VCCA Home Forums TECHNICAL - 1945 AND EARLIER 1937-1945 Gas gage issue

I'm helping a friend with a gas gage problem on his 36 Chevy--never shows more than a quarter tank, even when full. After replacing the sending unit twice with new ones and new wiring it seems the next step is to have the gage looked at. Any ideas?

Why did you replace the sending units? Is empty to 1/4 working? Did you test while the sender was out of the tank, or with an ohm meter?

I think you should look at the reading with the sender leed removed at the gauge and again with the same terminal grounded. The first should give you a full reading and the second an empty reading. I am basing this test on the 1938 and later gas gauges.

Check the grounding of the sender unit. Then put a wire from the sender body (clean a patch under a mounting screw) and a clean contact on the chassis. The sender doesnt always earth properly through the tank or mountings and this bad contact can cause incorrect reading on the guage.
Tony

Thrasher,
I replaced my 40's complete fuel system and associated electrical components. I couldn't get my fuel gauge to read correctly. I tested the ohms on the sending unit (as stated in the below "Filling Station" link) and found that my new sending unit was working correctly, but fuel gauge was not reading at all. When I moved the sending unit to the "full" position, the gauge would actually move below empty!

So, I sent my fuel gauge core for a replacement to bowtiebits.com. The gauge I received from them did the same thing. I called the technician and he said I have a bad ground where the gauge mounted in the dash panel.
He was correct. The culprit ended up being too much paint on the small panel the fuel gauge mounts to. (I restored my instrument cluster with decals/paint.) After removing the paint between the 2 poles (there's a copper contact between the 2 poles) the gauge worked perfectly.

If the problem points to your gauge, contact these folks and see if they can repair/replace your gauge. Also, as Tony stated in the previous post, a good ground is a must. I actually ran a separate ground wire from my fuel sending unit to the frame.

www.bowtiebits.com

Here is the Filling Station link for fuel gauge/sending unit troubleshooting.

https://www.fillingstation.com/articles/fuelgaugetesting.htm

Good luck,
Curt

My problem is that the dash unit reads 3/4 when the tank is full and read E when there is 3 to 4 gal still in the tank.
Can one bend the sending unit arm to change the reading on the dash unit ??? I would think by bending it down would change the reading in the direction to correct my problem, I realize it would be a trial & error situation but is my thinking correct or is my brain reading low also ????

Dick

The metal gauge retainer that the instrument gauges bolt to behind the instrument panel should not be painted.

Measure the resistance in the sending unit. If zero to 30 ohms then it works as it should. Next measure the position at zero and 30 ohms. If that is a longer distance than to top and bottom of the tank then shortening the arm is necessary. Many of the available sending units are universal and may not have the correct arm length for all tanks.

Thanks for responding. The sending unit is an original one and the arm is not adjustable. That's why I was thinking about bending the arm. Will check the ohm readings.

dick

All your ideas are exactly what my friend and I were hoping for!! We will run with them this weekend and see what happens. Thanks as always for your help. I'll let you know how it goes!

I wanted to give an update since all the help was offered. Checked all the ideas with no luck so now it's time to send the gage in for repair and hopefully that will be the solution.

All the symptoms point to a grounding issue at the gas tank. Newer tanks have a hole drilled in each corner nearest the axle. clean up the area around one of the corner holes (or drill one) and run a wire (about 14") from it to the hole in the rear fender support. Make sure all the surfaces are clean and bare metal only. Use a couple of good lock washers and then tighten every thing well in place. Years of corrosion will isolate a gas tank from its normal grounding points, especially if any restoration work involved repainting the the floor of the trunk, frame, car bottom, gas tank, or the gas tank straps.

Just my to cents, because I have been there done that. All the other advice about grounding the gas gauge is also important.

Good luck, Mike

I agree with Mike. During restoration I painted everything and now I have a problem grounding the gas tank to the frame. I've had to scrape some paint off the tank under one of the straps and scrape the strap to get a ground, but still have some intermittent grounding issues. This past winter I somehow lost my ground connection and didn't know it. When I loose ground the gauge reads a little over a 1/4 when the tank is empty, a problem if you forget how much gas was in the tank last fall.

Brian

The timing of this post was very fortunate for me. I am in the process of diagnosing an erratic fuel gauge reading in my '37 Master coupe (GB).

My conclusion is that I need to replace the cork float. I am using the original sending unit.

Here are the tests I did and the results.
1) When I disconnected the wire at the sending unit the needle in the gauge went to Full +. That is correct because the gauge is reading "infinite" resistance.
2) When I grounded the wire that connects to the sending unit the gauge went to "E". That is correct because the gauge is reading 0 resistance.
3) I removed the sending unit from the tank and tested it with an ohm meter. With the arm in the full up position it read 32 ohms and at the bottom it read 1 ohm. This matches very well with the specs.
4) I connected the wire to the sending unit while it was out of the tank and grounded it. When I moved the arm up and down the gauge responded correctly throughout the range. At the maximum up position the gauge read "F" and at full drop it read "E".

