Differential Detection Circuit ................41

3-16. Agilent ADS 2003 Simulation: Output Voltage to Swept ZnO Nano-Rod
Resistance ........................ .............................42

3-17. Protel PCB Top and Bottom Layout ............ ................................................ 43

3-18. Fabricated and Assembled Differential Detection Interface Board with Packaged
ZnO N ano-Rod Sensor ......... .......................... ........ .................................... 43

3-19. Measured Output Voltage vs. ZnO Nano-Rod Resistance Sweep for Fabricated
D ifferential D election Circuit........................................................ ............... 45

4-1. M microprocessor State Flow Diagrams. ............................................ ............... 48

4-2. Initialization Power Required for MSP43 OF 1232IPW.............................................49

5-1. Tradeoffs Between Performance and Architecture Complexity of t/4 DQPSK and
O O K .......................................................................5 1

5-2. Signal Constellations. ......................................... ................... ........ 52

5-3. Path Loss Attenuation(dB) with Respect to Carrier Frequency. ..............................53

5-4. Schematic of Ming TX-99 Taken from Datasheet ...................................................56

5-5. Ming TX-99 Transmitter in OOK Mode. VDD Is Replaced with Data Stream.........57

5-6. Ming TX-99 Transmitter. Red Outline Highlights Onboard Antenna .....................57

5-7. Test Setup for Output Power and Power Consumption of Transmitters ..................58

5-8. SMA Connector Soldered to Antenna Tap on Ming TX-99 Transmitter .................58

5-8. Schematic of Ming RE-99 Taken from Datasheet..................................................59









5-9. M ing RE-99 R eceiver........................................................... .. ............... 60

5-11. Floorplan of First Floor Atrium of New Engineering Building .............................61

5-12. Experimental Setup for Distance M easurements........................... .....................61

5-13. Received Power vs. Distance With Reference to Room Shape.............................62

5-14. Screen Capture of Received Power Spectrum ........... .... ............... ................... 63

5-15. System Level Architecture of LINX TXM-315-LR................... ................64

5-16. Pin-Out of TXM -315-LR Transmitter............................ .................. ....... 65

5-17. System Level Architecture for RXM-315-LR..................... ........................66

5-18. Pin-Out of RXM-315-LR Transmitter ......... .................. ............... 67

5-19. Tektronix TDS5104B Digital Phosphor Oscilloscope Screen Capture of Power
Analysis Performed for RF Transmission of One Bit.................... ..................69

5-20. Tektronix TDS5104B Digital Phosphor Oscilloscope Screen Capture of Power
Analysis Performed for RF Transmission of Multiple Bits .................................70

5-21. LINX ANT-315-SP SPLATCH Antenna From Datasheet................................. 71

5-22. Testing Board for SPLATCH Antenna. ...................................... ............... 72

5-23. S-Parameter for SPLATCH Antenna .............. .............................................74

5-24. S-Parameters of Measured (red) and Simulated (blue) in Ansoft Designer ...........75

5-25. Matched Antenna (Red) vs. Unmatched Antenna (blue) ......................................75

5-26. 5 MHz Bandwidth of Matched Antenna...... ...................... ............76

5-27. Microstrip Inductance Measurement for Ming TX-99 Onboard Antenna ..............77

5-28 Agilent ADS 2003 Simulation Environment.................................. ............... 79

5-29. Flowchart for FCC Part 15-231 Requirements.............. ...... ...............81

5-30. Graphical Representation of Field Strength Limitations for Part 15.231 Section
e .......................................................................... 82

5-31. Moving Average Filter Example .. ................................... .. ............... 85

5-32. Labview Code .................................... .................. ......... .. ............. 86









6-1. Power Consumption Reduction per Additional Redundant Bit Comparison for
M multiple Source C oding Schem es ................................................. .....................92

7-1. Redrawn Schematic with Higher Detail of Hydrogen Chamber Components ..........95

7-2. Microcontroller with RF Transmitter Attached to Microcontroller USART PORT..96

7-3. Output Data Bit-Stream Pattern of Received Data From Ming RE-99 During
System Integration Testing............................................. .............................. 97

7-4. Solar Cells (left) with Solar Power IC (right)................................. ............... 98

7-5. Vibration Energy Harvesting Components ............ ..........................................98

7-6. System Integration Testing Block Diagram with Screen Caps From Video..............99

8-1. Protel PCB Top and Bottom Layout For Fully Integrated Board ..........................102

8-2. Fully A ssem bled Single M odule ........................................ ......................... 103















Abstract of Thesis Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Master of Science

DESIGN AND TESTING OF A SELF-POWERED WIRELESS HYDROGEN
SENSING PLATFORM

By

Jerry Chun-Pai Jun

May 2006

Chair: Jenshan Lin
Major Department: Electrical and Computer Engineering

Within the ongoing interdisciplinary hydrogen research at the University of

Florida, a self-powered wireless hydrogen sensor node has been designed and developed.

By using multi-source energy harvesting circuitry designed and developed at the

University of Florida, scavenged or "reclaimed" energy from light emitting and

vibrational sources serve as the source of power for commercial low power

microcontrollers, amplifiers, and RF transmitters. After system power up, the sensor

node is capable of conditioning and deciphering the output of hydrogen sensitive ZnO

nano-rods sensors also developed at the University of Florida. Upon the detection of a

discernible amount of hydrogen, the system will "wake" from an idle state to create a

wireless data communication link to relay the detection of hydrogen to a central

monitoring station. The system's sensitivity is on the order of parts per million, and

hydrogen concentrations starting as low as 10 PPM can be detected.









The thesis will discuss the performance of the self-powered wireless hydrogen

sensor node, and also focus on the design and optimization of the detection circuitry,

digital processing considerations, modulation scheme and wireless communications link

to maintain an accurate and reliable system, while expending a minimal amount of energy

scavenged from ambient light or vibration for very long lifetime operation.














CHAPTER 1
INTRODUCTION

Motivation

Self-powered wireless sensors themselves are becoming a popular topic for

implementation in future systems. The idea of inexpensive sensor devices with very

long-lifetime operation involving minimal maintenance, is poised to make a very strong

impact on the engineering community. Because these devices can be deployed in very

harsh and dangerous environs, environments can be monitored remotely in real-time to

update the end-user of ambient conditions, and report any slight deviations from safe and

normal operating conditions. This can help prevent endangerment to the quality of life for

surrounding bodies, without placing expensive machinery and humans in harm's way.

As fossil fuel supplies are depleted, alternative fuel supplies such as hydrogen

which can be quickly replenished, are swiftly growing in popularity, and with a self-

powered wireless hydrogen sensing platform capable of sensitivity on orders of parts per

million (PPM), the applications of these sensors can include monitoring of hydrogen

powered machines, combustion gas detection in spacecrafts, and solid oxide fuel cells

with proton-exchange membranes [1,2].

With a self-powered wireless hydrogen sensor, and the current advances in system-

on-chip technology and MEMS processes, an integrated circuit capable of harvesting

energy supplies through ambient conditions such as lighting and vibration, can be

realized in a physically small package to control and report the dangers involved in the

development of hydrogen as a viable fuel source for the future. The overall block










diagram of the self-powered wireless hydrogen sensing platform can be found in Figure

1-1.


f-----------------------------------
PIEZOELECTRIC ENERGY ENE
RECLAMATION EERG
CIRCUIT STORAGE
SOLAR CELL

ENERGY HARVESTING CIRCUITRY
-,- _---


Figure 1-1. System Level Block Diagram of Self-Powered Wireless Hydrogen Sensor

Contributions

The objective of this research is to develop a self-powered wireless hydrogen

sensor node powered through energy harvesting techniques and capable of PPM

hydrogen sensitivity to report and transmit data, via a low-power wireless

communications link, to a central monitoring station. The specific goals of the research

provided in this dissertation, are to:

* Present limitations involved in the design of a low-power wireless sensing platform
given current restrictions placed by government requirements, and those
restrictions involved in using commercial off-the shelf analog components.

* Develop a low-power sensor interface to convert the reaction of the hydrogen
sensitive mechanism into a conditioned signal which can be accurately represented
in digital form, while consuming minimal scavenged energy.

* Develop, test, and optimize a wireless communications link for communication
between the self-powered wireless sensor node and a central monitoring station,
while meeting the regulations imposed by governmental agencies.









* Develop a source coding scheme to minimize the required power to transmit a
single message.

* Perform a full system integration and test system with a live hydrogen source


Thesis Organization

In chapter 2, the limitations involved in the design of a self-powered wireless

hydrogen sensor are presented. Chapter 3 introduces the hydrogen sensing mechanism,

and the design and testing of a differential detection circuit to interface the hydrogen

sensing mechanism to a digital back-end. Chapter 4 iterates the microcontroller system,

and the operation of the controller system with regard to the sensor platform, while

Chapter 5 emphasizes the design, testing, and optimization of the wireless

communications link. A minimum energy source coding scheme is discussed in chapter

6, and a full system implementation between energy harvesting devices, RF front-end,

sensing mechanism and sensor interface, and controller system are described in Chapter

7. Finally, conclusions, current work, and future work are detailed in Chapter 8.














CHAPTER 2
WIRELESS SENSOR SYSTEM LIMITATIONS

There cannot be the successful design of a fully wireless sensor platform without

the designation of limitations to the system. Once a thorough analysis of those

limitations which would hinder or stop the progress of the design is completed, design

orientated specifications that the system must meet can be set to assist in the realization

of the system.

To be a truly self-powered wireless hydrogen sensor node, a study of the potential

short-comings of the energy harvesting devices serving as the power sources for the

system must be completed. By using the available power provided from energy

harvesting techniques, the design of the sensor system must strive for both accuracy and

ultra-low power operation. With the requirement for a low-power sensor design, also

includes those problems associated with available commercial low-power and low

voltage analog and digital components. A thorough investigation into the problems and

solutions to the issues expected to arise in the design of the low-power analog and digital

blocks of the system will be required. In addition to a power analysis, an analysis of the

limitations that will make a direct impact on the design of a wireless transmitter and

receiver will also be discussed.

Energy Harvesting and Reclamation

This section reviews the solar and vibration harvesting devices and circuits

designed, fabricated, and tested at the University of Florida. The objective of these

designs was for an end product capable of extracting energy from both photovoltaic, and









piezo-electric(PZT) energy harvesting devices. This energy would then be primed by

efficient power converters for use by the hydrogen sensor, sensor interface,

microcontroller, and transmitter. Interested parties should refer to the original references

[3,4,5,6,7]. For the purposes of this dissertation, a system level implementation of the

energy harvesting devices can be found in Figure 2-1.


\ -----------------------
Solar Power Processor I SenIor
Panel (Solar) ------
Micro
Processor
^^^^ ^^^ ^^^^ ^^^ ^^^^ ^^ ----------_------------


,p ......................
P"T Pow r Pmeessor II Transmitter





Figure 2-1. System Level Implementation of Energy Harvesting Devices

Solar Energy Harvesting

Photovoltaic devices are a mature commercial product, and offer the attractive

availability of high energy density per area. They are however, limited to real-time

lighting and temperature conditions. The IXOLAR XOD17-04B Solar Cell seen in

Figure 2-2 was used as the solar harvesting device, and the energy yielded from this

device was conditioned by a Pulse-Resonant Power Converter designed at the University

of Florida for use by the electronics of the wireless sensor system.

The Pulse Resonant Power Converter, whose functional block diagram can be seen

in Figure 2-3(a) was designed to be self-powered and self-controlled for maximum power

point tracking, low switching loss, and to convert an input voltage of 0.8 1.2V to a









steady 2V output voltage to be used by the sensor, sensor interface, microcontroller, and

RF transmitter. The bare die photo of this power IC can be seen in Figure 2-3 (b).

-- -m-


I 6mm
I4I


Figure 2-2. IXOLAR XOD17-04B Solar Cell

Functional Block Diagram (Solar)
rr 1


1.5mm

b)


Figure 2-3. Pulse Resonant Power Converter. (a) Functional block diagram (b) Bare die
photo of Pulse Resonant Power Converter IC

Vibrational Energy Harvesting

For the harvesting of vibrational energy, piezo-electric devices are attractive as

sources in that no light is required, and the collection of energy is proportional to the

volume of the devices. The limiting factor however, is the magnitude and frequency of









the vibrations. As MEMS PZT devices are currently being developed, for a proof of

concept design, commercial PZT bimorph beams were purchased and used. Four PSI

D220-A4-203YB Double Quick Mounted Y-Pole Bender seen in Figure 2-4 (a), were

selected as the PZT devices, and were connected to a direct charging circuit (full-bridge

rectifier and shunt capacitance) seen in Figure 2-4 (b) constructed at UF.


IPZT

PZT
| Al |
j PZT


BAT54 Shottkey Diode
(Fairchild Semiconductor)
PSI D220-A4-203YB
Double Quick Mounted Y-Pole Bender C:
330uF Electrolytic Capacitor
Beam/PZT Specs:
1.25" x 0.25" x 0.02" Battery:
VL1220, 2V, 7.0mAh
(Panasonic)


a) b)

Figure 2-4. Vibration Energy Devices. (a) Four mounted PSI D220-A4-203YB Double
Quick Mounted Y-Pole Bender (b) Direct Charging Circuit

The vibration energy harvesting system was tested under lab conditions and used a

mechanical shaker tuned to 1 grams @ 130 Hz (the resonant frequency of the bimorph

beam), as the source of vibrations. This proof of concept design was able to deliver 250











uW of power. The efficiency and power transfer to battery versus the mechanical input

power of the PZT bi-morph beams can be seen in Figure 2-5.


Direct Charging Efficiency vs. Mechanical Input
Power to Battery vs. Mechanical Input Power Pno^r


300
? 250
_ 200
3 150
100
50
a. 0


80.00
60.00
0 40.00
LU
n 20.00
uJ


000 10000 20000 30000 40000 50000 60000 0.00 200.00 400.00 600.00
Mechanical Input Power [uW] Mechanical Input Power [uW]


a) b)


Figure 2-5. Performance of Vibration Energy Harvesting Devices (a)Power to Battery vs.
Mechanical Input Power (b) Direct Charging Efficiency vs. Mechanical Input
Power

Design Limitations of Low-Power and Low-Voltage Discrete Components

System designers working with discrete commercial integrated circuits (IC) are

moving towards single supply, low power, and low voltage designs. Such designs

are becoming more popular within design communities due to their reduction of cost,

component count, and power consumption. An innate feature of using low-voltage

single-supplies is the reduction of quiescent current, which is necessary for battery-

powered systems, or in the case of self-powered operation, scavenged energy.

