| 00:34.18 | *** join/#qi-hardware fengling (~fengling@58.135.95.135) |
| 01:18.55 | *** join/#qi-hardware fengling (~fengling@58.135.95.135) |
| 03:31.28 | *** join/#qi-hardware sulky_ (~sulky@148.163.189.33) |
| 03:46.30 | *** join/#qi-hardware sulky_ (~sulky@148.163.189.33) |
| 04:01.03 | *** join/#qi-hardware DocScrutinizer05 (~saturn@openmoko/engineers/joerg) |
| 04:02.22 | *** join/#qi-hardware sulky (~sulky@148.163.189.33) |
| 04:21.17 | *** join/#qi-hardware qi-bot (~qi-bot@turandot.qi-hardware.com) |
| 04:21.17 | *** mode/#qi-hardware [+o qi-bot] by ChanServ |
| 04:56.54 | *** join/#qi-hardware sb0 (~lekernel@121.139.198.203.static.netvigator.com) |
| 06:38.42 | *** join/#qi-hardware pcercuei (~pcercuei@92.177.36.45) |
| 08:08.03 | *** join/#qi-hardware pcercuei (~paul@137.71.226.54) |
| 08:12.08 | *** join/#qi-hardware eintopf_ (eintopf@zzz.strace.org) |
| 10:52.09 | *** join/#qi-hardware DocScrutinizer51 (~lagrange@openmoko/engineers/joerg) |
| 11:52.50 | *** join/#qi-hardware sb0 (~lekernel@210006059234.ctinets.com) |
| 12:33.38 | *** join/#qi-hardware fengling (~fengling@106.37.236.225) |
| 14:05.55 | *** join/#qi-hardware fengling (~fengling@106.37.236.225) |
| 14:13.28 | *** join/#qi-hardware pcercuei (~paul@137.71.226.54) |
| 15:47.09 | *** join/#qi-hardware roh (~roh@yamato.hyte.de) |
| 15:47.40 | *** join/#qi-hardware whitequark (whitequark@fehu.whitequark.org) |
| 16:58.37 | *** join/#qi-hardware SSomeOne (~Plissken@apolo.padtec.com.br) |
| 17:18.44 | *** join/#qi-hardware pcercuei (~pcercuei@92.177.36.45) |
| 18:44.51 | *** join/#qi-hardware dandon (~dandon@185.16.52.103) |
| 19:55.48 | *** join/#qi-hardware wildlander (~wild@unaffiliated/wildlander) |
| 20:06.22 | roh | hm. anybody played with miniature (<3mm diameter) cmos cameras before? |
| 20:07.20 | roh | i got some for free and am wondering how to get them to talk to me.. it got i2c, some vout line and a '4mhz' labeled one. still not sure if i need to provide a clock in there or what |
| 20:42.08 | whitequark | probably has ntsc output |
| 21:07.37 | roh | whitequark: maybe... any idea why 4mhz? |
| 21:08.33 | roh | ntsc should eat 4.433 something i thought |
| 21:09.00 | roh | or 3.5something |
| 21:09.54 | roh | well.. i found a 3.6864 oscillator in the misc parts box and hooked it up.. voila.. there is the i2c working somehow. a device on 0x36 which wasnt there before |
| 21:14.59 | *** join/#qi-hardware luke-jr (~luke-jr@unaffiliated/luke-jr) |
| 21:57.31 | DocScrutinizer05 | hook your scope to vout |
| 21:58.03 | DocScrutinizer05 | but I'm afraid you will need to initialize the thing via I2C |
| 21:58.32 | DocScrutinizer05 | not exactly simple to guess or RE how to do that |
| 22:01.16 | DocScrutinizer05 | wpwrak: http://bucket1.glanacion.com/anexos/fotos/30/medio-ambiente-2239330w640.gif is deviation from long term average |
| 22:01.46 | DocScrutinizer05 | no 'icy spot' |
| 22:20.53 | roh | hm. scoped it. i2c dumped it... and i guess my clock is too far off... anyhow.. the available brief-sheet says its not anything normal like regular video... 400x400 pixels says that quite directly. |
| 22:31.17 | roh | i fear i need to build some digizer extra for this |
| 22:45.38 | DocScrutinizer05 | err, teh horizontal "pixels" (correctly lines) are irrelevant in a CVBS signal. And 400 lines vertical would be "normal" for NTSC, I.E should still show a picture when the frame rate is 45 to 65Hz |
| 22:49.39 | DocScrutinizer05 | worst case use an old CRT TV, those are usually even more tolerant against out-of-specs framerates, vertical resolution lines etc. Just the line frequency should be 15kHz to 16KHz |
| 22:50.36 | DocScrutinizer05 | attaching a scope to vout you's easily see the line frequency, either it's TV ~16kHz, or VGA ~32kHz |
| 22:52.49 | DocScrutinizer05 | problem would be progressive n H * 400 V for TV, you have interleave there and so a usual TV could only show cope with max ~700/2 lines per frame |
| 22:53.00 | DocScrutinizer05 | progressive |
| 22:53.57 | DocScrutinizer05 | VGA works also with 60Hz frame rate but 400 progressive, so the line frequency is twice as high |
| 22:55.05 | DocScrutinizer05 | aka 640H*400Vprog@60 |
| 22:57.00 | DocScrutinizer05 | while PAL TV is nH*~512Vinterleave@50 |
| 22:57.49 | DocScrutinizer05 | oops sorry NTSC has 486V |
