Posting again about my fuel guage, as the info has changed a bit.

Original problem: when the car is warmed up, the fuel guage dies.

Rather than pulling the back floor to get at the sensor, I pulled the instrument pod, to start by testing the guage. With battery voltage applied, it shot right up. Right. Good guage.

Next, I connected it from the grey wire on the half-moon shaped connector (the one that goes to the in-tank sensor), to the proper mounting bolt on the guage, and connected the other mounting bolt to battery power. The guage read properly, given the amount of gas in the tank. Everything seems great, right? just put it back together and go.

Wrong. It took me a moment to remember, but I figured in that nasty variable: intermittency. The problem only happens when the car is warm. And at that, it might be a matter of driveing 10 miles, or it might be a matter of driving 40.

So, here's my question: with such a simple circuit, and with the sensor being so far from any source of heat, why the heck might having the car running for a period of time cause the circuit to fail?

-EdM.
--
'90 240DL Wagon 'Lola' -- '72 1800ES 'Galadriel'