Another not-so-obvious reasoning for using a single-supply system is the increase

of the reliability of the system. Due to the desire for a long-lifetime sensor node

requiring minimal maintenance, by designing a system with discrete components

operating at levels much lower than their maximum ratings, this inherently increases the

lifetime of the device. The trade-offs of designing a system with low power and low -


voltage operation however, comes the adverse affects associated with the slew rate,










bandwidth, and "head-room" problems [8]. These trade-offs must be considered in the

design of a sensor interface with the necessity to obtain an accurate real-world signal.

This following section provides insight into the careful design of a sensor interface.

Dynamic Range

As previously mentioned, "Headroom" is the associated dynamic range of a low-

voltage system. Since the system's usable resolution is dependent upon the signal to

noise ratio, the dynamic range of a system is perhaps one of the most significant trade-

offs in low-voltage single-supply design[8]. If one considered conventional u741 +/- 15

V dual supply operational amplifiers, due to the architecture of this device, the

input/output has a "fixed" head room of 2 V. This 2V denotes that the maximum

input/output swing of the u741 is between +/- 13 V--this swing is the dynamic range of

the operational amplifier and can be seen in Figure 2-6 from [8],

HEADROOM +15
+13








-13
HEADROOM -15
-2

Figure 2-6. Dynamic Range ofu741 Operational Amplifier

If one were to then translate this fixed headroom to a single-supply operational

amplifier operating from 0 to 3 V, the input/output maximum swing does not exist

because the headroom required is 4 V, and the maximum voltage span provided is only 3

V. The dynamic range of this device is absolutely unacceptable and will prove to be









disastrous in the final design. Thus, the single-supply design community is constantly

increasing the demand for rail-to-rail amplifiers with very low headroom requirements to

increase the dynamic range of the system. [8]

Input Offset Voltage

Due to the unbalances of the transistors and resistors within an operational

amplifier, it is often the case that an input offset voltage can be found between the input

terminals of an operational amplifier. Unfortunately, input offset voltages do not scale

down with supply voltages, and by reducing the power supply voltage, there may be a

disproportional shift in input offset voltages [9]. At heart, an operational amplifier is

simply a differential amplifier which will amplify the difference between two input

terminals. If one were to consider a system with high large-signal gain, along with the

signal, the output would also include the amplified error of the input offset voltage.

Conventional operational amplifiers such as the u741 usually include offset-null

trimming points. Although these points will compensate for input offset voltage in a per

case basis, the trimming potentiometers, as well as the offset voltage drift of the

operational amplifier, are plagued with their dependence to temperature. Although most

operational amplifiers are factory trimmed, a system designer should also place a heavy

emphasis on searching for operational amplifiers with low input offset voltage, and low

offset drifts. Often more times than not, the system designer will place more importance

on the selection of a device based on errors caused by drift due to temperature or time

than due to absolute magnitudes of an error. This is due to microprocessor computing

horsepower becoming fairly inexpensive and allowing system designers to easily linearly

correlate data using software such as a crude lookup-table approach to compensate for

absolute magnitudes of error. Such software techniques however, cannot compensate for







11


effects due to time and temperature, which makes the design of a good sensor system that

much more important for a device with a desired long lifetime of operation with minimal

maintenance.

Least Significant Bit (LSB)

With a wavering voltage span, is the relative precision required to acquire a signal.

As compared to a +/- 15 V system, given an ADC of 12-bits, a 3 V span would require up

to eight times of precision as compared to that of a +/- 15 V system. A comparison seen

in Figure 2-7 from [9] graphically displays the increasing precision of the LSB to varying

system voltage supplies.

+10V
System Supplies and Voltage Span

+/- 10 V System



Single 5 V System
+/- 5 V System 4- 4
+4.4V
+3.5V
+/-3 V System +2.5V
+2.2V Single 3 V System
OV OV



I I I
-2.2V


-3.5V


4.8mV I

SRelative Size of a 12-bit LSB
I I
1.7mV m1.2mV 1.1mV
One LSB _0.6mV

-10V


Figure 2-7. The Shrinking LSB LSB for Multiple Voltage Supply Spans









Limitations of a Wireless System

Once a system is successful in overcoming the requirement for low power, and has

the ability to accurately obtain the real world signal of interest, if a wireless sensor node

is unable to communicate with a central monitoring station, all efforts made in the design

of power harvesting and of signal processing are wasted. Alike to the design of an

accurate analog to digital interface, a RF wireless system suffers just as many if not more

limitations. In addition to the physical limitations of a reliable wireless front end, include

those limitations imposed by the United States Federal Communications Commission's

(FCC) Code of Federal Regulations (CFR) to further aggravate a system designer. Such

limitations will be further discussed in this section.

Wireless Channel Estimation Techniques

Due to the effects of multi-path, and signal power-loss, the accurate modeling of

the wireless channel within which a transmitted message will propagate through becomes

a fairly difficult and complex parameter to model. An understanding of the wireless

channel is essential in the build of a link budget, to estimate the data range of a wireless

system. A link budget can be expressed as [10];


PathLossdB = PTdB PRdB + GTdB + GRdB

Where (Pr) is the power of the transmitter, (Gr) and (GR) are the gains of the

transmitter and receiver antennas respectively, (PR) is the received power, or receiver

sensitivity in the case of a link budget, and Path loss is the attenuation due to the

propagation of a RF signal. Once we can verify the output power, gains of antennas, and

receiver sensitivity, we can use an appropriate path loss model to extract a theoretical

estimate of our distance. Unfortunately, an accurate path loss model is difficult to









realize, however, there are certain techniques one can use to estimate and calculate a

theoretical data range for our wireless system.

Free space path loss

Free-space path loss is the attenuation of the electromagnetic waves of a radio

signal traveling from a transmitter to receiver. [11] Free-space loss is usually the first

unknown parameter to be estimated, and alike to its moniker, free-space path loss is the

attenuation based strictly on the ideal propagation condition for the spherical expansion

of the RF wave front [12] through "free-space" where there is only one clear line-of-sight

path between the transmitter and receiver, and does not include those losses incorporated

with reflections from objects, diffraction, refraction, absorption and any other variables.

This of course results in a path loss with significantly less attenuation that can be found in

real-world practice. [12]

The principle behind the free-space path loss is-- signal attenuation is proportional

to both the square of the distance as well as the square of the frequency while taking into

account the expanding spherical wave-front of the electromagnetic signal. As a

electromagnetic signal travels from a transmitter to a receiver, the signal spreads in all

three-dimensions creating an expanding spherical surface. A sphere with radius R has a

surface area of:


SurfaceAreaSphere = 4. T (R)2

If R, the radius is doubled, then the surface area is increased by a factor of four.

This relationship is known as the "Inverse Square Law". Given these details, we can

derive that the free-space path loss of a wireless channel is:










FreeSpacePathLoss 4. -= R 2



Where A = c is the speed of light (3 x 108), fc is the fundamental frequency of
fc

interest, and R is the distance between the receiver and transmitter.


Using this path loss estimate, H. T. Friis presented his channel model; the Friis

Free Space Path Loss Model which is used to calculate the receiver signal power (PR)

with respect to antenna gain of the transmitter (Gr) and receiver (GR ) antennas,

attenuation due to path loss, and transmitted power (Pr). The Friis Free Space Equation

is [13]:


PrT GR 2
PR(R) =
(4 -T- R)2


The Free-Scale Path loss model makes the assumptions of the most ideal of

conditions for the transmission of data. Unfortunately, in the real world, these

assumptions are rarely, if ever true. Thus, more sophisticated modeling is required.

Two-ray ground reflection model

A more common approach for the estimation of propagation signal attenuation is

the plane earth propagation model which models the average attenuation associated with

the distance between a stationary transmitter and receiver with direct line-of-sight, while

taking into account the ground reflection path. This method is considered to be more

accurate of a model and can be used to roughly estimate the attenuation for fundamental








frequencies within the ultra-high-frequency bands between 200 MHz and 5 GHz. [10]

The path loss for the two-ray ground reflection model can be seen in the following

equation, where (R) is the distance between the transmitter and receiver, and (ht) and (h,)

are the heights of the transmitter and receiver from the ground, respectively.

R4
TwoRayGroundLoss =
ht2 hr2

From this path loss equation, it is assumed that R is much greater than the heights

of either antenna. From this path loss equation, we determine the power received (PR)

given the transmitter power (PT), the heights (ht,h,) and gains (GT,GR) of the transmitter

and receivers respectively.

SPr .Gr.GGR. ht2 hr2
PR(R) =
R4

Up to now, both models of propagation have been using the dependence of distance

of the propagation path to estimate the received power for the wireless system. This

technique of estimation represents the communication range as that of a sphere. We have

yet to consider the effects due to random multi-paths, which is a more credible model for

channel estimation.

Shadowing model

Both free-space and two-ray ground models neglect the reality that the power

received at a given receiver is more of a random variable due to multiple path effects, or

fading effects. [13] The use of a shadowing model extends the ideal sphere of the

previous methods, where the predicted received power is more of a "mean" of received









signals, and creates a richer more statistical model based on the environment or "terrain"

of the propagation path. The shadowing model consists mainly of two parts. One part,

like previous modeling techniques, calculates a mean path loss for the system within the

path of propagation, while the second part reflects variations of power at certain distances

[13]. Table 2-1 and Table2-2 show typical values for channel estimation using the

shadowing model.

Table 2-1. Typical Values for Path Loss Exponent 1
Environment P
Outdoor Free Space 2
Outdoor
Shadowed Urban Area 2.7 to 5
In Building Line-of-Sight 1.6 to 1.8
In Building
Obstructed 4 to 6


Table 2-2. Typical Values of Shadowing Deviation adB
Environment GdB (dB)
Outdoor 4 to 12
Office, hard partition 7
Office, soft partition 9.6
Factory, line-of-sight 3 to 6
Factor, obstructed 6.8


In the path loss portion of the shadowing model, the mean received power (PR) at

distance (R) is found by referencing (R) to the power received at a distance closer in,

(Ro). The mean received power for a given distance R is computed through the following

equation where 0 is a typical path loss exponent found in Table 2-1.


PR(Ro) R P

PR(R) Ro)


This can also be expressed in terms of dB.









PR(R) R
-(R) dB = -10 p -log R
PR(Ro) _Ro)

For the variations of received power at certain distances, it can be modeled as a

log-normal random variable, or a Gaussian random variable XdB with a mean of zero, and

a standard deviation of GdB. Thus, the shadowing model can be fully represented by:


SPR(R) R
--- d = -10- (log R + XdB
PR(Ro) Ro)

The shadowing model gives a more authentic estimation of a wireless channel.

However, a common radio practice for interior wireless channel estimation is to use any

path loss models presented in previous sections, and to assume an addition 15 to 20 dB of

fade margin in the link budget calculation. This margin should account for multi-path

phenomena, shadows, reflections, system losses, and other divergences from an ideal

system model. [10]

FCC Part 15 Regulations
Within the United States of America, there are strict impingements placed on the

use of the radio spectrum between 9 kHz up to 3 THz. The responsibility of regulating

the radio spectrum falls in the hands of the FCC who administrates the spectrum band for

non-federal governmental use, and the National Telecommunications and Information

Administration (NTIA), a unit under the Department of Commerce, who regulates the

spectrum for use by the federal government. Since this design, is for non-Federal

Government use, only the details of restrictions placed from the FCC will be assessed in

detail. Interested parties should visit www.fcc.gov for more information.








18



The Federal Communications Commission


The FCC, established by the Communications Act of 1934, is an independent


government agency whose duties and responsibilities are directly decided by Congress.


The FCC consists of five commissioners appointed by the president and confirmed by the


senate, and are slated for five year terms except when filling an unexpired term. One


commissioner is appointed chairperson by the president, and only a maximum of three


commissioners may be allowed to be of the same political party [14].


Offie of Office of
Engineering General

Elromnogdir on(l iDrItlIDiI AdImnAinrhrie LowDr.
LborNPax ~I Lillgn Drr
h atknodAlogy hL
Amhdlmad aroll


Offie of Oace of
Admintrlotfe SIrteg Planning
aw Judges & FPolly Aay sis


Office of
Managing
Direcar
Hmoa IR-oure Momgeml
Ini.nmtbn etdAoflgi Crte
FinndIIl0para onn
AdrrnnlshollLt perarions
PrlormanmtErd. &l Rmsordam
ft 5rmwAy jn


OfHie of
Media
IhelianM
bdr IsnwesSka
Inhmndui S ,iruioAl
u-Vlisi Sernul Slan


Ofice of Offike of
Communications Workplace
Busies Diversity
pporinites


WIrele Enforcement WiNls Me
Tdecmmmmhsnkal
jm ipelftt ki Bureau T elecan w igi ltd m l Lure sU
Bureau urre Bu
A;... LMin iii il0 ofHMnlMrnat Rer &eRinci LManamlen ltresu Strm of AIgmI &Psourr-l5f
l.'Tol.T '.u Of Maiid nfre ldmitary ADInaI tSisaainum kesCIr DiliRnofon &dustrylrId.
I',, r.F.:I.I .D' Telnmmunianionsa nsulelrrDOln Sm freomgmrlMtI ourc! F eiydi T
T4e-. -,r c A.l I Il., Sped crumn Enl~rsn ntDi TeThmlogesDir Itrdiry AmlalisDIv
Il.. ,rI, ...., I T4...J .i D Markt Dputs Rei inr1nD Spednm & f[nsmiwiin PdolinbT Englimlng DI
In.ieigliums& He..ingsis PubhDSfDtry I ffdill BoImdma Ltm ftlqly
hglIoml & Fielfla O frilial M fmln urni Di AudieDIl .
Mobiwtlv DI video Li
Enadahnd MIY


(onsuner &
b"ro me ntl,
Affirs Bueau
MmUniMpTt Mi.
I..r,. ...r.. ..i-,, f ..,, fllH-
flliqy 01
Dibiillas Rilight Ike
lanmner faltsn Oureaih Cl.
RIferaenoe forlom Conrer


Office of
legislative
Atairs



Intemeilonal
Boeau
Monargemrsn &ainiprteiSruiiff
rJ.i, rn..
'....|.,,. riB
1 rphal|tr;BNh 1 ]1 .C,


Figure 2-8. FCC Organizational Chart


The FCC is maintained by a staff organized by function, and has jurisdiction over


the entire 50 states, the District of Columbia, as well as all U.S. territorial possessions.