This is where it gets interesting. I reinstalled the sending unit with the tank full of gas. The gauge read about 3/4 to 7/8 full and about 24 ohms. My thinking was that I could compensate for that by bending the float arm. However I did not want to make that adjustment until I saw how the gauge read at lower fuel levels.

Over the next few days I drove the car maybe 30 miles. I noticed that the gauge was reading lower but a lot lower than the gas I was using. On the 3rd day it read 1/4 and on the 4th day almost empty. Plus the gauge barely moves around when I am driving the car.

My conclusion is that even though the cork float looks good it has deteriorated such that it absorbs fuel over time and sinks.

Am I on the right track?

Thanks,

Sounds like you probably need a new float like you suspected.

To verify, once you have the tank drained and out of the car, before you remove the sending unit, turn the tank upside down and check the resistance. See if it will go to 30 ohms.

You might have to shake it a little, because there is probably a brake on the float arm to keep the needle from waving around too much.

This will verify that the "not full" symptom is not due to a bent rod, or the float getting hung up on a tank baffle.

Going by previous discussions on fuel guage issues it wont hurt to put an extra earth wire from a sender unit mounting screw direct to a good earth point on the chassis. You mentioned 24ohm with near full tank so the earth may not be your problem but isnt going to do any damage.
Tony

Yes, do that. Even though you aren't having the issue right now, it will prevent problems in the future, and it is so easy to do with the tank out.

Just wanted to provide an update on my progress with the sending unit for my '37 Master coupe.

The new cork floats seem to be working fine. I say "floats" because the replacement part is really 2 #16 bottle stopper corks from Hobby Lobby ($3.99).

The original float measures 1 1/4" in diameter and is actually 2 cylindrical corks each 1.5" long. Each new float tapers from 1 3/8" to 1 1/8" in diameter over a 1 1/2" length. So I now have the same volume of cork as the original.

Things seem to be operating very well. I have driven about 60 miles over the last few days and the fuel level on the gauge has dropped about 1/4. The coupe has a 17 gallon tank so the numbers are about as accurate as you can expect with any gas gauge from this era.

We'll see how things go as I get closer to empty and do a full refill.

Thanks for all the inputs.

Did you use any shellac or sealer on the cork?

The problem with cork floats is eventually the gasoline and ethanol will deteriorate a natural cork float.

If the float is coated with shellac (Shellac is a resin secreted by the female lac bug, on trees in the forests of India and Thailand. It is processed and sold as dry flakes (pictured) and dissolved in ethanol to make liquid shellac, which is used as a brush-on colorant, food glaze and wood finish.). It is not impervious to ethanol.

I have found that even when coated (several layers) with gas tank sealer cork floats first saturate with liquid and then deteriorate. The Filling Station sells coated cork floats "FUEL SENDER FLOAT CORK - Alcohol Resistant FS-336A". I have one sender with those floats installed. Will see if they work???? I prefer to use "FUEL SENDER FLOAT Brass SA-99" even though they require modification of the sender wire. Have not had a problem with them. Model A part suppliers have a plastic float to replace the cork. I have not used one of them yet.

Great additional comments!

I had similar thoughts when I installed the new corks. I agree with Chipper aka the "Chemist" (at least to me) that shellac would be of little if any value.

I also thought about gas tank sealer but did not have immediate access to any. Plus my father came out in me and I was not going to spend all the money to buy a quart when I only need a little bit! I now know where I can get that little bit I need if I want to go that route.

I agree that the real long term solution is either a brass or plastic float (one made to work in ethanol fuels).

We'll see how this works over the next few months.

I have read here and elsewhere that cork floats need to be sealed. On the other hand, cork is a closed-cell material, and that suggests it shouldn't need to be.

As you say, the original sealer is generally known to be shellac. Some modern substitutes I have heard of are styrene, or gas tank sealer.

When I overhauled the sender in my 1936 Pontiac, there was no sign of shellac on the floats. I assume it was eaten by alcohol in the gas. At the same time I was doing this, I had a couple of brand new sealed cork floats laying on my bench that belonged to a friend who was restoring a Buick. Those came from Bobs Automoblia.

Just for fun, I took the old original bare cork floats from the Pontiac, weighted them and sunk them in a coffee can full of modern ethanol-laced gasoline. A couple of days later I compared the buoyancy to the new sealed cork floats. One of the old ones was slightly more buoyant, and one slightly less. The net buoyancy was about the same.

I put those old unsealed cork floats back in. They have been working fine for 6 months now.

I will not go into the chemistry and physics involved. Suffice it that there is a rate that any gas or liquid will pass through any solid material. Oxygen molecules will go through steel! Thankfully the rate is extremely slow. However that is not the case for hydrocarbons (read gasoline) or alcohols (read ethanol) and cellulose (read cork). The rate is significantly faster. A barrier material (read coating or sealer) will reduce the rate. Nuf for now.

Gas actually seeped through the cast iron bowl of a Carter W-1. When new the inside of the bowl had a sealer applied.