Gas Gauge Problem

[toddbailey]

Hi all

I'm a bit stumped here.

in my 83 944, My gas gauge always reads full.
I've tested the gauge and the float and both work as expected.
the gauge when connected to a pot will read between empty and full depending on resistance usually 0 to 80 ohms.

the float also will read between 4 and 80 ohms depending on the float position
so it too works as expected.

the fun begins once every thing is reconnected, the gauge reads full when connected and empty when disconnected. Regardless of where the float is located.

I've cleaned the float, applied de ox-it and also checked for any shorts. Nothing out of the ordinary

Any ideas what I've missed?

[grendiers]

Sounds like a grounding problem. Those hard to reach ground points behind the dash might be the ticket. Or the gauge itself, the circuitry behind it is faulty.

[toddbailey]

but the gauge works when test equipment is connected or when I ground the G lead the gauge goes from empty to full. I more inclined to suspect the float assy.

[Dano944]

Welcome to the head-scratch of 1983 models. That’s what I have and there are many gremlins inside my car. I have similar problems with my gas guage. I’ve had the float cleaned, but my gas gauge bounces around, go low then goes high, and after driving a bit, it seems to settle down with the proper reading. Maybe my gas tank needs to build pressure to get the float to read properly. Weird problems like this are what keeps me from taking long drives to the middle of nowhere.

[originalowner]

What is the resistance when you measure the sender while it is in the tank? Has to be low to make the fuel gauge read full.

Is the float physically stuck at the top when you reassemble it into the surrounding tube?
Is the sender wire shorting when you tighten the assembly into the tank?

[Monkey Wrench]

if you suspect a wiring fault aside from the gauge or sender, could you substitute a run of wire that's out in the open as a test? I dont think pressure inside the tank is related. hard to see if the float is physically getting stuck but the resistance should prove that. I'm wondering if that sensing wire is being grounded through body contact somehow.. maybe rubbed through somewhere due to vibration, or similar?