Broken up into six operating Bureaus, and ten Staff Offices, the Bureaus' are responsible


for processing license applications, analyzing complaints, conducting investigations,


developing and implementing regulatory programs, and taking part in hearings while the


, I









Offices provide support services. An organizational chart of the different operating

offices and bureaus of the FCC can be seen in Figure 2-8.

FCC rules, regulations, and safety

The rules and regulations of devices capable of emitting radio waves within the

radio spectrum are located in title 47 of the code of federal regulations (CFR). Although

the Office of Engineering and Technology (OET) maintains and is responsible for parts 2,

5, 15, and 18 of title 47, the official rules are published and maintained in the Federal

Register. [15] The OET is an office within the FCC and is responsible for allocation of

the radio spectrum for public, non government use, and provides advice on technical and

policy issues governed by title 47 of the CFR. The OET is also responsible for the

maintenance of the Table of Frequency Allocations. The Rules of title 47 of the CFR are

divided into part 0 through 101, and organized into four sub chapters. Rules applicable to

the design of a sensor are outlined as follows:

* Part 2: Frequency Allocations and radio Treaty Matters; general rules and
regulations

* Part 5: Experimental Radio Service(other than broadcast)

* Part 15: Radio Frequency Devices

* Part 18: Industrial, Scientific, and Medical equipment

The regulations set by the FCC to police the spectrum and mitigate unfair

interference between intentional and unintentional radiating devices (such as computer

monitors), ultimately limits the radiated power from radio systems. Thus, these

regulations directly affect both the RF transmitter and any gains associated with antennas.

Given the premise mentioned earlier where distance is proportional to the output power

of the transmitter/antenna pair, the FCC creates further difficulties in the design of a









wireless front end. In addition to those rules and restrictions set forth by the FCC,

additional insight into the human-electrical interaction serves as the basis for another

regulatory standard.

As required by the National Environmental Policy Act of 1969, the effect on the

quality of the human environment from the emissions of transmitters regulated by the

FCC needs to be evaluated by the FCC. Although there are currently no federally

mandated standards for RF exposure, several non-governmental organizations such as the

American National Standards Institute (ANSI), the Institute of Electrical and Electronics

Engineers, Inc. (IEEE) and the National Council on Radiation Protection and

Measurements (NCRP) have recommended limitations for human and RF

electromagnetic field exposure. Several of these recommended limitations are [15]

* Limitations set for the span of 3 kHz to 300 GHz.

* Controlled environments (where energy levels can be accurately determined and
every person on premise is aware of the presence of EM fields) allow for higher
power than that of uncontrolled environments ( where energy levels are unknown
and where personnel on facilities may be unaware of presence of EM fields)

* Lowest E-field exposure limits occur at frequencies between 30 and 300 MHz.( 1
mW/cm2 (61.4 V/m) controlled, 0.2 mW/cm2 (27.5 V/m) uncontrolled)

* Lowest H-field exposure limits occur at 100-300 MHz. (1 mW/cm2 (0.163 A/m)-
controlled, 0.2 mW/cm2 (0.0728 A/m) uncontrolled

* Above 100 MHz, safety limits for E and H fields remain the same.

* Below 100 MHz, E-field radiation has lower power density limits than do H-field.

The reasoning behind setting more stringent limitation on power densities within

the frequency band of 30 MHz to 300 MHz, is that the natural resonant frequencies of the

human body occur between 30 to 300 MHz, and at frequencies above and below, the











human body should absorb less energy. [16] A more graphic representation of IEEE RF

Safety Guidelines can be seen in Figure 2-9.




.....I IIII...w..L. E..rC T,

"11 1111 | I 1 II i | I, I I






l-l



103
10 10-2 10 10o 101 102 103 104 105 106
FREQUENCY, MHZ

Figure 2-9. Proposed IEEE RF Safety Guidelines

The guidelines set by the FCC and the safety standards recommended by IEEE,

along with the uncertainty of the wireless channel between the transmitter and receiver,

show that the design of a reliable wireless front end is not a menial task; instead, it

requires an intensive selection of proper components to optimize an already limited

system.














CHAPTER 3
SENSOR INTERFACE DEVELOPMENT

The proper development of the hydrogen sensing mechanism and the analog to

digital interface for this device is essential to the success of the design of a robust

hydrogen sensing wireless sensor. Extensive efforts must be made to fully understand the

hydrogen sensing mechanism before a proper design can be made for the interface

between the sensing mechanism and an Analog to Digital converter (ADC). The

hydrogen sensing mechanisms used were ZnO nano-rods developed and fabricated at the

University of Florida. These nano-sensors proved themselves to be a robust solution for

the missing role of a hydrogen sensing mechanism.

The nano-rods were interfaced to an analog to digital converter with careful

considerations as previously described in Chapter 1. After careful planning, a differential

detection circuit consisting of an instrumentation amplifier topology is used to ease in the

transition from a real world signal to a digitized representation. This topology showed

strong immunity to error sources as detailed from before, and served as a successful

interface between the reactions due to hydrogen on the ZnO nano-rod and the analog-to-

digital converter of the digital realm.

The background, analysis, design, and implementation of the ZnO nano-rod and

instrumentation amplifier are detailed within this chapter.









ZnO Nano-Rods

The Zinc-Oxide (ZnO) nano-rods developed at the University of Florida between

the collaboration of students and faculty of the Department of Materials Science and

Engineering and the Department of Chemical Engineering, were used as the hydrogen

sensing mechanisms of the system. These lightweight hydrogen sensors were designed

with the goal of achieving high sensitivity, rapid response to stimuli, reversibility, and

low power consumption, all within a physically small and light package.

The unique characteristics of ZnO nano-tubes and nanorods make them

fundamentally appropriate candidates for the sensing of hydrogen. ZnO is a material

currently used in the detection of humidity, UV light, and gas, and has shown to change

resistance with respect to both temperature and hydrogen exposure. Because of its wide

bandgap of 3.2eV, the ease of synthesizing nanostructures, the availability of

heterostructures, and the bio-safe characteristics of this material, ZnO is a most attractive

material for the specific sensing application at hand [1,2]. With ZnO nano-rods placed in

an array, as a gas sensor, they are able to create a large chemically sensitive surface-to-

volume ratio which is needed for high sensitivity in hydrogen sensing. ZnO nano-rods

can also be produced cheaply, and are highly compatible with other microelectronic

devices.

To enhance the device's sensitivity to hydrogen, catalytic coatings or dopings of

platinum (Pt) or palladium (Pd) were used to further increase the ZnO nano-rods

hydrogen sensing mechanisms [1,2].









ZnO Nano-Rod Fabrication Process

The processes of growing and packaging the ZnO Nano-Rods for use as the

hydrogen sensing mechanisms in the sensor system are presented. Interested parties

should read [1,2].

ZnO Nano-Rod site selective growth was accomplished by nucleating the Nano-

Rods on discontinuous Au islands of nominal thickness 20 A coated with a substrate. It

was previously shown that synthesis of ZnO Nano-Rods on wide assortment of substrates

is a fairly uncomplicated task, increasing the ease of Nano-rod synthesis [1] Deposition

of ZnO Nano-Rods were achieved using molecular-beam epitaxy with a base pressure of

5 10-8 mbar using high purity (99.999%) Zn metal and 03/02 plasma discharge as the

source chemical. Growth time was approximately 2 h at 600 oC. Resultant ZnO Nano-

Rods grew to a typical length of 2 15 um, with diameters in the range of 30-150 nm. A

schematic of a multiple ZnO Nano-Rod Sensor can be seen Figure 3-1 (a), while Figure

3-1(b) shows a scanning electron micrograph of the home-grown ZnO Nano-Rods.

Selected area diffraction patterns showed the ZnO Nano-Rods to be of single crystalline

form.

Additional coatings of Pt or Pd were deposited by sputtering in some cases,

forming Pd thin films coatings of approximate thickness of 80 A or Pt thin film coating of

approximate thickness of 10 A. Electrodes on multiple ZnO Nano-Rods were created by

using a shadow mask to pattern sputtered Al/Ti/Au electrodes contacting both ends of

multiple ZnO Nano-Rods on A1203 substrates. Electrodes were separated through a

spacing of approximately 400 um. Au wires were then bonded to the contact pads of the

package for transportation and current-voltage measurements. The packaged hydrogen

sensing mechanism can be seen in Figure 3-2.





























Figure 3-1. ZnO Nano-Rods. (a) Schematic of Multiple ZnO Nano-Rodss (b) Scanning
Electron Micrograph of ZnO Nano-Rods.





















Figure 3-2. Close-Up of Packaged ZnO Nano-Rod Sensor

Performance of ZnO Nano-Rods

The fabricated and packaged ZnO Nano-Rods were tested under laboratory

conditions where their reactions to hydrogen can be monitored under a controlled

hydrogen environment. This controlled environment is within a hydrogen chamber


Al .0
Svl D
A4ZnO M-NRMs
ZnO M-NRs~^ l^









located at the University of Florida, and allows for an accurate and quick assessment on

the performance of the fabricated ZnO Nano-Rod sensors. Figure 3-3 shows the

schematic of the hydrogen chamber, and a more detailed description on the operation of

the chamber will be included in a later chapter.


Sensor

Sgas


Furnace


Turbo
pump


Figure 3-3. Simple Schematic of Hydrogen Chamber Used for ZnO Nano-Rod Testing


ZnO Nano-Rod


0.5 V


Figure 3-4. Schematic of Biasing for ZnO Nano-Rod Hydrogen Sensitivity Testing

For the purpose of this section, the only information pertinent to understanding the

performance of the ZnO Nano-Rods are the chamber's functions of injecting controllable

amounts of N2 and H2 to create different concentrations of hydrogen on the orders of

parts per million (PPM) within the chamber. All measurements were performed using a









HP 4156B Semiconductor Parameter Analyzer with the packaged ZnO Nano-Rods biased

with a 0.5 V supply on one terminal, and ground on the other as seen in the schematic of

Figure 3-4.

Uncoated ZnO nano-rods

The relative transient response of uncoated ZnO Nano-Rods as the gas ambient is

switched from pure N2 to concentrations of hydrogen in nitrogen ranging from 10 to 500

PPM can be seen in Figure 3-5.

0.008 ZnO nanorod without Pd
500ppm

0.00 250pp


nE 0.004 lOOppm


0.002 -ppm Ir
10ppm Alr
N O2 2 Air
0.000
0 30 60 90 120 150
Time(min)

Figure 3-5. Uncoated ZnO Nano-Rod Relative Resistance Change for Various
Hydrogen Concentrations

This shows a relative resistance change of approximately 0.70% for 500 PPM of H2

in N2 after 10 minutes of exposure, with inconsistent results for lower concentrations.

The gas-sensing mechanisms include the desorption of adsorbed surface hydrogen and

grain boundaries in poly-ZnO, exchange of charges between adsorbed gas species and the

ZnO surface, leading to changes in depletion depth and changes in surface or grain

boundary conduction by gas adsorption/desorption [2]. Thus, it is shown the ZnO

Nano-Rods are a suitable candidate for the sensing of hydrogen. The performance of un-









coated ZnO Nano-Rods serves as a starting point for comparison to the increased

hydrogen sensitivities of Pt or Pd coated ZnO Nano-Rods.

Pd coated ZnO nano-rods

Pd coated ZnO Nano-Rods have a relative transient response shown in Figure 3-6.

Once again, this shows the reaction of the Pd coated ZnO Nano-Rods as the gas ambient

is switched from N2 to increasing concentrations of H2 in N2 starting from 10 PPM to 500

PPM.

-0-Z-nO nanorod with Pd
20ppm 500ppm
0.04 lOOppm25


0.03 10ppm


< 0.02 -

0.01 -
N1 0O Al Air Al Air
0.00
0 30 80 90 120 150
Time(min)

Figure 3-6. Pd-coated ZnO Nano-Rod Relative Resistance Change for Various
Hydrogen Concentrations

By comparison, there is an approximate five-fold increase in response to hydrogen

as compared to that of uncoated devices. This shows that the addition of Pd appears to be

effective in catalytic dissociation of the H2 to atomic hydrogen. The relative response of

Pd-coated Nano-Rods can be seen as a function of H2 concentration in N2, where Pd-

coated ZnO Nano-Rods were capable of detecting hydrogen down to less than 10 PPM

with a relative response of greater than 2.6% at 10 PPM. They have also shown to have

greater than 4.2% relative resistance change to 500 PPM after 10 minutes of exposure.









Additionally, as seen in Figure 3-7, ZnO Nano-Rods show no response to the presence of

02 at room temperature.

680 .






8660 1 1 1 10ppm H






0 5 10 15 20 25 30
Time(min)

Figure 3-7. Pd-coated ZnO Nano-Rod. Absolute Resistance Change for Various
Hydrogen Concentrations

The reversible chemi-sorption of reactive gases at the surface of metal oxides such

as ZnO can provide a large and reversible variation in conductance of the material. Thus,

the recovery of the initial resistance upon removal from the hydrogen ambient is quick

(less than 20 seconds). The sputtered Pd-ZnO Nano-Rod also shows to have an increased

effective conductivity due to the presence of metal. Palladium has proved to be a catalyst

which can enhance the sensitivity of the hydrogen sensing mechanisms.

Pt coated ZnO nano-rods

Similar fabrication and testing procedures were done with Pt Coated ZnO Nano-

Rods, except for the replacement of Pd with Pt as the sputtered metal. Figure 3-8 shows

the relative transient response of the Pt-ZnO Nano-Rods as the ambient gas is switched
from N2 to various concentrations of 2 in N2 hanging m 10 to 500 PPM.
i0, 0- 250ppm H2
-o- 500ppm H2
845
0 5 10 15 20 25 30
Time(min)

Figure 3-7. Pd-coated ZnO Nano-Rod. Absolute Resistance Change for Various
Hydrogen Concentrations

The reversible chemi-sorption of reactive gases at the surface of metal oxides such

as ZnO can provide a large and reversible variation in conductance of the material. Thus,

the recovery of the initial resistance upon removal from the hydrogen ambient is quick

(less than 20 seconds). The sputtered Pd-ZnO Nano-Rod also shows to have an increased

effective conductivity due to the presence of metal. Palladium has proved to be a catalyst

which can enhance the sensitivity of the hydrogen sensing mechanisms.

