MicroHacker Forum

I ever wondered why nobody had looked into the BCM firmware to find out where it set's the level for the current detection of some lights (e.g. headlights). There must be some value as they can switched between Xenon (HID) and normal Halogen. If we would put some LEDs instead of glowlamps then these value just needs some tuning to make the BCM happy, not to emitt DTCs.

Halogen lamps are switched on using PWM... there are some MOSFETs involved.
For XENON headlights the power supply is "full on", using the same MOSFETs.

From what I know, it only needs a resistor to emulate the closed circuit.

Usually LED lights don't have any resistance.. or it is high enough to be "detected" as not connected.

The solution would be to kill the entire function that monitors this resistance.

How the lamp is driven has nothing to do with the way they are surveilled. In fact LEDs are also PWM driven and of course they have a resistance. Only the power profile is different. PWM is most often an easy way to prevent overvoltage problems, so some kind of regulation. And of course for dimming.

Checking a lamp for function can be donw "cold" and "warm". Cold means, that before the lamp is powered on a circuit checks with a low voltage if current can flow through the connections (this is why LEDs sometimes glow, when used in a place where conventional bulbs are). It is some sort of easy resistance measurement. The warm check uses a shunt (a very low resistor in series of the lamp) to make some sort of simple current measurement. And there are also short-checks (short between, short to GND, short to BAT) which can be done with simple port-IO (if the output can't be pulled to GND, it is a short to BAT and vice versa).

I'm very shure that inside the firmware-surveillance-routines, there are measurement values which represents the resistance and current of the lamp to be driven. I won't kill the function at all, because you then loose monitoring. Better to tune it in a way, it can handle the lower load.

Is this the beginning of a Disassemble Tutorial of the BCM? Would even make a lot of fun, since i bet that firmware can be extracted the same way as from other modules