| 22:58.59 | DocScrutinizer05 | and VGA 640*480 |
| 23:03.28 | DocScrutinizer05 | bottom line: as long as you see a semi-decent CVBS signal on vout, you'll most likely find a monitor (or TV) to sync and display that signal |
| 23:04.35 | DocScrutinizer05 | nasty when vout has a serial digital signal though |
| 23:08.34 | DocScrutinizer05 | dunno if https://en.wikipedia.org/wiki/Camera_Serial_Interface#CSI-1 would come with one non-differential line, for example |
| 23:16.48 | DocScrutinizer05 | lane, not line |
| 23:18.51 | DocScrutinizer05 | nope, even CSI-1 comes with differential lanes (and clock) only, at least CSI-1 actually has only one lane |
| 23:19.39 | DocScrutinizer05 | http://www.ti.com/general/docs/lit/getliterature.tsp?genericPartNumber=sn65lvds315&fileType=pdf |
| 23:21.35 | DocScrutinizer05 | more convenient URL: http://www.ti.com/product/SN65LVDS315 |
| 23:32.40 | roh | its not digital as far as i can guess. its a coax line and the sheet says analog out |
| 23:32.43 | roh | http://www.buyemp.com/product/ambu-ascope-3-5-0-2-2 |
| 23:32.46 | roh | this is what i got ;) |
| 23:33.11 | roh | and i do NOT have this one to just 'sniff it' ;) http://www.buyemp.com/product/ambu-aview-monitor |
| 23:35.02 | roh | because the monitor gets reused, the camera-endoskop gets thrown away as medical trash usually |
| 23:38.36 | roh | my plan is to recycle those for weird uses... maybe my friends with their beehives want some more sensors? |
| 23:40.54 | DocScrutinizer05 | nice |
| 23:41.27 | DocScrutinizer05 | yeah, coax and analog out are very good signs |
| 23:41.52 | DocScrutinizer05 | I'd suggest to connect it to a cinch and plug it into a CVBS-in |
| 23:42.00 | DocScrutinizer05 | ;-) |
| 23:44.10 | DocScrutinizer05 | could you mod one to wiretap the I2C and connect to a monitor device while sniffing? |
| 23:45.03 | DocScrutinizer05 | since you will most likely need to confirure one or two dozen registers to make the critter behave |
| 23:47.10 | roh | i dont have access to a monitor. it sits in a sterile room which i dont have access to ;) |
| 23:48.06 | roh | well.. next i'll get some better oscilator.. like 4mhz instead of 3.56someting |
| 23:48.50 | DocScrutinizer05 | wore, not only you but also any laptops or whatever you would bring doesn't have access too |
| 23:48.54 | DocScrutinizer05 | worse* |
| 23:48.56 | roh | atleast ive spoken a few bytes i2c or SCCB like omnivision calls it with something today |
| 23:49.42 | roh | SCCB seems to be i2c with no ack bits... which is kinda stupid but well... |
| 23:49.56 | DocScrutinizer05 | meh |
| 23:50.35 | DocScrutinizer05 | did a i2cdump? |
| 23:50.49 | roh | http://pastebin.com/UnZNpRzy |
| 23:51.24 | DocScrutinizer05 | hmm 19 to 20 registers |
| 23:51.42 | roh | some omnivision devices have an 16bit id register at address 0x300a |
| 23:51.46 | DocScrutinizer05 | to 25 |
| 23:52.01 | roh | so i just guessed that 69 30 is this.. at 0xa |
| 23:53.10 | DocScrutinizer05 | :nod: |
| 23:53.52 | DocScrutinizer05 | no manuf logo? |
| 23:54.17 | roh | and voila.. some google brought up http://www.ovt.com/products/sensor.php?id=49 , which is the first sensor ive looked at today which had some kind of analog out at all and is small enough to fit what i see as case (its 5.5mm total diameter and has a 2.2mm tool-shaft) |
| 23:54.47 | roh | http://www.ambu.de/deu/produkte/area_of_interests/anaesthesia/product/ambu%C2%AE_ascope%E2%84%A2_3_5-0-2-2-prod17261.aspx |
| 23:54.48 | DocScrutinizer05 | :-D |
| 23:54.50 | roh | this one |
| 23:55.30 | DocScrutinizer05 | nah on the sensor, but you already found it |
| 23:57.03 | DocScrutinizer05 | here you are 400*200@60, "full @30" |
| 23:57.47 | roh | well.. thats my best guess so far. i am not sure at all... the numbers is just plain guessing from the i2cdump |
| 23:58.38 | roh | if that pcb in the handpiece would not have been labeled with silkscreen at the solderpads for the cable from the camera and to the monitor... i would still not have power on it. |
| 23:59.55 | roh | the 2 leds in the head get their own vcc and seem to like 20mA at 6.0-6.1V just fine... and my 'doctor friend' who uses these on a daily basis gave me his best guess when testing out different currents about how bright its on the original device |