Pt coated ZnO nano-rods

Similar fabrication and testing procedures were done with Pt Coated ZnO Nano-

Rods, except for the replacement of Pd with Pt as the sputtered metal. Figure 3-8 shows

the relative transient response of the Pt-ZnO Nano-Rods as the ambient gas is switched

from N2 to various concentrations of H2 in N2 ranging from 10 to 500 PPM.









10 '
-- ZnO nanorods with Pt
8 500pp
8 250ppm H2
HH

H
4 lOppm
<| H2
2-


0 20 40 60 80 100 120
Time(min)

Figure 3-8. Pt-coated ZnO Nano-Rod Relative Resistance Change for Various
Hydrogen Concentrations

Similar transient characteristics can be seen between Pd-ZnO Nano-Rods and Pt-

ZnO Nano-Rods such as rapid recovery of initial resistance, with a 90% recovery within

20 seconds upon removal of the hydrogen from the ambient with the replacement of 02

or air, and no response to the presence of ambient 02 or N2 at room temperatures.

However, the change from initial resistance after 10 minutes of exposure to H2 has an

almost two fold increase as compared to Pd coated ZnO Nano-Rods. This shows that as a

hydrogen sensing mechanism, Pt coated ZnO Nano-Rods have increased sensitivity to

hydrogen, and would prove as a better sensing device than both Pd and uncoated ZnO

Nano-Rods.


Detection Interface for Hydrogen Sensitive Devices

The challenge in designing the interface between a sensor and the Analog-to-

Digital (A/D) converter of a system is found in the necessity to obtain an accurate real

world signal with the limitations of low power and reduced voltage swings. These

limitations were previously discussed in chapter 1, and will be reiterated as the design of









the differential detection circuit is described. Furthermore, for accomplishing a long-

lifetime mode of operation, this sensor system is faced with the limitations of analog

development, as well as the requirement for low-power operation.

One of the most important objectives of the sensor interface design is for the

compatibility between the sensing mechanism to the detection circuitry and ADC.

Thankfully, ZnO Nano-Rods were chosen due to their high compatibility with

microelectronic devices. The design will focus on meeting the demands of the ZnO

Nano-Rods, power requirements, and meeting those requirements set by the resolution of

the ADC input of the microcontroller. Additionally, the differential detection interface

should only detect reactions due to hydrogen, and void all changes caused by other

variables such as temperature. Given that the ZnO Nano-Rod's initial response to any

exposure of hydrogen is distinct and immediate, this intrinsic characteristic will serve as

an ally for the successful detection of hydrogen within the sensor system.

This section will detail in depth the design of a differential detection sensor

interface-from theoretical concept, to considerations of component selection, and finally

an evaluation on the performance of the fabricated and packaged differential detection

circuit.

Wheatstone Resistive Bridge

The intrinsic characteristic of ZnO Nano-Rods that makes them suitable as

hydrogen sensing mechanisms is their change in resistance with respect to how much and

how long the device has been exposed to hydrogen. To accurately detect the presence of

hydrogen in the ambient, since the nominal resistance of the ZnO Nano-Rods is a

function of hydrogen concentrations, the precise measurement of the change in resistance









of the ZnO Nano-Rods can be used to correlate the change in resistance to the

concentration of hydrogen in the ambient.

Bridge circuits, although old and primitive, continue to commonly serve as the best

solution for the measurement of resistance, capacitance, and inductance. The resistive

bridge instrument used for the measurement of an unknown electrical resistance is known

as a Wheatstone bridge. The Wheatstone resistive bridge as seen in Figure 3-9, invented

by Samuel Hunter Christie in 1833 and popularized and improved by Sir Charles

Wheatstone in 1843 [17], illustrates the concept of using a difference measurement for

obtaining the value of an unknown electrical component.




R, R3


Vs V V2

R2 R4




Figure 3-9. Wheatstone Resistive Bridge

A difference measurement is taken across Vg seen in Figure 3-9, where Vg is

dependent upon the nominal resistance values of R1, R2, R3 and R4. With R1 and R3

resistance values set and non-changing, the voltage divisor combinations of R3/R4 and

Ri/R2 respectively, form the voltage Vg across V2 and Vi. Vg can be equated as:

Vg =V2-VI So,










R4 R2
g = R3 + 4 R +R2 s Where Vs is the supply voltage


Thus, it can be seen that if R4 and R2 are of equal values, the voltage across Vg

should be 0 V, and any variation of R2 with respect to R4 will create a voltage drop across

Vg. By measuring this voltage drop, assuming the reference resistor R4 has remained

constant, we can decipher the change in resistance of R2.

Because the ZnO Nano-Rods exhibit a change to both temperature and hydrogen

concentrations, the relationship between R4 and R2 serves to be advantageous in

providing a way for making the system impervious to the variable of temperature. By

using a passivated ZnO Nano-Rod device encased in glass as the reference resistor R4 and

an un-covered and exposed ZnO Nano-Rod device as R2, additional changes in resistance

for the exposed ZnO Nano-Rod due to effects of temperature will be compensated by R4,

and so the only changes seen at Vg are those caused by the exposed ZnO Nano-Rod

reacting to the presence of hydrogen gas in the ambient.

For the mode of low power operation, by setting resistors R1 and R3 as bias

resistors much larger magnitude in comparison to that of R4 and R2, these resistors can be

used to limit the power consumed by the bridge circuit Using the relationship of Voltage

= Current Resistance, and Power = Voltage Current, using very large resistance values

for R1 and R3 should limit the current passing through the two legs of the bridge circuit,

and since power is directly proportional to current, limit the overall power consumption

of the Wheatstone Bridge.

This limiting of current however, lowers the dynamic voltage swings of V2 and Vi,

which in turn makes the output voltage Vg very small in comparison to the supply









voltage. To successfully detect the change in resistance, the voltage Vg must be

amplified to meet the resolution requirements of the ADC. This requires the addition of

an amplifier stage to buffer and amplify the signal before processing by the ADC of the

sensor system. Since the Wheatstone resistive bridge uses the concept of differential

measurement, the amplification stage must adhere to the design of a differential detection

interface.

Differential Detection Interface

The design of a differential detection interface must meet the requirements of

several issues noted in the previous section. Firstly, the design of the interface must

remain steadfast to the original concept of using differential measurements to determine

the detection of hydrogen, and lastly, the interface must have high large-signal gain to

amplify the output of the current-limited Wheatstone bridge.

Difference amplifier

To remain a differential measurement instrument, a difference amplifier as seen in

Figure 3-10 from [18] is employed. The architecture of an operational amplifier with no

feedback by itself is already a differential amplifier with an uncontrollable gain. Typical

inverting or non-inverting amplifiers are restrictive in that there is the practical loss of

one of the two inputs; however, by using a difference amplifier topology, a nominal gain

can be set, with both inputs remaining intact. By keeping both inputs intact, the internal

differential architecture of an operation amplifier can be exploited to be used as the

differential measurement inputs.

In a difference amplifier configuration, the analysis of the circuit is essentially the

same as that of an inverting amplifier. The only difference is the non-inverting input of








the operational amplifier being set to a voltage that is a fraction of V2 rather than ground,

where V2 is dependent upon the resistance values of R3 and R2.

R2 R3


V,









V,


Vout


Figure 3-10. Difference Amplifier

The output of the operational amplifier, VOUT, can be found using the following

equation:


R3
.(V2- V1)
R2


Unfortunately, as compared to a non-inverting amplifier configuration, the input to

the difference amplifier has fairly low impedance. Because V2 and Vi of the Wheatstone

resistive bridge is to be connected to the differencing inputs of the interface, if the inputs

of the interface are of low input impedance, this will pose a serious problem to the

accuracy of Wheatstone bridge due to the presence of a separate path for current to cause

an inaccurate representation of output voltage at the output nodes of the Wheatstone


VOUT =









bridge. Fortunately there is a simple solution. All that is needed is a non -inverting

buffer, or a voltage follower to be added as seen in Figure 3-11.


OUT


Figure 3-11. Difference Amplifier with Non-Inverting Buffer to Differential Inputs

An improvement to a simple voltage follower can be seen in Figure 3-12. This

topology is known as an instrumentation amplifier.


OUT


Figure 3-12. Instrumentation Amplifier









Instrumentation amplifier

Alike to the addition of a voltage follower as seen in Figure 3-11, Figure 3-12

shows an improved version of Figure 3-11 through the addition of three resistors

connecting the two input buffer circuits to the difference amplifier. Using this topology,

there can be the establishment of a gain stage before the large-signal gain of the

difference amplifier, while maintaining a high input impedance to isolate the Wheatstone

resistive bridge from the feedback resistor network of the difference amplifier.

With the introduction of the three resistors connecting the high impedance input

buffers to the difference amplifier, additional gain can be provided before the gain of the

difference amplifier. This gain is achieved by creating a voltage drop across Rg from the

isolated input voltages at Vi and V2. This voltage drop induces a current through Rg, and

since the feedback loops of the input buffers draws little to no current, the same amount

of current is drawn through the two resistors labeled R1. This produces a voltage drop

across nodes 3 and 4 equal to:

2+-R
V3-4=(V2-Vl)- 1+ 2-R
Rg

With the combination of the input buffer stages, three connecting resistors, and

difference amplifier, the total large-signal gain of the instrumentation amplifier is found

to be equal to:

2-Ri R3
VouT = (V2-VI). 1+ )R-
Rg R2

The schematic for the full differential detection circuit can be seen in Figure 3-13.

Here we see the Wheatstone resistive bridge serving as the input to the instrumentation









amplifier, with the inclusion of the exposed ZnO Nano-Rod, passivated ZnO Nano-Rod,

Wheatstone current limiting bias resistors, and the feedback network of resistors, Rg, R1,

R2, and R3 of the instrumentation amplifier. At this point, the design of the differential

detection interface is completed, and the selection of the proper components for the

fabrication of the differential detection interface can start.

VDD


Figure 3-13. Full Schematic for Differential Detection Circuit


Realization and Testing of Differential Detection Circuit

Thorough considerations from the limitations of commercial discrete components

noted in Chapter 2 remained in mind for the selection of components. Instrumentation

amplifiers intrinsically include very low DC offset, low drift, low noise, high open-loop

gain, high common-mode rejection, and high input impedances, making them highly

accurate, and stable circuits for long and short term use [19]. From the discussions of









chapter 2, heavy emphasis should be put on the selection of an operational amplifier with

requirements of low supply voltage, low supply current, low DC input offset, low drift,

high Common-mode rejection, and rail to rail voltage swing capabilities.

Selection of operational amplifier

A thorough internet search of available low power operational amplifiers was

conducted, and the results of this search can be summarized in Table 3-1.

Table 3-1. Various Commercial Operational Amplifiers
NAME Vsuppy(IN) supply(uA) VOS(MAX)(uV) VOS(TYP)(uV) Rail to Rail
MAX4289 1 9 2000 200 Output
MAX406 2.5 1.2 500 250 Output
MAX478 2 17 70 30 no
INA321 2.7 40 500 200 Output
OPA336 2.3 20 125 60 Output
TLV2401 2.5 0.88 1200 390 Input
Output

From Table 3-1, Maxim IC's MAX4289 was chosen as the operational amplifier of

choice. At the time, the MAX4289 showed to have the best input offset as compared to

power consumption. The most attractive part of the MAX4289, is the minimum voltage

range it was rated for. Out of all the other operational amplifiers studied, none were able

to operate from a single-supply voltage as low as 1 V, while draining only 9uA and

maintaining an input offset voltage of 200 uV.

Simulation of differential detection circuit

The assembly of all components involved both the components required for the

instrumentation amplifier, as well as the resistive bridge. For use as the exposed ZnO

Nano-Rods, new Pt-Coated ZnO Nano-Rods were grown and packaged for use. The

nominal resistance change to the injection of 500 PPM of H2 in N2 into the ambient can

be seen in Figure 3-14.










ZnO with increase Pt catalyst

1580 -
1560 -
E 1540
o 1520
1500
S1480
S1460
1440
W 1420
1400



time(min)

Figure 3-14 New Pt-coated ZnO Nano-Rod Grown and Packaged for Differential
Detection Circuit

Because current testing only involves testing at room temperature, the requirement

for a passivated ZnO Nano-Rod to compensate for temperature deviations is not required.

This simplifies the design in allowing the passivated ZnO Nano-Rod to be replaced by a

resistor. In order to tune the gains of the amplifier, the system was simulated using

Agilent ADS, with the operational amplifier's parameters entered with the values seen in

Figure 3-15, which are the parameters found on the datasheet for the MAX4289.

After gain tuning, the Differential Detection circuit seen in Figure 3-13, was

simulated with the component values found in Table 3-2. In the simulation, the exposed

ZnO Nano-Rod's resistance value was swept from 1565 ohms to 1461 ohms, which

correlates to the resistance span of the Pt-ZnO Nano-Rods with 500PPM of H2 in the

ambient seen in Figure 3-14. The simulation set up of the Differential Detection circuit

including parameters of the MAX4289 can be seen in Figure 3-15. The output of the









41




instrumentation amplifier was plotted against the swept resistance span, and can be seen


in Figure 3-16.


