A little more info:
I got analog out set up on the camera now. As per the documentation (D8X3CUserManual attached earlier in this thread) analog video out is pin 1 of the 50 pin header on the board. A little continuity testing showed that that pin is connected to the following traces, picture 1 shows where it is on the camera module and picture 2 shows it on the unit main board.
I spliced in so that the analog video connector in the rear cable is connected direct to the camera board instead of the main board, circumventing the main board entirely.
This in interesting, because it appears the analog out directly from the camera is a much higher quality signal than what is output over the ethernet port. I wonder if the digital capture being done by the ethernet board is not taking the full 14bit depth or some other bottle neck, it is less sharp and has less dynamic range and sensitivity than the analog channel.
One interesting next direction to go would be to directly control the camera module. This can be done with pins 3-7, and should allow full control of the camera without using the ICI main board at all. As is, if you try to run the module without connecting it to the main board it only NUCs once at startup, which quickly makes it not a great image. Very interesting that the analog video looks so much better, definitely points to the fact that a custom data capture 'backpack' could squeeze much more performance out of this camera.
To add to this thread on different variants of the ICI FM 640 P...
I snagged one off eBay for ~$300, still a good price IMO for what's inside. It was absolutely caked with dust on the inside of the housing and must've been used for a while or in an industrial setting. It has a variation of the visible camera I've never seen before in teardowns (although noted as being on the original model by radziorr). A HIKVision (DS-2CD7027FWD) camera equipped with electronic focus control.
The core is the
DLD640 and the backpack/core combo is the
Dali-Tech D843NT, which is Dali's own module. I'm hoping that this will provide better implementation possibilities if I can get Dali to provide core documentation, plus it's got a much smaller footprint than the later ICI "Plus" models which had a separate processing board.
The login password is different when connecting via IP 192.168.1.102 than the newer modules noted in this thread. Exact same interface otherwise.
Username:
adminPassword:
adminOne thing I did notice is this whole module gets rather warm, which I'm sure isn't great for image sensitivity. I used another thermal camera to image the Dali module (I recognize the irony) and it seems the Altera Cyclone is the main heat source. It would be interesting to see if I can add some heatsinking or cooling to remove this heat. After several minutes the whole core housing becomes heated. I do have a stirling crycooler at my disposal, the Superlink Hybrid 98, which may call for some experimentation.
Bobertawsome,
Thank you for sharing details of your cameras internal modules. Very interesting
Regarding localised heating of the thermal camera module, this is not uncommon and not normally a problem. The Microbolometer normally operates in thermal equilibrium at around +32 Degrees Celsius. The die temperature is monitored by the electronics package. Cryogenic cooling of a microbolometer is not normally needed or recommended. If a component in the electronics package is generating too much localised heat it is often dealt with by adding forced air cooling to the enclosure and a small finned heatsink to the offending component. Another option is to use a copper heat pipe or solid copper ‘finger’ attached to the offending component and a large heat sink or thermal reservoir. As I say, heat from the electronics package is not normally a problem in larger cores where air can circulate. It is the miniature imaging cores with hot processors in close proximity to the rear of the microbolometer that can cause thermal gradients and other undesirable effects on the image output.
Fraser
Thanks for the additional info
Looking at the product page I believe the camera core is TEC temp controlled, so it would make sense that the die is probably a lower temp and the additional heat is sinking off the hot side of the Peltier.
If a Peltier module is being used that is good news as it greatly improves microbolometer stability. The down side of the Peltier temperature stabiliser is that the core will be quite power hungry, especially at first switch on. You will see the Peltier stabilising the microbolometer by monitoring the current draw when the core is switched on. The current will drop significantly once operating temperature is reached. I personally like cores that have temperature stabilisation so it is good that yours has this feature
Fraser
Several nice thermal cameras were rehomed in private sales. I have multiple examples of some cameras so sold some excess inventory
I still buy the odd thermal camera if it is interesting enough and the price is right
(it has to be inexpensive!) I no longer buy cameras purely for teardown and analysis posts on this forum though as my funds are very limited. Times have changed from when I used to buy many cameras to learn about the designs and technology within. I have now accumulated enough technical knowledge to repair both old and new cameras
I have also moved onto repairing specialist fully rugged laptops, as used by the military, law enforcement and hazardous environment teams. That new interest distracts me a little from the thermal imaging scene
That said, a nice high-end industrial thermal imaging camera connected to a mint Dell Latitude Rugged 7414 laptop is a thing of beauty to me
