The Cadillac cameras are basically the standard Raytheon Thermal Eye 300 design modified to suit an automotive application. Sadly the schematics for these and other Raytheon cores were never released or leaked into the public domain. This is typical of most/all thermal cameras and you have to reverse engineer parts or all of the circuits in order to carry out normal fault tracing procedures. I am used to working without schematics so tend to identify key sub systems and check them for signs of normal operation. If I find something unusual, I reverse engineer that part of the circuit and delve deeper with diagnostics. If no obvious issues are apparent I have to reverse engineer each sub system of the camera design and test it for correct functionality. It is hard work and time consuming. I never reverse engineered the Cadillac or Thermal Eye 300 cameras as I did not have a fault to track down. Hence why I cannot help you on the fault finding front.
I can make some generic comment on the Cadillac camera though......
The camera is relatively old, has been exposed to a broad range of temperatures and may have suffered water ingress. Always inspect the case interior and PCB’s for any evidence of water ingress. If found, track down all damage. This sometimes requires the removal components as fluid tends to accumulate under Chips and even smaller components and corrosion occurs over time. The most common corrosion occurs on rails that are carrying DC power rather than data lines. Power rails and chip power pins can get eaten away or may look OK, but have high resistance. Water is pretty deadly to these camera cores if enough gets past the seals. The heat inside the camera and its age can lead to failure of the tantalum capacitors. It does no harm to replace all the tantalum capacitors but ensure you fit the same specification or tears can result when they burn-up ! Just read the data sheet for the types fitted and buy the same surge tolerant types. The motor that drives the chopper wheel is generally reliable but ensure if is not suffering worn out commutator brushes as that can cause RFI and poor motor performance. The motor spindle should be very easy to spin. Any sluggishness or resistance in the shaft rotation points to stuffed oilite bearings. You may feel the magnetic influence on the rotor but that us very different to a sticky shaft feel.. The motor should quickly spin up the chopper wheel to the correct speed and it should be locked-in using a PLL as this chopper wheel is essential to the correct refreshing of the BST sensor after each sensor Row is read. If the chopper wheel drops out of sync, the image will be awful, or non existant. Very poor contrast or no image can be chopper wheel synchronisation related. Chopper wheel speed stability can sometimes be checked by listening to the motor note but a more reliable way is an optical tachometer of an oscilloscope monitoring the chopper wheel IR reflector sensor output that reflects off of a silver dot on the rear of the chopper wheel. A damaged chopper wheel membrane will also degrade the imaging performance. A camera that appears functional but does not produce an image may have a BST sensor issue, such as coming up to atmosphere due to a vacuum seal leak. This can be checked by temporarily installing another (known good) BST sensor from another camera. The display, if it appears will have dead pixels present but if an image appears then you know the original BST sensor has a problem. Clean all its pins and re-install it. If no image appears, it is scrap. The camera system is calibrated to the sensor and the dead pixel map is part of that ‘calibration’. As such you cannot usually rob a good sensor from another camera for permanent use. The image will likely be covered in little black dots (dead pixels).
A lot can be learnt about a BST Raytheon core by monitoring its current draw and chopper wheel at start up. The current draw will initially be quite high, say 800mA as the heater within the BST sensor brings the sensor up to around 32C (it’s nominal operating temp). Within a short period of time the current will drop to say 380mA as the heater drops to idle running current. Whilst the camera is starting, the chopper wheel should spin up to speed, overshoot a little and then settle to a steady speed, producing a constant hum. If the chopper wheel sounds like it’s speed is ramping up and down, it means the camera electronics are not able to lock its RPM and phase to the sensors readout electronics and a very poor, or no image will result. The spiral mask on the membrane must be in exactly the right position as the BST sensor is read-out. It effectively follows the read-out rows and erases them ready for the next scene capture. A small error in the chopper wheels phase with respect to the read-out process is detrimental to the whole systems performance.
The BST core applies certain bias voltages to the BST sensor. If any of these are missing or incorrect, there will be either no image or a very poor image. I cannot provide the pinout for the BST sensor or the voltages that you should expect to see. You can use a known good camera as a reference against which to compare though. All voltages and signals should be similar enough to be of use in diagnostics. All power supplies should be checked for presence and correct level. Again, a known good camera can provide a reference for such checks. Beyond this, you are entering into the digital World with a fast ADC converting the analogue signal stream of the Sensors read out chip into a digital signal that is presented to the digital video processing stage. And then on to the image creation stage with its SRAM and TV system compliance. A microprocessor oversees the cameras operation and detects errors and faults. Sadly I cannot help you with any of the digital stuff as I have not delved into such in these cameras.
You have a number of cameras. I suggest you take the one that works best out of all of them and make a series of voltage and signal readings on key components and power supplies. Use that unit to learn about how the camera should behave when working. Once you know the camera, take a look at the faulty units and compare readings and data. You may well be able to identify the cause of the problem. Robbing parts from the worst condition unit may be an option if there is not too much damage evident within it.
Hope this helps a little
Fraser