Table 3-2. Differential Detection Circuit Component Values

PART VALUE

Rg 470 kohms

RI 2 Mohms

R2 39 kohms

R3 2 Mohms

Rbias 270 kohms

Passivated ZnO 1565 ohms

Operational Amplifier MAX4289


SRC1
- Vdc=2 0 V


Rout=100 Ohm
RDiff=50 MOhm
CDiff=0 F
RCom=1 MOhm
CCom=0 F
SlewRate=6e+3
IOS=0 5 nA
VOS= 2 mV
BW=17kHz


DC
DC1






out

VAR
VAR1
R=1565





1 PARAMETER SWEP
ParamSweep
Sweep1
SweepVar="R"
SimlnstanceName[1]="DC1"
SimlnstanceName[2]=
SimlnstanceName[3]=
SimlnstanceName[4]=
SimlnstanceN ame[5]=
SimlnstanceName[6]=
Start=1565
Stop=1461
Step=1


Figure 3-15. Agilent ADS 2003 Simulation Setup for Differential Detection Circuit









Ouput Voltage to Swept ZnO Nano-Rod Resistance
0.8-


0.7 -





0.5


0.4 -



1480 1470 1480 1400 1500 1510 1520 1530 1540 1550 1580 1570

R

Figure 3-16. Agilent ADS 2003 Simulation: Output Voltage to Swept ZnO Nano-Rod
Resistance

Given the supply voltage of 2 V, the designed output power of the power IC power

converters designed for the sensor system, and assuming a 10-bit ADC, the A/D has a

resolution of about 2mV, with 1024 voltage levels between 0 and 2V, and The output of

the interface must be able to meet this requirement and provide at least a 2mV per ohm

(output voltage to ZnO resistance change) output. Fortunately, from the graph of Figure

3-16, we can see that this requirement of resolution can be achieved, with an approximate

4mV/ohm falling slope. The simulation also reveals the effects of the input offset

voltages of the operational amplifiers. From Figure 3-16, it can be seen that for the case

where both the exposed ZnO Nano-Rod and passivated ZnO Nano-Rod is matched, there

exists an approximate 200 uV DC offset. This is unfortunate, but can be compensated









through software coding. The next step is for the design of a printed circuit board (PCB)

for use of fabricating and testing of the differential detection system.

Fabrication of differential detection circuit

The next step was to layout the schematic of the instrumentation amplifier onto

PCB, and assemble the device for testing. The designed PCB layout can be seen in

Figure 3-17. Same component list as seen in Table 3-2 were used in the assembly, and

the final assembled device with a packaged ZnO Nano-Rod device can be seen in Figure

3-18. Initial measurements for the fabricated sensor interface can be seen in Table 3-3.











Figure 3-17. Protel PCB Top and Bottom Layout

















Figure 3-18. Fabricated and Assembled Differential Detection Interface Board with
Packaged ZnO Nano-Rod Sensor










Table 3-3. Initial Measurements of Differential Detection Interface
Supply ZnO H2 Supply Sensor
Voltage Resistance PPM Current Output
2V 1565 0 42uA 84uW 30uV
2V 1522 10 44.2uA 88.4uW 152mV
2V 1500 500 44.3uA 88.6uW 210mV


Given the A/D resolution of 2mV, from Table 3-3, we can see that the sensor and

sensor interface is capable of meeting the resolution requirements of the system while

consuming minimal power, and remaining sufficiently impervious to the limiting effects

described in chapter 2.

A study was then conducted to measure the linearity of the whole assembled

differential detection device. For this experiment, the exposed ZnO Nano-Rod was

replaced by discrete chip resistors ranging from 1460 to 1562, while the passivated ZnO

Nano-Rod was replaced by a chip resistor with nominal resistance of 1562. This

procedure is similar to the simulation detailed earlier in the chapter. 1562 ohms was used

instead of 1565 due to the available value of resistors. A plot of the performance for the

linearity of the instrumentation amplifier with respect to the swept value of the exposed

ZnO Nano-rod bridge resistance is seen in Figure 3-19. From these results, it is shown

that the performance of the differential detection circuit is actually better than simulated

results.

The detection circuit shows to have good large-signal linearity with an approximate

-4mV/ohm slope, which matches the slope of the simulated system. In addition, for the

case when both exposed ZnO Nano-Rods and passivated ZnO Nano-Rods are matched,

the DC offset is only a mere 30 uV. This shows that the instrumentation amplifier is










indeed impervious to the effects of input offset voltage, and the architecture has managed

to compensate for the absolute input offsets of the MAX4289 operational amplifiers.

Software compensation of input offset voltages mentioned earlier is unnecessary for the

differential detection circuit. The next process would be to integrate the sensor interface

to that of the digital signal processing portions of the wireless hydrogen sensor node.


400

300
o -*
E200

S100
0
0
14


Output voltage vs sweep of exposed Pt-ZnO Nominal
Resistance


60


1480


1500 1520
Nominal Resistance (Ohms)


1540


1560


Figure 3-19. Measured Output Voltage vs. ZnO Nano-Rod Resistance Sweep for
Fabricated Differential Detection Circuit














CHAPTER 4
MICROCONTROLLER DEVELOPMENT

Microcontroller Selection

The proper selection of a microcontroller was essential to the success of the design.

The system required the microcontroller to include an onboard ADC with enough

resolution to track the changes of the sensors and be capable of conditioning and

processing the data received from the sensor interface. Enough onboard memory is

needed to retain both the runtime code as well as store the data sampled by the ADC.

There is also the requirement for our microcontroller to have the ability to encode and

send this data to the transmitter via a serial output. The system would be optimized if the

microcontroller included a serial output port capable of sourcing enough power to drive

and power the transmitter. There also exists the requirement for the microcontroller to be

easily reprogrammable and consume a minimal amount of power. Interested parties

should reference to [20].

Because the initial goal is for a truly self-powered system, an emphasis must be

placed on the assumption that the bulk of power will be consumed in the active states of

the microcontroller and transmitter. However, because the time required for the ZnO

Nano-rods to saturate is on the order of minutes, the system can operate with a very low

duty cycle, and so, idle or standby current also becomes a significant factor in the

decision of which microcontroller to select.

Eventually, Texas Instrument's MSP430F1232IPW was selected. This specific

microcontroller was chosen because of its many features, large searchable









knowledgebase, and the quality of assistance and samples given by TI. Table 4-1

highlights all the pertinent features of our microcontroller selection.

Table 4-1. Features of Texas Instruments' MSP430F1232IPW
Type of Program Memory Flash
Program Memory 8 kB
RAM 256 Bytes
I/O Pins 22 pins
ADC 10-bit SAR
Interface 1 Hardware SPI or UART, Timer UART
Supply Voltage Range 1.8 V 3.6 V
Active Mode 200uA @ 1 MHz, 2.2 Vsupply
Standby Mode 0.7 uA
# of Power Saving Modes 5

Modes of Operation

Currently the microcontroller is programmed to run as a state machine, and has two

different reprogrammable modes of operation. In each mode of operation, the

microcontroller operates within the following states: initialize, collect data, transmit data,

and sleep. The first mode of operation is for the level monitoring of hydrogen. This

mode runs through each state until a discernable threshold of hydrogen is detected. This

threshold is set so that although hydrogen is present, the level of hydrogen is not enough

to pose any serious danger. Once hydrogen is detected, the microcontroller forces the

RF front-end to transmit an emergency pulse to the central monitoring station, and returns

back to an idle mode.

The second mode of operation is of data transmission. In this mode, the

microcontroller collects data from the sensor interface, and queues this data to the RF

front-end to be transmitted to the central monitoring station. This mode is for a constant

tracking of hydrogen levels, while the level monitoring mode is to alert the end user that










hydrogen has indeed been detected. The state flow diagram for the Level Monitoring

Mode and Data Transmission Mode can be seen in Figure 4-1.

Initialization Initialization
LESLEEPP
SLEEP I
NO Set I
Timer Set YES
Collect
CollectData NO Analyze Threshold YE ra SmiMt

Threshold YNO
AnalyzeData Dete. ed. TransitPulse YES
Data Nt
N
(a) (b)


Figure 4-1. Microprocessor State Flow Diagrams. (a) Level Monitoring State Flow
Diagram. (b) Data Transmission State Flow Diagram

Power Requirements of Microcontroller

To analyze the power consumption of the microcontroller during various stages of

operation, a 383 ohm resistor was connected in series between the power supply, and

microcontroller. Differential probes were used to measure the voltage across the 383

ohm resistor. Calculation for the average power consumption is as follows:

* Total Area (TA) under the measured curve is calculated (units of V-sec)

* Peak power is calculated as:
VMAX
PeakPower = -- VSUPPLY
383
Where VMAX is the maximum point of the measured curve

* Average power is calculated as

TotalArea
AvgPower -= r VSUPPLY
S Duration 383t )
Where Tduration is the duration of the measured curve









A power analysis was done by David Johnson at Cisco to examine the power

requirements of the controller system. From this analysis, it is observed that the power

consumption for the microcontroller to remain idle, output data via serial power (either

high or low bit), and to scan the ADC's input, is a constant 2.5 uW. The most power

consumed by the microcontroller at any time, is in the microcontroller's initialization

state, which occurs only once during initial power up of the microcontroller. The

initialization time for the microcontroller is only for 12.5ms, where average power

consumption is 3.07mW with a peak power of 7.3mW (as seen in Figure 4-2).


Figure 4-2. Initialization Power Required for MSP430F1232IPW














CHAPTER 5
LOW-POWER WIRELESS COMMUNICATION LINK

In the design of wireless sensors, the most power consuming component is often

found in the wireless front-end. To make matters worse, components within a RF

transceiver/transmitter/receivers such as power amplifiers or oscillators, at best, only

have efficiencies slightly better than 50% [10]. This means at best performance, to

transmit 100mW of power, the device will require 200mW of power. However, the

effects of low efficiency can be mitigated through several techniques. Because typical

sensor nodes remain in idle states much longer than in active states, the sensor nodes

themselves are of very low duty-cycle. By using low duty-cycle and low data rates,

components within the transmitter for the wireless sensor can be turned off when no data

is present to be transmitted, and the entire transceiver/transmitter/receiver can be placed

into a low-power sleep mode. The considerations for the design and implementation of

the wireless communications link are detailed in this chapter.

Selection of a Modulation Technique

From the previously mentioned reality of both the nature of the sensor system, as

well as the limitations of the components within a RF transceiver, transmitter, and

receiver, there exists a modulation technique which can take advantage of all the

mentioned limitations and requirements of the system. Firstly, the system must be able to

obtain power from scavenged sources, and allow for long life-time operation with

minimal maintenance, which reduces the complexity of the wireless front end from that

of a transceiver to that of a lone transmitter. Additionally, because the system itself is of









low duty-cycle and lower data rate, and because of the low efficiencies of RF discrete

components, a modulation scheme is required which can exploit the concept of

consuming power only when transmitting data, and requires a simple transmitter

architecture, with few discrete RF parts counts. By using a modulation scheme of low

complexity, the depth of modulation can be realized through a transmitter architecture

with fewer discrete power consuming components. Because these components

intrinsically show poor efficiency, a lower component count will further reduce the

power consumption for the RF front end.

.At1 I- -DQPSK
,:: ..... 4
,I. I.........._ T

LF 1 x






a) b)







absent" technique, also known as "On-Off Keying" (OOK). What made OOK an
OOK s w p








1only be "on" and consuming power when the RF front end was transmitting a "high" or a
r rr r r


a) b)

Figure 5-1. Tradeoffs Between Performance and Architecture Complexity of 71/4 DQPSK
and OOK. (a) BER performance (b) Architecture Complexity

The simplest modulation scheme available is that of a "carrier present, carrier

absent" technique, also known as "On-Off Keying" (OOK). What made OOK an

appealing modulation scheme was the intrinsic premise that an OOK transmitter would

only be "on" and consuming power when the RF front end was transmitting a "high" or a

"1", and that the transmitter has the advantage of going into an "idle" state so that little to

no power would be consumed on the transmission of a "low" or "O". A comparison of

the transmitter architecture and performance for 7c/4 DQPSK as a reference to OOK [21]










can be seen in Figure 5-1. It is shown that the n/4 DQPSK transmitter architecture is

significantly more complex than that of an OOK transmitter. This complexity is a trade

off between performance and complexity (correlating to power consumption) for the

selection of a modulation scheme.

The disadvantage of OOK modulation however, is found in the error caused by the

presence of unwanted or undesirable signals. Typically, OOK is an unappealing

modulation scheme for networks of heavy traffic, but because a sensor system rarely will

transmit with a duty cycle of more than 25% of the time, OOK modulation is suitable for

use as the modulation scheme of the wireless sensor node. OOK differs from Amplitude

Shift Keying (ASK) in that there is no carrier present during the transmission of a zero.

This allows for additional power reduction on the transmitter side, however, it allows

OOK to be more susceptible to an interfering signal making detection by the receiver

more difficult as compared to ASK. This is a performance trade-off between ASK and

OOK.

(P 2 'P 'P 2
Threhol1d Threshold



"', .. / ",, .... P ",. -... ",,... .*'' ..... .'" 1
OOK
ASK FSK




a) b) c)

Figure 5-2. Signal Constellations. (a) OOK(b) ASK(c) and as a reference, FSK

The signal constellations for ASK and OOK can be seen in Figure 5-2 from [22],

with the signal constellation for FSK as a reference for comparison. For the design of the









wireless sensor node, because power requirements out-weigh any other requirements,

OOK is selected as the choice of modulation due to the characteristics of OOK which

make it a highly power conserving form of modulation.

Selection of Operating Frequency

Once a modulation scheme is selected, the selection of which frequency band to

operate within is needed. The considerations for the selection of which frequency band

are heavily dependent on the limitations stated in Chapter 2 considering impairments of

the wireless channel, FCC regulations, and the modulation scheme derived in the

previous section. Because OOK is highly sensitive to interference signals, the selection

of the operating frequency is very important for the success of the wireless system.

Path Loss (dB) vs Frequency (Hz)

100 I I I I--
100




50

PL(f)

0 -




-50 I I I I
1.105 1.106 1.10 110 8 1.109 1.1010
fc
Frequency(Hz)

Figure 5-3. Path Loss Attenuation(dB) with Respect to Carrier Frequency.

As seen in Figure 5-3, regarding free space path loss, it can be seen that as

operating frequency is increased, for a given distance, the attenuation of the propagation

path increases. Thus, with a lower frequency, longer transmission distances can be









achieved with lower power. Since lower power is required to attain a particular distance,

less output power is needed, which in turn reduces the power consumption requirements

of the transmitter.

This however comes at a cost. From the discussion of Chapter 2, it was also

mentioned that the most stringent output powers were placed on the frequency band of 30

to 300 MHz. To reiterate, this was done to limit RF power absorption by the human

body, because the human body is naturally resonant between those frequencies. Thus, a

trade-off exists in that although limited in output power, using a lower frequency

increases the transmission distance, but decreases the required output power to transmit a

certain distance.

