Originally posted by Jon [in CT]
I wonder whether your "accuracy" problem might be due to "offset voltages."
The card is grounded at the same point the analog signal output from the meter is grounded.
The problem lies in the sophistication of the inverter/amp card. It's basically a rack of Zener diodes and a voltage divider network to give the reference voltages. It's output varies greatly on the type of power supply given to the card.
The inverter card as it ships is powered by two 3v lithium batteries in series (making a 6volt supply), although it really only needs 3volts for the zeners to function to give their reference voltages. I first hooked up a variable power supply @ 5volts I have at work, which is basically a 10:1 step down transformer and a full wave bridge rectifier along with a few rheostats to give various desired voltages. That gave me one cal curve based. Then I wired the batteries back up to the power supply, giving me yet another cal curve. Using the 5volt sensor power supply from the ECU's 5volt bus gives me yet another cal curve.
Bleh... I don't want to hijack this thread anymore.
Rob - would you please forward my questions to Klaus