I'm puzzled by the above conversation. My 1965 TR4A fuel gauge reads between 1/2 and 3/4 full when the tank itself is full. My troubleshooting research describes the gauge as a current sensing device, where higher current creates more deflection of the spring holding the indicator needle. The gauge has regulated 10 volts on one side and the current is highest when the other side is at ground. It was recommended that shorting the connector at the sending unit to ground would either send the needle all the way up, or a continuity problem would cause it to read low. I did that and found it to be true and my continuity is good, thus a FULL indication. That means the resistance of the sending unit when the tank is full should be close to zero. Resistance goes up as the fuel level goes down, so current goes down and the gauge indication goes down. I have a new sending unit from Moss Motors, pn 763-020. I tested the resistance of the sending unit with the tank about 3/4 full and found 47 ohms. The gauge read 1/2 full. I then placed a potentiometer in parallel with the sending unit and adjusted it to 120 ohms, which allowed the needle to increase to a reading of 3/4. (Putting the resistances in parallel causes the net resistance to be lower, so higher current and a higher reading.) I can use this strategy to calibrate the full tank reading to be FULL. However, I don't yet know how an empty tank would show on the gauge when FULL is reading correctly. It would show higher than without the parallel resistance in place, but I don't know how much yet. I'm a bit reluctant to test this until I run out of gas if the EMPTY reading is too high! So, I'm curious to learn if anyone out there has had a similar experience and, if so, how they ended up.