Products > Thermal Imaging

Identifying an imaging failure mechanism in AMBER Sentinel infrared camera

(1/2) > >>

DoubleG79:
I have recently acquired an operational AMBER Sentinel camera.



However, the image quality is severely degraded when contrast is increased.
This can be viewed when imaging a blank background:



Any clues to a possible reason why the image fails at contrast? While low contrast can be used to get images the camera constantly applies an automatic correction resulting in a high, albeit smooth background level. For example a low contrast image of a coffee cup:



Upon opening the camera up, there appears to be nothing damaged inside. The surfaces on either side of the NUC shutter appear to be clean and free of any residue.
Upon a NUC, the shutter is used however this does not seem to fix the issue.



It is not clear how to open the camera optics, especially without damaging it further. Perhaps there is something akin to residue or mold on either the sensor or lens? Does this kind of image ring any bells as a clear cause of something?

Any help restoring the camera is greatly appreciated, but if the fix appears to be too involved, perhaps best to leave this camera as the working relic it currently is!

Fraser:
It looks to me like a camera that has lost its NUC table and dead pixel map. Can you locate the memory battery in the unit ? I suspect that you will find that it’s voltage has dropped below that needed to maintain the NUC and configuration memory. With the Radiance 1 this is corrected using the camera controller utility that creates a new NUC table and Dead Pixel map. I do not have a user manual for the Microbolometer based Sentinel camera so cannot advise how you can recreate the required data for normal operation. Sadly such configuration and calibration may have been a ‘return to service agent’ event.

Fraser

Fraser:
I just looked up the Sentinel specifications on the old Amber Infrared site.

https://web.archive.org/web/19980211073715/http://www.amber-infrared.com/products/sentinel.html

I see that Amber state a “Package life” of 10 years. This suggests to me the expected life of a common Lithium memory maintenance battery. The Lithium battery often lasts longer than 10 years, but expecting it to last 25+ years is asking a lot of it.

The specifications also state that the microbolometer is thermoelectrically temperature stabilised. It would be worth checking the current draw of the camera at first power on to see if there is the classic higher current draw at start and gradual decrease as the Peltier module gets the microbolometer to the correct operating temperature. You can also monitor the power feed to the Peltier module within the microbolometer capsule. There should be a TEC somewhere on one of those PCB’s.

Fraser

hap2001:
I've never seen a TIC store it's calibration relying on battery, it's such an awful idea.
Do many old cameras do that?

Fraser:
Thankfully, very few cameras store NUC and other crucial data in battery backed memory. I can only think of Amber and a Inframetrics (PM series) cameras being examples of such. Science cameras could use battery backed memory for calibration as they usually include a calibration utility so that the user can carry out their own calibration updates after a battery change. Why place the crucial data in battery backed memory rather than flash ? I do not know. The Amber Radiance 1 uses a PCMCIA linear flash card for its OS, yet stores calibration data in battery backed memory. That said, it uses a large BR123 Lithium Cell to power that volatile memory for many years.

I do not know whether the Sentinel continued Ambers tradition of storing vital calibration data in battery backed memory, but it is a possibility. The Sentinel is a completely different technology to the Cooled Radiance 1 as it is Microbolometer based. That said, the design team for both cameras could have been the same.

We shall have to see whether the OP updates this thread with news of whether a lithium memory backup battery is present in the Sentinel.

A final thought…. If a manufacturer wants to “encourage” a customer equipment refresh at around 10 years…. a predicted 10 year life lithium battery supporting crucial calibration data makes some sense ;)

Fraser

Navigation

[0] Message Index

[#] Next page