Need help, fuel tank sending unit installation

[FastbackDon]

My 71 Mach 1 has shown a full fuel tank for the year I've owned it. I replaced the sending unit 8 months ago, but it's still shown full from day 1. I replace the unit again yesterday, and it's still showing full.
Before installing the new unit, I tested it this time. I plugged in the wire connector (it's a 2 wire system) and set the float level at low, medium and high positions and the gauge read pretty accurately.
When it was installed, the gauge went back to showing a full tank when the tank was totally empty.
It's like there's a secondary ground in the system. Has anyone had this issue, or know of any possible reasons for this? Yes, the round gasket was installed, hopefully correctly. I'm at a loss.
The wiring diagram I have (A Prospero's Garage color diagram) shows a single wire to the tank, and the tank acting as a ground.

 

[midlife]

When you tested the unit before installation, did you use the car's wiring and connector? If so, there's a problem with the installation of the tank and routing of the wire that is shorting the sending line when installed. If you didn't use the car's wiring, then the problem is still in the wiring, most likely a pinched wire somewhere.

 

[FastbackDon]

Midlife,
Yes, I used the stock wiring when testing the sending unit. I attached the connector (only one orientation will work) and let the sending unit hang from the wires. I used a test light connected to the wiring at the sending unit before starting any of this. With the key on and the test light connected to both wires at the fitting at the sending unit, I received a pulsing signal. If the YouTube videos I watched are accurate, that indicated that the wiring was 'good' and the constant voltage unit was good.
I will attempt to check continuity to ground on both wires. I'm 'assuming' one wire comes from the gauge, and one wire is a ground. Not sure why it would matter in the ground wire was pinched or not..

 

[FastbackDon]

After playing around, I hooked up a jumper from one port on the connector to one of the studs on the fuel tank and the gauge worked. I'd still like to get this issue corrected, but this will suffice for now and I don't have to worry about running out of gas because of a faulty gauge.

 

[MikeGriese]

If I am understanding correctly, you ran a jumper from the 2 lead connector to the sender and it worked, but plugging the connector to the sender doesn't work. If this is correct, the female socket in your harness connector has spread so that it doesn't contact the pin on the sender. Clean out the socket as best you can and you can try to crimp it a bit through the shroud with a pair of stout needlenose pliers.

 

[midlife]

If I am understanding correctly, you ran a jumper from the 2 lead connector to the sender and it worked, but plugging the connector to the sender doesn't work. If this is correct, the female socket in your harness connector has spread so that it doesn't contact the pin on the sender. Clean out the socket as best you can and you can try to crimp it a bit through the shroud with a pair of stout needlenose pliers.

If poor contact on one of the pins was the problem, the gauge would read zero (low). His problem is that the gauge reads high. It only reads high when the sending unit is installed, which suggests to me that the sending unit (yes, this is the second one) is shorting out when installed.

 

[Kilgon]

If you can pick up some resistors you can check the gauge. Just put the resistor inline to the gauge at the connector and then to ground. I got a pack of 12 ohms off of amazon for about $5. The following is the read out vs resistance. Just put them in series to get close to the needed ohms for the gauge reading. This will tell you where your problem lies.

12ohms - full - 1 resistor
26 ohms - half - 2 resistors
70 ohms - empty - 6 resistors

 

[FastbackDon]

@MikeGriese, I double checked the sockets with an 8-32 screw (matched the stud on the sending unit), both were good tight fits.

@Kilgon, I will check to see if I have any resistors laying around any more....

I did think of one other thing. I believe the wire from the float arm and potentiometer is connected to the lower stud when the sending unit is installed in the correct orientation. They might be reversed on a stock unit. If anyone has a stock unit, would you mind checking so I can cross this off my list please?
Thanks,
Don

 

[Stanglover]

If poor contact on one of the pins was the problem, the gauge would read zero (low). His problem is that the gauge reads high. It only reads high when the sending unit is installed, which suggests to me that the sending unit (yes, this is the second one) is shorting out when installed.

I could be wrong, memory is fading, but I think there should be a gasket between the sender and the tank and the sender is then help by the locking ring. If he's missing the gasket or has it in the wrong place, wouldn't that cause the short?