Sorry, I read the power switch pins the other way around. I was measuring garbage
I've also managed to follow most of the dimming circuit, unfortunately with a gap in the middle. Anyway, here are high rez picts:
http://www.hotpen.co.uk/others/panelfront.jpghttp://www.hotpen.co.uk/others/panelreverse.jpgThis is the trace with the voltages (dimmed-non dimmed).
1) 0.8V - 12V … direct entry from the dash switch
2) 1.8V - 12V … after SJ35
3) From here it goes to various clusters of 2 resistors and then feeding the other non-display LEDs. Voltage after the 2 resistors is 0.7V - 3.8V
gap
5) 5V - 5V … after pin 71 on the Motorola MC912D
6) 4.7V - 4.7V … after 4991 resistor (goes to pin 3 power switch input)
7) 12V - 12V … pin 1 power switch output
Edit: this seems to be more complicated than I thought, there is a light sensor in the dash which acts like a switch - in broad daylight it deactivates the dimmer. Maybe that's the problem right there, that panel is stuck on 'daylight' regardless of what signal is receiving from the dash sensor. There's no wire with a voltage difference corresponding to this ambient change in the plugs so the signal might be transmitted through the k-bus wire (networking).
Also, the microprocessor feeding the switch doesn't look like it has a dimming function, I think it's just turning the fan on and off
along with the LEDs by sending voltage to the switch.. When the fan is off, the display goes off too and that is controlled by the microprocessor.