Another factor to consider is the traffic involved in each frequency band. Common

operating frequencies such as the 902-928 MHz band, are constantly being used with

continuous transmissions of voice, data, video, and offer high level interference from

microwave ovens and spread spectrum devices. Because an OOK modulation scheme is

highly susceptible to interference, the overcrowding of frequency bands within 900 MHz,

2.4 GHz, and 5 GHz can prove to be treacherous to a low complexity modulation scheme

such as OOK. Other lower frequency bands such as the 260 to 470 MHz bands are much

more open and less crowded. Typical frequencies within these bands such as 315 MHz,

418 MHz, and 433.92 MHz only compete with garage door/keyless entry systems, or

interference from amateur radio users [23].

Due to these factors, and the availability of commercial products, the operating

frequency of 300 MHz to 315 MHz within the 260-470 MHz FCC frequency band was

selected. This operating frequency offers a fairly interference free band, but is shown to









have some unusual restrictions specific to the frequency band of 260 to 470 MHz set by

the FCC under part 15.231 which will be detailed later on in the chapter.

Selection and Performance of a RF Transmitter

To simplify the design of the wireless sensor node, commercial RF transmitters

were chosen to be used as the RF front end of the sensor system. After selection of

modulation scheme and operating frequency, an internet search was performed to find

available transmitter/receiver pair packages capable of OOK within the 315 MHz

operating frequency band. A brief listing detailing available transmitters with their

corresponding performance specifications can be seen in Table 5-1.

Table 5-1. Comparison of Available Commercial ASK/OOK Transmitters
Output Output
MODEL VMIN VMAX ISUPPLYMAX ISUPPLYMIN wer(min) Power(max) Freq TYPE
LINX TXM- 315
LIX5M- 2.1 3.6 5.1mA 1.8mA -4 dBm 8 dBm Hz SAW
315
MAX1472 2.1 3.6 9.1mA 1.5mA 3.3 dBm 6 dBm C5rystal
MHz
315
MAX1479 2.1 3.6 6.7mA 2.9mA 2.7 dBm 5.3 dBm 3 Crystal
MHz
Ming TX-99 ? 5 1.6mA ? ? ?300 LC
MHz
Atmel 315
274mB 2 5.5 12.5mA ? 1.5 dBm 5 dBm MHz Crystal
The list found in Table 5-1 was whittled down to two choices for a RF front-end-

the TX-99 manufactured by Rayming Corporations, and the TXM-315-LR manufactured

by Linx Technologies. These specific transmitters were selected due to their low power

consumption, low component count, and low complexity for the ease of rapid prototyping

and development while meeting the requirements for a low-power OOK transmitter

operating within the 260 to 470 MHz frequency band.

Rayming Corporation TX-99 300 MHz AM Transmitter/RE-99 Receiver Pair

From a previous project, there was a Ming TX-99 transmitter/ RE-99 receiver pair

available for use for this project. The difficulties of the Ming TX-99 were that little









documentation was provided for both the company, and the device. The datasheet for the

Ming TX-99 offered very few maximum and minimum operating conditions. However,

what made up for the difficulties of the Ming TX-99 was the simplicity of the design.

Ming Tx-99 transmitter

The architecture of this transmitter is based on a colpitts oscillator design seen in

Figure 5-4 and consists of a single high frequency NPN BJT transistor and a LC tank to

tune the transmitter to oscillate at a specific frequency.





(5 .53 iT2t

[1ATA -- ------ U -----1---- ^ ----- 4
EiN







Figure 5-4. Schematic of Ming TX-99 Taken from Datasheet

Initial performance tests show that when biased at 0.6 V, the transmitter drains

850uA, which translates a power requirement of 510uW to transmit a constant 50% duty

cycle 580mV peak to peak square pulse train of 1 kHz. The LC tank included a variable

capacitance with a range of 2-7 pF to allow for the tuning of the operating frequency.

Another advantage of the Ming TX-99 transmitter was the onboard antenna. The

onboard antenna served as the inductor for the LC tank. With the Supply Voltage and

Data nodes tied together as seen in the schematic of Figure 5.5, the transmitter can be

used as an OOK transmitter. The whole transmitter module can be seen in Figure 5.6

with the onboard antenna highlighted for detail.









VDD
T











GND

Figure 5-5. Ming TX-99 Transmitter in OOK Mode. VDD Is Replaced with Data Stream











Figure 5-6. Ming TX-99 Transmitter. Red Outline Highlights Onboard Antenna

Additionally, the printed micro-strip inductor which serves as the onboard antenna

can be tapped for attachment of an external antenna. For output power measurements,

the signal pin of a SMA connector was soldered and tapped to the micro-strip inductor

where an external antenna would be attached to, with ground of the SMA connector tied

to ground of the Ming TX-99. The SMA was then attached to one end of a Ift SMA

cable (FLX402#1), and the other end to a DC blocker before finally attached to the input

of a HP 8563E 9kHz to 26.5 GHz Spectrum Analyzer. The DC blocker was used to

prevent damage to the spectrum analyzer by blocking DC current from directly entering

the input of the spectrum analyzer. Both the supply voltage and data node of the Ming









TX-99 were connected to a power supply (Agilent E3631A) set to 2 V to send a 100%

duty cycle signal (transmitter continuously on). Test setup can be seen in Figure 5-7, and

the transmitter with attached SMA connector can be seen in Figure 5-8.













Figure 5-7. Test Setup for Output Power and Power Consumption of Transmitters






Figure 5-7. Test Setup for Output Power and Power Consumption of Transmitters


a) b)


Figure 5-8. SMA Connector Soldered to Antenna Tap on Ming TX-99 Transmitter










The result was an output power of approximately -4.67 dBm while draining

1.95mA continuously from a 2V supply. To test for minimum voltage operation, for

100% duty cycle, the transmitter was capable of operating at 1.2 V while draining 290 uA

at an output power of -21.17 dBm. Assuming efficiency can be calculated as:


PEf y OutputPower
Efficiency = --
VSupply I ISupply

The data found in Table 5-2 can be tabulated.

Table 5-2. Performance ofMing TX-99
SUPPLY ISUPPLY(mA) Output Power(dBm)
1.2 0.29 -21.17
2 1.95 -4.5


U


Power(mW)
0.007638358
0.354813389


Efficiency(%)
2.194930413
9.097779211


Ming RE-99 receiver

Since a central monitoring station can be assumed to provide as much power as

needed, the power requirements are not as stringent on the receiver side, making the RE-

99, the receiver complement of the TX-99 a suitable receiving unit. Alike to the TX-99

transmitter, the RE-99 also lacks in documentation. From the schematic found in the

datasheet as seen in Figure 5-8, it is seen that the RE-99 is an envelope detection circuit

which is typical for AM/ASK/OOK receivers.


Figure 5-8. Schematic of Ming RE-99 Taken from Datasheet









Alike to the TX-99 the RE-99 provides an onboard antenna which also serves as

the LC tank resonating at 300 MHz as the input for the receiver. It differs from the TX-

99 in that it is no longer a micro strip antenna, but a loop antenna of 2 turns, and diameter

of 5mm. It also offers a tap for an external antenna, and as recommended, a quarter wave

monopole antenna was created by cutting a 22 gauge copper wire down to a length of

9.36 inches, which is approximately 14 the wavelength for 300 MHz, and soldered to the

antenna tap. Unfortunately, the sensitivity of the receiver was unable to be measured, but

instead, a distance measurement was performed. A picture of the RE-99 receiver without

the external antenna can be seen in Figure 5-9.












Figure 5-9. Ming RE-99 Receiver

Ming distance measurements

Because the sensitivity of the receiver is unknown, it was decided to perform an

experiment to find the maximum transmission distance. The experiment was conducted

in the atrium on the first floor of the New Engineering Building at the University of

Florida. The floor plan of the atrium can be seen in Figure 5-11. The setup is detailed as

follows:

* Transmitter was tied to serial output (USART) of a microcontroller (MSP430)
outputting a constant data stream. The serial output of the microprocessor was tied
to both the supply voltage and data node of the transmitter, forcing an OOK
modulation scheme, while providing power to the transmitter. The microprocessor
was powered with a 2V supply.










* The transmitter remained stationary and the receiver was attached to a cart for
mobility. The height of the transmitter and receiver were 0.45m and 0.55m
respectively. The receiver was powered via another power supply. A diagram of
this setup can be seen in Figure 5-12.

* The output of the receiver was tied to the input of a Tektronix TDS210 Two
Channel Digital Real-Time Oscilloscope

* For received power measurements, a 22 gauge copper quarter wave monopole
antenna soldered to a SMA connector was connected to an Agilent E4448A PSA
Series Spectrum Analyzer


Atrium


0 m


3.5 m


10m


Figure 5-11. Floorplan of First Floor Atrium of New Engineering Building


Receiver


Transmitter



C
3
3


Distance (m)i


20 m


Figure 5-12. Experimental Setup for Distance Measurements

Within the testing, the placement of a 22 gauge copper quarter wave monopole

antenna was placed on the transmitter, receiver, or both. The maximum transmission










distances for these test cases can be seen in Table 5-3. From this data, it shows that with

a quarter wave monopole antenna on both the transmitter and receiver, the maximum

distance for the successful detection of the original data stream serially outputted by the

microprocessor, was found to be 19.4 m.

Table 5-3. Maximum Transmission Distances with Varying Antenna Locations
Antenna Location Maximum Distance
Receiver Only 14.5 m
Transmitter Only 16.8 m
Transmitter & Receiver 19.4 m



In addition to this maximum transmission distance experiment, measurements for

received power were taken as well. These measurements were taken from an Agilent

E4448A PSA Series Spectrum Analyzer connected to a quarter wave monopole antenna.

For all measurements, the transmitter also had a matched quarter wave monopole

antenna. Figure 5-13 shows the received power versus distance, with reference to the

layout of the testing environment, the atrium of the New Engineering Building.


Distance (m)
0 5 10 15 20
-35 -
E -40 -
-45
-50
C -55
-60
-65
S-70
-75



Figure 5-13. Received Power vs. Distance With Reference to Room Shape

From this Figure, it can be seen that at around 10m, the hallways of the floor plan

began to act as a sort of waveguide. This caused the received power to increase after 10








meters was reached. From the maximum distance experiment, it is shown that the

maximum distance achieved was set to be around 19.4m. At that distance, the received

power was approximately -70 dBm. From this, it can be concluded that the sensitivity of

the receiver is approximately -70 dBm. Figure 5-14 shows the power spectrum taken

from the spectrum analyzer at im, and 8m respectively.


MarErIl"'.
300A.00000 Hll
-3 .4 I


a) b)

Figure 5-14. Screen Capture of Received Power Spectrum. (a) at im (b) and 8m

Linx Technologies LR series Transmitter and Receiver

Unfortunately, Rayming Corporation no longer exists. As time goes by, locating

the Ming TX-99 or RE-99 for purchase becomes increasingly difficult as suppliers have

depleted their stocks with no new shipments coming in to replenish their supply.

This prompted the selection of the Linx LR series transmitter and receiver, which is

an update from their LC series line of transmitters and receivers. With these transmitters

and receivers, comes a plethora of application notes and documentation to aide in the

design of a wireless communications link. Linx Technologies also provides pre-

fabricated low-profile antennas for use with the transmitters and receivers. The

outstanding product provided by Linx Technologies, as well as the plethora of


Markerm
3U000000 I~ MIz
-6S.17~ d,':]









documentation notes, makes rapid prototyping and development with the LR series

transmitter and receiver fairly painless.

Linx Technologies TXM-315-LR

The LR series transmitter from Linx Technologies is a high performance

synthesized ASK/OOK transmitter which has the ability to reach a serial data rate of 10

kbps. The transmitter consists of a PLL synthesized architecture offering low-power

consumption, accurate operating frequency, and power-down functions with an antenna

serving as the only external part needed. The system level architecture of the transmitter

can be seen in Figure 5-15 from the TXM-315-LR datasheet.



DATA
PDN


PLL VCO -- PA -- RF OUT



XTAL

Figure 5-15. System Level Architecture of LINX TXM-315-LR

The components of the transmitter consist of a Voltage Controlled Oscillator

(VCO) locked through a phase locked loop (PLL) which is referenced to a high precision

crystal. The output of the VCO is then amplified and buffered by a power amplifier

before the carrier is filtered to attenuate and suppress harmonics and spurious emissions

to within legal limits. The carrier is then output to free space via the 50 ohm antenna port

of the transmitter. The pin-out of the TXM-315-LR transmitter can be seen in Figure 5-

16.










1 JGND PDN 8
2 DATA VCC 7
3 4GND GND 6
4 LADJNCC ANT 5


Figure 5-16. Pin-Out of TXM-315-LR Transmitter

Several unique features such as a Power Down line (PDN) which is used to power

down the transmitter's power amplifier when no data is present for transmission, and a

Level Adjust line (LADJ) which is used to limit the output power for the transmitter, can

be used in unison to even further reduce the power consumption of the transmitter. The

LADJ line can prove to be even more useful during FCC testing and verification to

compensate for antenna gains. By tying the PDN, supply voltage, and Data input node

together, and driving this node with the serial output of the microprocessor, the

transmitter, like the Ming TX-99, can be used as an OOK transmitter.

Initial test setup where the antenna port is soldered to a 50 ohm micro-strip, and

connected to a spectrum analyzer (Agilent E4448A PSA series Spectrum Analyzer) via a

1 ft long SMA cable show the minimum bias of 1.6V to send a 100% duty cycle signal

(transmitter continuously on) requires 6 mA to output 0.34 dBm of output power.

Additionally, at 2V bias, the transmitter can output approximately 3.08 dBm while

draining 8 mA of current. A list of power specifications, with efficiency calculated in the

same fashion as for the Ming TX-99 analysis detailed previously, is seen in Table 5-4.

Table 5-4. Performance of LINX TXM-315-LR
VSUPPLY ISUPPLY(mA) Output Power(dBm) Power(mW) Efficiency(%)
1.6 6 0.34 1.081433951 11.26493699
2 8 3.08 2.032357011 12.70223132










As compared to the Ming TX-99, although output power and efficiency are both

better than the TX-99, the minimum power required to turn the transmitter on is

significantly higher, and may pose a serious problem when trying to obtain power from

scavenged energy. This shows that although Linx Technologies' TXM-315-LR is a more

robust and mature commercial product, the most important requirement of minimal

power expenditure is not met, and may not serve as a suitable RF transmitter for the

wireless hydrogen sensor node.

Linx Technologies RXM-315-LR

Alike to the assumption made for the Ming RE-99, the same assumption that a

central monitoring station can provide as much power as deemed necessary is also made

for the RXM-315-LR. Unlike the products by Ming, and alike to the TXM-315-LR, there

exists a plethora of documentation for the RXM-315-LR, as well as several application

notes.

