MicroHacker Forum

Hey guys!

Today i got a CRM module onto my desk i try to figure out whats wrong with it.

I've bought it as faulty and my first thing was to put it into my repair workshop setup, i've made on my desk. This setup includes a powersource, a PC with a CAN-card to generate arbitrary messages, the plugs to connect the module, a genuine camera and to view the video a small monitor with RCA video-in. First i tested the setup with an working module, applied power and as soon as i put "IGN" and "REVGEAR" CAN-messages to the bus, i see a picture in the small monitor.

Then i switched the CMR modules and no picture is seen. The current consumption is about 350 mA (module only, no CAM connected) when sending CAN messages, which is about 100 mA higher than with the other module. Well, at least the module seems to work "somehow" as it reacts on CAN messages.

Next i connected an ELM to this setup and it detects the module. Nice, so the module is able to communicate via CAN. I then look if it may be corrupted by an aborted update, but UCDS tells that it has the appropriate software so i cannot update: (maybe i should try this with an forced/dead module update?)

Then i read DTC and there are some. Ok, i expected this, because the module is out of car and cannot communicate with the BCM and PAM. What catches my attention is the U200D which tells there is a short to +12V for the "Output Power A". The other DTC is complaining about a missing signal on "Video Input A". So maybe this means that the interface to the camera itself is faulty and therefore it cannot communicate with it, and for shure cannot show a picture?!

The CAM is connected through this Molex port of the module: I measure about 10V on the upper left pin on both modules, the working one and the defective one. Also i can see on the Oscilloscope that there is a binary communication on one pin, also on both modules.

I then checked the power-regulators, and the big one on the lower side (15) is getting hot fast on the defective module: and stays normal on the working one.

So i suspect something is drawing much more power than it should, and therefore overheats the regulator. Maybe something short to ground.

I've switched over to a working module and read DTCs. This shows that most of them are also present there. Even the one is suspected to be related to the issue: "B100E: Video Input "A" Failure Type 0x87: Missing Message".

The only DTC which is solely on the faulty module is this one here: I've then tried to get what parts belong together to achieve a specific funktionality. It's most likely that parts which are close together also are relates to another. So (1) (the graphics-processor chip) shurely uses the RAM (2) and Flash (9) for executing his image processing software.
Part (3) is an LVDS-Decoder for the camera-signal and (10) to build the video input part for the camera.
Was causes my attraction then was the high-side-switch (4) which switches the powersupply for the camera on/off. As for the datasheet it is capable of detecting failures, like short-to-GND, short-to-VCC, and such. I guess is it driven by (8) the module-controller which also containts the CAN message handler.

I've checked the operation of the (4), but it seems identical with the working board. I also desoldered it, but the error keeps the same. I also desoldered the part from the working module to see if that gives any additional DTCs, but is doesn't. It seems that even if that part could be monitored, it isn't.

So i'm pretty shure that this is the one i need to find: Maybe this is the voltage-regulator "LTC3407" part (15) which is monitored somehow. It has no internal monitoring or failure detection functionality, so it may be done externally.

Well... you have a working one so you can compare readings of VCC on different places and also ressistance (diode mode) on pins of ICs when the boards are not powered.