I figured what the hell and added a new ground. The is now 0.051V with the lights on and 0.020V with the lights off. Seems strange that the new ground doesn't pull the circuit to zero but reduces the ground voltage by a factor of 10. Just a coincidence?
The 51mV and 20mV values are good. Zero voltage drop can't be reached, unless
you have access to superconductors. Just an example: the voltage drop on a 1.6ft
copper wire (with 1 sq mm cross section) is 35 to 45mV between the 14.4V power
source and the consumer while the current is 1A. Thinner wires have higher resistance
and voltage drop under the same circumstances.
If you want to do some more checks, here are my ideas:
- With ignition on check the voltage (with and without lights) between the red and black wires towards the speedo. The expected voltage is greater than 12V in both cases.
- There might be defective soldering joints in the speed signal preprocessing unit.
This unit can be found on the back of the RPM counter under a white, box shaped plastic
cover. It has a five pin connector. This unit should be inspected as well.
- Supply problem of the closed loop motor controller circuit inside the speedometer
(this can't be accessed from outside, it requires to disassemble the speedo)