Flir Tau: strange video output...

[ivaylo]

Haven't used the Tau for two months or so. I turn it on now and video output is as if the sensor isn't seeing anything. Communication with camera via Camera Controller software is normal.
Flir Part number: 427-0077-51-00
FPA Size: 336x256
Symptoms:

• Again, digital communication with PC is fine, all settings can be read and changed (camera_controller.JPG)
• I do see the test pattern when turned on from Camera Controller, so video signal does come out (test_pattern.JPG)
• Otherwise video is just a gray blob as if camera has no lens, no matter what lens I try output is just gray no matter where I point it. The pixels do flicker a little like in normal operation, i.e. image is not frozen (after_fcc.JPG)
• Auto FFC seems to be working (I hear it clicking) but one thing I noticed is that after a FFC cycle the image starts deteriorating (develops these diagonal 70deg? lines) but the next FFC clears them (before_fcc.JPG - contrast is increased on purpose).

Can the sensor die with everything else working like this? The camera has been taken apart in the past, but I believe it was before I've seen it working last. I feel I am missing something simple here, appreaciate your help... 

[LesioQ]

Perhaps irrevelant, but in the past I had my Seek Thermal invert it's shutter action, so it was mostly closed and opened only for calibration.
This resulted with almost washed-out image. What may be the cause ? Electrically inverted shutter coil ?
Maybe an external magnet re-polarized the shutter in reverse ? Bit logic fail ? I don't know. But eletrical reversing the shutter drive (wires) fixed that.

[ivaylo]

Thanks for your response. Yeah, that’s a weird one, nice you fixed it. The shutter here closes correctly, the lens is removable and shutter is big enough I can see it. But yes, the video output is as if there is something in front of the sensor.

[Fraser]

If that Tau core has been taken apart in the past I would start by carefully disassembling it again, inspect for issues like past repairs and damaged flex connectors and cables. If all is OK reassemble and re test. You may be seeing a poorly connected microbolometer cable but I have not seen these symptoms before except on a stuck FFC flag core and a core that was set up for high temperature imaging. Have you tried moving a soldering iron across its field of view ? I know the measurement is stated as 70F but it is worth checking whether the microbolomter is blind or just in a ‘low sensitivity-high temperature’ bias state.

It could be a fault in the pixel bias area of the core.

The worst case scenario is a microbolometer core that has come up to atmosphere after a vacuum leak. Bill_W described that failure mode to me and it causes massive loss of sensitivity in the microbolometer.

Diagnostics on these cores through direct connections can be a PITA without inter-board connection extensions. I would start by checking all power rails in the core and then look for the video output from the microbolomter. Monitor the output as a hand and soldering iron is passed across the field of view. If no change is seen at the output of the microbolometer then I would move on to the various bias inputs etc. If there is no image signal from the microbolometer you can effectively discount an image processing stage issue and focus on what makes the microbolometer function.

Fraser

[Fraser]

I have just had another thought. What is the history of this Tau core ? Some that were used in HS series cameras were prone to crashing. It was a fault that produced no image but they could be accessed via the communications port. FLIR produced a reset routine and firmware update to correct this issue. That update was available for download from the FLIR support pages but that was a long time ago.

Fraser

[ivaylo]

Thanks, Fraser! It came from a FC-309S security camera. That was actually the reason for disassembly in the past - replace whatever web interface board it came with with the VPC board it uses now plus replacing the lens holder in front (both original Flir parts I got from eBay). Remember admiring how well it was built when I opened it, must have jinxed it (could also static have zapped the sensor or something?) Otherwise this unit hasn't been abused, I do take care of my stuff (it was also supposedly refurbished by Flir before coming to me).

I did take apart and put back together again yesterday. Don't like that fat rigid ribbon cable from the sensor board, too rigid for the angle it's bent under. It does go in, but can't be 100% sure everything reliably connects. Tried using its original security camera interface with the same result - digital communication fine, video test pattern fine, no image. Edit: I did wave hot things in front of it too - it doesn't register.

Where do I get the sensor pin layout from? I do find the Tau teardown thread here, but I only noticed a block diagram or something?

Thanks again...

[1os]

to understand the damage, I would like to know if the image is displayed when loading first with the label flir, a green square appears before the shutter flag is triggered.

[1os]

a picture similar to this one

[ivaylo]

Yes, for shorter than a second. Same image as yours but on a white background. Then I hear the first click. What is the significance of this? Thx...

[johnelot]

sorry if this has been said but could the shutter be stuck across the sensor ?  on the pic it looks like a diagonal line from the top that might be the edge of a shutter

[ivaylo]

sorry if this has been said but could the shutter be stuck across the sensor ?  on the pic it looks like a diagonal line from the top that might be the edge of a shutter

Nope, this was LesioQ's guess as well. I just checked again, I see the sensor when I remove the lens with the shutter up in the corner (luckily it's easy to check on these cameras). I thought dirty or damaged lens, wrong focus distance, etc. but hopefully I eliminated those. Thanks for your input!

[1os]

pay attention to this chip .this is the flash memory that is used when loading the camera usually when it is broken the camera cannot output an image

[1os]

the same chip is used in the old iPhone

[ivaylo]

Got it, thanks. Will go read on it. Any tips how to diagnose without desoldering highly appreciated...