The system level architecture of the RXM-315-LR is seen in Figure 5.17, and

unlike the envelope detection circuit of the RE-99, the RXM-315-LR receiver modules

employs a single-conversion super-heterodyne architecture to demodulate the received

signal.

SOaF RF IN
(Antenna)
Band Selet 10 7 MHz OataSlicar
Filter 0 IF Filter


9Y,- RS,/,Anabg


Figure 5-17. System Level Architecture for RXM-315-LR









The RF signal entering from the 50 ohm matched antenna is band-pass filtered

before being amplified by an NMOS cascade, Low Noise Amplifier (LNA). The

amplified signal is then down-converted to a 10.7 MHz Intermediate Frequency (IF)

which is done by mixing the amplified signal with a VCO controlled by a PLL,

referenced to a high precision crystal. The mixer stage, which down-converts the signal,

consists of a pair of double balanced mixers, and includes an image rejection circuit.

Once down-converted to an IF frequency, the signal is further amplified, filtered and

demodulated to recover the original base-band data bit-stream. This baseband signal is

squared by a data slicer and output to the DATA pin of the receiver.

This architecture, along with the high IF frequency and ceramic IF filters, helps

reduce the susceptibility to interference, which is a problem associated with OOK

modulation. The pin-out of the receiver can be seen in Figure 5.18. Due to the

architecture and components of this receiver, it is able to achieve a very high sensitivity

of-112 dBm, while remaining unsusceptible to interfering signals which plague OOK

communication links.

1 NC ANT 16
2 NC GND 15
3 NC NC 14
4 GND NC 13
5 VCC NC 12
6 PDN NC 11
7 RSSI NC 10
8 DATA NC 9

Figure 5-18. Pin-Out of RXM-315-LR Transmitter

Compared to the Ming RE-99 receiver, Linx Technologies RXM-315-LR receiver

shows to have the best performance out of both receivers, while the Ming TX-99 shows









to have the most favorable performance between the two transmitters. To optimize the

wireless communication link, the combination of the Ming TX-99 and Linx RXM-315-

LR should be used together. Unfortunately both components are set to operate at

different frequencies. As stated before, one of the attractive characteristics of the Ming

TX-99 is the tunable variable capacitor which can change the resonant frequency of the

LC tank. If the LC tank were to be tuned to 315 MHz, the Ming TX-99 can be used in

conjunction with the RXM-315-LR to create an optimized wireless link for the wireless

hydrogen sensor node.

Wireless Link Optimization

The optimization of the wireless link includes not only the combination of the

Ming- TX-99 transmitter with the Linx Technologies RXM-315-LR receiver, but also the

minimization of power consumption by the Ming TX-99 transmitter, development of a

low-profile antenna to reduce the size and increase the compactness of the overall

wireless sensor package, and work on the receiver side to gather data from the transmitter

for use as a central monitoring station. Additionally, studies should be done to develop a

wireless node operating within the restrictions set by FCC part 15.231, which regulates

operation within the 260-470 MHz range. Prior to all the additional work, a starting

ground should be set by performing an initial power analysis of the system.

Ming TX-99 Power Analysis

A power analysis on the Ming TX-99 was performed by David Johnson, at Cisco

Systems in Bradenton, FL. Similar to the power analysis performed for the

microcontroller in Chapter 4. To reiterate the test setup, a 383 ohm resistor was

connected in series between the power supply, and microcontroller/RF transmitter. The

USART output of the microcontroller was used to power and drive the Ming TX-99









transmitter. P6248 Differential probes attached to a Tektronix TDS5104B Digital

Phosphor Oscilloscope were used to measure the voltage across the 383 ohm resistor.

Calculation for the average power consumption is as follows:

* Total Area (TA) under the measured curve is calculated (units of V-sec)

* Peak power is calculated as:
VMAX
PeakPower = -- VSUPPLY
383
Where VMAX is the maximum point of the measured curve

* Average power is calculated as
TotalArea UPP
AvgPower= VS- r UPPLY
STduration 383 )
Where Tduration is the duration of the measured curve

It was previously shown in Chapter 4, that for any task other than for initialization,

the MSP430 microcontroller only consumed a low 2.5uW of power. To test the average

power consumption for the transmission of a 500 uS bit length pulse, the procedure above

is applied, and the measured curve taken from a screen dump of the Tektronix

TDS5104B Digital Phosphor Oscilloscope can be seen in Figure 5-19.

522uW



------------------- 6ms ---------- ----|

|--0-5 ms







Figure 5-19. Tektronix TDS5104B Digital Phosphor Oscilloscope Screen Capture of
Power Analysis Performed for RF Transmission of One Bit









From this curve, it can be calculated that for the transmission of a 500 uS pulse, the

transmitter consumes an average power of 261 uW with a peak of 522 uW. As expected,

the RF transmitter does not consume power for the RF transmission of a logical "low" or

"O". The charging and discharging characteristics of Figure 5-19 may be due to the LC

resonant tank which is used to set the operating frequency of the RF transmitter to 300

MHz.

Another experiment was performed to analyze the power consumption for the RF

transmission of multiple bits. The screen capture from the Tektronix TDS5104B Digital

Phosphor Oscilloscope can be seen in Figure 5-20. From Figure 5-20, it can be seen that

all rising slopes and falling slopes are equal.



















Figure 5-20. Tektronix TDS5104B Digital Phosphor Oscilloscope Screen Capture of
Power Analysis Performed for RF Transmission of Multiple Bits

By grouping all rising slopes together at the front, and grouping all falling slopes

together in the back to arrange a simple triangle, and assuming that the power is not

completely discharged between "high" bits, it can be gathered that the worst case










average power consumption (maximum average power consumption) to send N bits, can

be equated as:


1
WorstCasePower = PeakPower Nbits
2

Where PeakPower is the peak power required to transmit a single bit (261 uW), and

the worst case data input to the transmitter is a continuous train of"1" or "high" pulses,

one after the other with no "low" or "0" bits in between.

Low Profile Antenna

To increase the compactness of the wireless sensor, Linx Technologies provides the

ANT-315-SP, or "SPLATCH" antenna. The features of this antenna are an ultra-compact

package and good resistance to proximity effects. The SPLATCH uses a grounded-line

technique to achieve a quarter wave type antenna centered at 315 MHz, with a bandwidth

of 5 MHz. The "SPLATCH" antenna with dimensions can be seen in Figure 5-21.


1.12"
(28.4)

0154 PLA
(13.7) -06
(1 5)j

S (5.1)


No ground plane or traces
PLA I under the antenna
nS^^ GPCB rJ;, A... r1 Splatch
SVas to ground plane
S1.S5 x .0" mln. Ground plane on bottom
ground plane layer for counterpoise
50 ohm microstrip line


Figure 5-21. LINX ANT-315-SP SPLATCH Antenna From Datasheet









Unfortunately, unlike the TXM-315-LR transmitter or RXM-315-LR receiver, the

SPLATCH antenna does not provide much in documentation. The only details given are

the antenna's requirements for a 1.5" x 3.0" ground plan, and a 50 Ohm micro-strip line

between RFIN of the antenna, and the RF output node of the transmitter. An assembled

FR4 testing board for the SPLATCH antenna can be seen in Figure 5-22.
















a) b)

Figure 5-22. Testing Board for SPLATCH Antenna. (a) Front(b) and Back

Initial tests to obtain the gain of the SPLATCH antenna were performed with an

Agilent E8316A 10 Mhz to 6 GHz PNA series Network Analyzer, and an Agilent

E4448A PSA Series Spectrum Analyzer. An antenna test structure was tied to the

Agilent E8254A 250 kHz to 40 GHz PSG-A Series Signal Generator via a FLX402#1

SMA cable to serve as the transmitting antenna, and an identical antenna structure was

tied to the Agilent E4448A PSA Series Spectrum Analyzer outputting a RF power of 10

dBm at 315 MHz, via a FLX402#1 SMA cable to serve as a receiving antenna. With

respect to the free space path loss equation, where










PT.GT.GR *A2
PR(R) =
(4-.r-R)2

Assuming frequency, distance, and transmitted and received power are all

controlled parameters, since the transmitter and receivers antennas are identical, their

respective gains can be calculated. For these calculations, cable losses were also taken

into account. Initial gain measurements from different distances showed the data found

in Table 5-5. As a comparison, the gains of the 22 gauge copper quarter-wave monopole

antennas were also tested to serve as a comparison to the SPLATCH antennas.

Table 5-5. Antenna Gain Measurements
Received Transmitted
Distance Received TransmPath Loss Cable Loss Gain
S Power Power ( Antenna
(m) (dBm) (dBm) (dB) (dB) (dB)
(dBm) (dBm)
4 -58.11 10 34.44918309 0.41 -16.6254 SPLATCH
5 -60 10 36.38738335 0.41 -16.6013 SPLATCH
4 -26 10 34.44918309 0.41 -0.57041 Monopole
5 -28 10 36.38738335 0.41 -0.60131 Monopole

From these measurements, it can be seen that the "SPLATCH" antennas exhibited

poor performance with an approximate gain of-16.6 dB, as compared to the -0.6 dB gain

of the monopole antennas. Clearly, there is a distinct performance loss of the SPLATCH

antenna over the monopole antennas, and a more comprehensive analysis of the

SPLATCH antenna should be conducted.

To measure the resonant frequency of the SPLATCH antenna, a 1 Port S-parameter

measurement was taken using an antenna, and an Agilent E8316A 10 Mhz to 6 GHz PNA

series Network Analyzer. Additionally, the antenna test structure was simulated in

Ansoft Designer. From the S-Parameter measurements taken by Agilent E8316A 10 Mhz

to 6 GHz PNA series Network Analyzer shown in Figure 5-23, the SPLATCH antennas







74


showed to have a resonant frequency located at 330 MHz with a bandwidth of 5 MHz,

rather than 315 MHz as shown in the specifications of the datasheet.


22 Mar 2006 Ansoft Corporation 19:24:43 Y1-0
XY Plot 5 dB(S(Portl Port1
PlanarEM1 jerry_splatch jer

0.00-






-10.00-







0
-20.00- f






-30.00- -- .
200.00 250.00 300.00 350.00 400.00
F [MHz]
X1= 330 00MHz
Y1= -27 63

Figure 5-23. S-Parameter for SPLATCH Antenna

Since the fundamental operating frequency is 315 MHz, this may be the reason for

the poor performance of the antenna. A planar EM simulation in Ansoft Designer was

performed, and showed similar results compared to the measurements taken from the

E8316A 10 Mhz to 6 GHz PNA series Network Analyzer as seen in Figure 5.24(a). Also

seen in Figure 5.24(b) is the corresponding antenna test structure model within Ansoft

Designer.

To shift the resonant frequency of the SPLATCH antenna down to 315 MHz, a

matching circuit consisting of a shunt and series capacitor were simulated in Ansoft

Designer, and then realized and tuned on the SPLATCH antenna test structure. The












frequency shifted matched antenna S parameters can be seen in Figure 5-25 with


comparison to the original un-matched antenna. Figure 5-26 shows that the matched


antenna retains an approximate -10 dB bandwidth of 5 MHz.


Figure 5-24. S-Parameters of Measured (red) and Simulated (blue) in Ansoft Designer.
(a) S 11 (b) Simulation Setup


08 Feb 2006 Ansoft Corparmwim 15:V1 -O--
XY Plot 2 IB(S(PoriPortl))
Circuil pery_ spltch erryj


dB(S(P oIPtI1))
_slatch_2: Jerr








-tj









200. 5 30 M 30. 10 40 M

= O.0Mz 2= 315.00MHz
I1=63 Y2=-24 85


Figure 5-25. Matched Antenna (Red) vs. Unmatched Antenna (blue)


- - -


niim


~~nF~~a~
Xl~il
~ln~lll











Ansdt Crpmrtion 15:08:02
XY PIkt 3
Cftlu


08 Feb 2006














aI


F 1U2z]


Xl = 313 OOMHz
Y1 = -8 74


X2 3180 OItH
Y2= -8.15


Figure 5-26. 5 MHz Bandwidth of Matched Antenna

Once matched, the SPLATCH antenna was again tested for gain. The experiment


setup is identical to the previously mentioned set up, and the data for this experiment can


be found in Table 5-6.


Table 5-6. Gain Measurements for Matched Antenna
Received
Distance Re d Transmitted Cable Loss
() Powe Power (dBm) Path Loss (dB) (dB) Gain (dB)
(m) (m)Po_________wer (dBm) (dB)_________ _____
2 -49 4 28.42858318 0.11 -12.2307
3 -50 4 31.95040836 0.11 -10.9698
4 -48 4 34.44918309 0.11 -8.72041



These results show an approximate 6 dB increase in performance, but the resulting


gain is still lower than that of the quarter wave 22 gauge copper monopole antenna. This


may be due to the poor radiation efficiency stemming from electrically small


characteristics of the antenna as compared to the 22 gauge copper monopole antennas. A


trade-off thus exists in the reduction of gain for a compact surface mount antenna. For







77


optimization purposes, since size is an unlimited parameter at the central monitoring


station, the quarter wave 22-gauge monopole antenna can be used as the antenna for the


receiver, while the SPLATCH antenna can be used as the antenna for the wireless sensor


node.


RF Transmitter Optimization

As previously mentioned, an attractive feature of the Ming TX-99 transmitter is the


simplicity of the design, and the exposed discrete components that can be changed and


replaced to tune both frequency, and power consumption. All components were extracted


and measured, with the micro-strip inductance measured by an Agilent E8316A 10 Mhz


to 6 GHz PNA series Network Analyzer. The impedance measurement of the inductor


can be seen in Figure 5-27.


22 Mar 2006 Ansoft Corporation 18:46:30 Y1
XY Plot 1 im(Z(Portl,Port1
PlanarEM1 , ,nd o


00 .0 250.00 300.U0 350.00 400.00
F [MHz]
1= 200 00MHz X2= 201 00MHz
S132 40 Y2= 32 45


Figure 5-27. Microstrip Inductance Measurement for Ming TX-99 Onboard Antenna


Y1-P-
re(Z(Portl ,Port1
Minglnductor N


60.00-



40.00- --



20.00



0.00



-20.00- _


I


I









By placing the value of these components into Agilent ADS 2003, the discrete

components can be varied to reduce the minimum power requirements of the RF

transmitter. A list of all values relating to the discrete components of the Ming TX-99

can be found in Table 5-7.

