You'd need a test sender with a confirmed range of 0 ohms empty, to 90 full, positioned outside the tank, or a variable resistor that can indicate 0-90. Position it well grounded and fed from the gauge ground feed. Attach a ohmmeter from the body of the sender across to ground. The beauty with the 66 (I'm assuming it's the 66?) is I think you can get to the sender feed.
Zero ohms (or close), float at the bottom of its travel, the gage reads empty. Swing the float up until the meter reads 90 Ohms (max up), the gauge reads at or close to the full mark.
A poor ground, poor continuity and more resistance in the ground feed from the gauge to the sender and beyond to the chassis, (higher resistance) could cause the needle to swing past the full mark??
Grounds are always where to start. Or perhaps an incorrect value sender was installed?
I don't think the gauge often fails does it?
Probably a far easier way then this.
__________________
65 Laurentian post, 67 Grande Parisienne 4 door HT.
