Pontiac renovations, part 8.

I refitted the now-clean headlight switch back to the dash. Unfortunately now the rest of the dash dials look dirty in comparison. I shall have to remove the heater controls and clean them also.

Headlight pull-switch (in second position)

Electrical work continued with gauges. The fuel gauge was similarly burned up inside, along with the temperature gauge. Measuring the diameter of the wire on the fuel gauge coils showed it was the same diameter as the new wire I had bought.

Measuring wire gauge with micrometer.

The mounting bucket was also showing signs of age; the paint was flaking off from heat and there were scorch-marks one the paint also. Cleaned down the original paint (which came off very easily with carburetor cleaner).

Clean mounting bucket.

Counted the number of turns on the fuel gauge and re-wound the spool carefully. Testing on a battery with 6V converter in place- I decioded to use 6V because the wire gauge was the same as original, lower current through the sender in the tank and everything could be set up identically to how it was removed.

Calibrated the irons to the correct resistance values -0, 15 and 30 Ohms. 15 Ohms showing middle-of-the-range.

Mid-way calibration mark- 15 Ohms.

I then began to readdress the temperature gauge. Operating it at 6 Volts showed there was insufficient magnetic flux in the middle of the needle’s range to hold it in place over the friction of the pivot.

Six volt test.

Operating it at 12V as originally intended solved this but I had wound the coils for the original spec sender unit, which is no longer available. The original sender unit had a very low range of resistance values for temperature, 100F being about 300 Ohms. I purchased a (slightly) more modern sender, a Wells TU-4. It has approximately twice the resistance per temperature, is physically the same size and is more readily available, as it fits mid 50’s Chevy trucks.

Sender unit in hot water, with thermometer.

I decided to abandon attempting to calibrate the gauge while connected and instead started plotting temperature versus resistance on paper.

Wells TU-4 resistance curve.

The gauge is wired in a Y-configuration- I had missed this previously. One coil is directly across the supply voltage, the other is across supply through the sender. I decided that I should be able to add a resistance across the other coil and reduce the pull of the magnetic field, allowing the higher resistance (and lower current) of the modulated coil to pull the needle off COLD sooner.

Bench testing to determine biasing resistance.

Initial testing proved successful, with the gauge capable of being adjusted to be accurate from the lower end of the range all the way up to maximum. It over-reads a little between 100-130F but is accurate at 180F and 230F. I decided knowing if the engine was too hot was more important to know if it was cold.

Resistance in place, heat-gun testing with sender unit.

Reassembled into the frame and tested the gauge again using the frame’s ground, as occurs in the car. Successful testing with good results.

Gauge mounted in dash.

I turned my attention to the paint on the dash, which was scratched and thin in places from years of polish and being handled. I started by stripping all the gauges out and taking the old paint off with a brass brush.

Paint removal.

The dash was then masked up, painted, reassembled and tested.

Ready to reinstall, looking very tidy.

Leave a Reply