Table 5-7. Component Values of Ming TX-99 Transmitter
Component Value
C1 2-7 pF
C2 12 pF
C3 3.3 pF
L1 28.22 nH
L2 1 uH
R1 47 kOhms
R2 100 Ohms
Transistor MMBTH10


By simulating the circuit found previously in Figure 5-5 in Agilent ADS 2003, the

discrete components directly correlating to output power and power consumption can be

found. The simulation setup in Agilent ADS can be found in Figure 5-28. From these

simulations, it is seen that the major components that control output power and power

consumption were related to the resistors found at the base and emitter of the transmitter.

The emitter resistance increases the stability of the transmitter, so the component with the

most effect on the output power and power consumption is the resistor in the base. Also

discovered, is the output power can be increased by the removal of the diode also found

at the base of the transistor.

Another component that had an influence on the power characteristics of the

transmitter was the high-frequency NPN BJT transistor. By choosing a different

transistor capable of high frequency operation, and had lower power requirements, the









79




overall power of the transmitter can be further reduced. Table 5-8 shows the results of an



internet search for high-frequency NPN BJT transistors.


PinDiodeModel DC
NLRNM1 DC1
s= Af=
VI= Ffe=
n= AiParams= HARMONIC BALANCE
Rr= HarmonicBalance
Cmln= HB
HB1
Tau Freq[1]=300MHz
Co= Order[1]=7
VIJ=
M=
Fc=
,ma= I TRANSIENT
Imelt=
Kf= Tran
Tranl
SStopTime=300 usec
MaxTlmeStep=10 mec


BJT Model
MMBTH10LT1
NPN=es Kc= Cex= Xtf=15 Ffe=
PNP=no Isc=1fA Cco= Tf 4nsec Lateral=no
Is=32868fA C4= Imax Vtf=1 RbModel=MDS
Bf=75328 Nc=1 05387 Imelt= Itf- Approxqb=yes
Nf-117427 Cbo= Cje=112827pF Ptf0 Tnom=
Vaf-10 Gbo= Ve=0 65 Tr= lusec Trse=
I1=-0262558 Vbo= Mje=035 Kf-0 Eg=105
Ise=1fA Rb=01 Cjc 165813pF Af-1 Xtb=01
C2= lrb=237229 Vjc=095 Kb= Xtl=1
Ne= 14992 Rbm=01 Mjc=023 Ab= AIIParams=
Br=0162338 Re=0001 Xcjc=08 Fb=
Nrl 5 Rc=9 5364 Cjs= Rbnol=
Var=100 Rcv Vjs=0 75 Iss=
1Ir=262558 Rcm= Mjs=05 Ns=
Ke= Dope= Fc0 533333 Nk-


Figure 5-28 Agilent ADS 2003 Simulation Environment


Table 5-8. Various High-Frequency NPN BJT Transistors

NPN Transistor Ic(mA) PTOT(mW) fT Noise Figure

MMBTH10 50 1250 650 MHz N/A

BFS19 25 500 260 MHz N/A

BFT25A 6.5 32 5 GHz 1.8 @ 1GHz

BFW92A 25 375 3.2 GHz 2.5 @ 800 MHz

BFS17A 25 300 3 GHz 2.5 @ 800 MHz



Agilent ADS was used to find the minimal operating point for the transmitter to



turn on and transmit present a carrier. The simulation setup within ADS 2003 was seen



previously in Figure 5-28. To actually test the performance changes, a manual sweep of



these values was performed by de-soldering and re-soldering different components onto



the transmitter board, and measuring the output power and current consumption given a



2V supply, in a similar setup as previously described. However, instead of using an










Agilent E4448A PSA Series Spectrum Analyzer, a HP 8563E spectrum analyzer was

used. Table 5-9 shows the results of this trial and error experimentation with the

transmitter.

Table 5-9. Performance of Various Transistors and Resistors for Ming TX-99 Transmitter
Base Emitter y u Output Output Efficiency
Transistor Resistanc Resistance l up Power Power
e (kOhms) (Ohms) oltae current (dBm) (mW)

MMBTH10 47 100 2 2.75 -3 0.5012 9.1124952
MMBTH10 47 300 2 2 -13 0.0501 1.2529681
MMBTH10 100 100 2 1.77 -5 0.3162 8.9329877
MMBTH10 100 200 2 0.911 -11 0.0794 4.35965
MMBTH10 200 100 2 0.996 -9 0.1259 6.3199067
MMBTH10 200 200 2 0.923 -11.83 0.0656 3.5544164
MMBTH10 200 300 2 0.86 -13 0.0501 2.9138793
BFT25A 47 100 2 1.73 -1.33 0.7362 21.277662
BFT25A 47 200 2 1.52 -3.33 0.4645 15.280108
BFT25A 100 100 2 1.095 -5.5 0.2818 12.869328
BFT25A 150 100 2 0.8 -8.5 0.1413 8.8283597
BFT25A 200 100 2 0.65 -10.93 0.0826 6.3541381
BFS17A 47 100 2 2.1 -2.67 0.5408 12.875103



In Table 5-9, the resistance combination which produces the highest efficiencies for

each transistor are boxed in red. It is found that the Philips' BFT25A high-frequency

NPN BJT transistor served as the best transistor for use in the RF transmitter. By using

the same nominal resistances from the original Ming Tx-99 transmitter, it was capable of

increasing the efficiency by increasing the output power, while lowering the overall

power consumption for the transmission of a constant 315 MHz carrier, as compared to

the original Ming TX-99 tuned to 315 MHz.

FCC Part 15.231

The operating frequency of 315 MHz lies in the FCC frequency operating band of

260 MHz to 470 MHz, which is regulated by Part 15.231 of CFR47. The specific








81



regulations of part 15.231 are rather unusual in that for many bands, the FCC specifies


only fundamental power, harmonic levels, and allowed bandwidth, while for the


frequency band of 260 to 470 VMHz, the FCC regulates this spectrum based on the


intended function and form of the transmitted data. Part 15.231 is broken into paragraphs


A through D, while paragraph E applies only if the rules specific to paragraph A are


broken. Due to the complexity and application specific regulations of part- 15.231, the


limitations of part 15.231 are best illustrated in a flowchart form as seen in Figure 5-29


taken from [23]. Interested parties should visit the FCC website at


http://wireless.fcc.gove/rules.html.



Is My Application Legal for Operation

Under Part 15.231 A-D?
I


Allowed
Trannsmisson Types
under ia 231 A-D You ay Tranrr :
* wntl Dr cornmain agna
* coofla in ofaerto iientir a
"yaBm canponent
* ado conol agnalsa during
Gmergences
* vanasfl aa as lg an acfmr ora
K cad Se Bent k l a


p-n Banned
Transmission Types
UndBsr 15.l21 A-O You my Not send:
voice orvMdeo
cmtre tonya
VerlatNtOs [atSwV u tacantcDfal e
(tama, premur. ternperatur, eo.)
* parxidc tranrtagor atE ragiar
pre-tnnfnOd ntI valS ptU
po inmegrmy of aysem cornponerim





Any Operato n la Allowed
UnderPafograpn EP Povided
1. ThB output pOci1er1 halved.
2. MaEimium transnmBison tne 1 sa.
a. Th minimum peod blstwoen
transm ERtonEl ISa3 M nMeG i
trunisfneon peoni but never lsee
than 10 seco8na
4. The hnamolC and baifdiMt
requ rIenfts or Para B-D are
compiled wift.
a ____^


Figure 5-29. Flowchart for FCC Part 15-231 Requirements


Cmn with ta outlliEm p e
anraqma, mid hawutni mpr nrcmerwi of
peragrapna -0 anatne procua imaurMMine
Tfo coruncmeon









The limitations of part 15.231 section A are expressed in sections B through D, and

can be found in Table 5-10, while the limitations of section E can be found in Table 5-11

and illustrated in Figure 5.30 from [24] .

Table 5-10. Limitations under FCC Part 15.231 (a-d) **Linear Interpolations
Fundamental Frequency Field Strength of Field Strength of Spurious
(MHz) Fundamental (uV/m) Emission (uV/m)
40.66- 40.70 2250 225
70-130 1250 125
130 174 1250 to 3750** 125 to 375**
174-260 3750 375
260 470 3750 to 12500** 375 to 1250**
Above 470 12500 1250

Table 5-11. Limitations under FCC Part 15.231 (e) **Linear Interpolations
Fundamental Frequency Field Strength of Field Strength of Spurious
(MHz) Fundamental (uV/m) Emission (uV/m)
40.66- 40.70 1000 100
70-130 500 50
130 174 500 to 1500** 50 to 150**
174-260 1500 150
260 470 1500 to 5000** 150 to 500**
Above 470 5000 500


EFwmeawntjapWn)

0 t00 -------------


----/


/i


Fc~.MMZ


Figure 5-30. Graphical Representation of Field Strength Limitations for Part 15.231
Section e.


43W
4133



16100


/

. .
*
*









For the frequency of 315 MHz, the field strength of the fundamental in section A

can be calculated as uV/m at 3 meters = 41.6667(F) 7083. 3333, while for section E, the

field strength can be calculated as uV/m at 3 meters = 16.6667(F) = 2833.3333, where F

is the operating frequency in MHz. This shows that the requirements for section E of part

15.231 are lowered by almost 40 % as compared to the limitations for part 15.231 section

A. Since multiple accessing schemes are currently being developed, and an ID tag will

be assigned to every transmitted message, The RF transmitter should be designed to meet

the requirements of Part 15.231 section a found in Table 5-10.

For an operating frequency of 315 MHz, the field strength fundamental is

calculated to be 6.041mV/m referenced to 3 meters. Assuming an isotropic antenna of

gain 1, the allowed transmitted power can be calculated as:

1 (E2d
PTransmitted = -- E2 -d
30

Where E is the field strength and d is the reference distance of 3m. From this

equation, it can be found that the transmitted power, assuming an isotropic antenna of

gain 1, is approximately 10 uW, or -19.6 dBm. Given the best case gain from the

SPLATCH antenna of -8.72 dB, an output power less than -10.88 dBm, or approximately

-11 dBm is required from the transmitter to meet the FCC part 15.231 requirements for

radiated power. From Table 5-9, it is shown that the Ming TX-99, with a base resistance

of 200k ohms, emitter resistance of 100 ohms, and the NPN BJT transistor replaced with

the Philips BFT25A, fulfills the -10.88 dBm output power requirement with an output

power equal to -10.93 dBm. Thus, by using this specific transmitter configuration, the

requirements of FCC part 15.231 can be fulfilled.









Given a receiver sensitivity of -112 dBm, receiver antenna of approximately -1 dB,

transmitter antenna of -9 dB, output power of-10.93 dBm, and the two-ray ground

reflection model, assuming the height of both the transmitting and receiving antenna is 1

meter, the theoretical distance of the wireless sensor node can be calculated to be

approximately 189 meters. If we include an estimated 15 dB path loss associated with the

loss due to random variables such as multi-path fading effects, this distance is lowered to

approximately 79.7 meters.

Central Monitoring Station

The transmitters of the wireless hydrogen sensor eventually will be required to

communicate to a central monitoring station. For the design of the central monitoring

station which consists of a RF receiver module, a Data Acquisition (DAQ) device, and a

laptop or desktop computer, it is assumed that power consumption is of little worry for

the development of the receiving unit. The central monitoring station was originally

designed for use with the Ming RE-99. Since the RE-99 is of an envelope detection

topology, the RE-99 is especially sensitive to interference and noise when a carrier is

absent. To remedy this problem, considerations into software decoding should be

emphasized in the design of the central monitoring station to filter and remove the effects

of noise and interference.

The central monitoring station currently consists of the Ming RE-99 receiver, and a

National Instrument USB-6008 DAQ device. The Ming receiver is powered and outputs

data through an analog output and input node on the USB-6008 DAQ device. The USB-

6008 DAQ device itself is powered via an USB port from a laptop running LabVIEW

7.1. The USB-6008 provides basic data acquisition functionality for the purpose of data

logging, and making portable measurements, while being fairly affordable, but powerful









enough for sophisticated measurement applications. The USB-6008 consists of eight 12-

bit analog input channels, 12 digital I/O lines, 2 analog outputs, and 1 counter, with a

maximum analog sampling rate of 10 k Samples/sec and an analog output range of 0 to

+5 V which is sufficient for powering and gathering received data from the Ming RE-99

receiver.

When a carrier is absent, due to the envelope detection topology of the Ming RE-

99, the receiver becomes susceptible to noise and interference. To mitigate the effects of

the noise and interference, software coding was developed to distinguish data pulses from

noise and interference. The proposed solution is to use the maximum sampling rate of

10kHz for the DAQ, and continuously sample, taking one analog sample every 100 uS.

Then these samples are buffered, and a moving average filter is used to differentiate

pulses from noise and interference. An example of a moving average filter can be seen in

Figure 5.31.





Pulse









Figure 5-31. Moving Average Filter Example

From Figure 5.31, it shows that the middle four samples are the pulse, and the rest

is noise. If the samples are continuously filtered over a 4 sample window, and if the

average of those 4 samples is equal to, or greater than the expected amplitude, then a






86


pulse has been detected. This moving average filter was implemented in LabVIEW code,

and can be seen in Figure 5-32(a). To provide a graphical interface, the LabVIEW front

panel user gui can be found in Figure 5-32(b).




-P1

Ue .B ... I:^..


a) b)

Figure 5-32. Labview Code. (a)Block Diagram Code (b) Labview Front Panel Gui

The considerations, design, and optimization of a wireless communications link

have been described in this chapter. A modulation scheme of OOK was selected, and a

carrier frequency of 315 MHz was picked as the operating frequency of the wireless

communications link. Additionally, a low-profile antenna was tested and matched, and

the transmitter set to operate within the confines of FCC part 15.231. The development

of a central monitoring station was also described. To reiterate the selection of an OOK

modulation scheme, OOK is attractive in that power is only consumed for the

transmission of a "1" or "high" bit, and consumes little to no power for the transmission

of a "0" or "low". By exploiting these characteristics of OOK, perhaps a source coding

scheme can be used to further reduce the power consumption required to send a fixed

length message. A source coding scheme for minimum energy expenditure is described

in the following chapter.