The Stovebolt Site Forums The Shop Area The Electrical Bay Another Fuel Gauge Problem

Fuel gauge works perfectly. Sending unit works perfectly. Wiring from the tank to the gauge works perfectly.

How do I know this? I connected the sending unit with alligator clips to the two wires at the tank before installing the sending unit in the tank. Dash gauge went up and down with the manual movement of the sending unit arm and float. Oh ya, float works perfectly , too. I submerged it in gasoline overnight to make sure it doesn't have holes in it.

Once installed in the tank, the gauge stays at Empty with no movement.

What do I need to do?

1)Put the battery cables back on.
2)Turn the key in the ignition switch to "ON" position.

an open circuit on tank ground would make the needle rise to full
an open wire on the sender from the tank to the gauge would make the needle rise to full
if your needle stays at empty with power on
your gauge in the dash probably has no power
52Carl maybe onto something...
-s

2MT,

You missed the part where the dash gauge works when connected to the sending unit outside the tank with key on. Arm goes up and down and immediately registers on the dash gauge.

Once installed in the tank and connected to the same two wires as before, gauge stays at empty with key on.

Most likely once you install it in the tank your are fully grounded the circuit somehow. . Open circuit would be full if I recall correctly. Ohm meter your wires while it’s still in the tank

I c
Okay, Power off, then ohm out the sending wire
It’s likely shorted to ground yes
Then disconnect it and find if the short is at tank end
Or if the gauge end got grounded in the cluster

Also don’t forget to make sure there is gas in the tank
Maybe it’s reading empty correctly?

Float rubbing on tank wall can also keep it at empty position
-s

You said two wires. Where do they go most senders use one wire to gauge and use metal to tank as ground.
Do you have a picture of your sender top

One is the ground on one of the sending unit screws and the other is the hot wire to the sending unit terminal.

We had the same problem with my 50. Drove us crazy. It ended up being a short on the center wire stud, positive that would short out when screwed down. Worked perfect outside of the tank.Sending unit was bought from Classic a few years ago. Bought another and bingo, worked perfect. I bet we had that first sending unit out 15 times.

Could be a short somewhere between the sending unit and the gauge. Easy to nick a wire while working on other things. I remember nicking a wire while installing upholstery.

But it works with the same wiring and gauge when out of the tank.

Ohm out the sending unit before and after installing in to
Tank
Better yet
Connect the ohm meter up and watch it as you move the sender into place and start installing the screws
And if it’s going weird then you know precisely at what point it’s getting shorted out and can further trouble shoot
This method works better if there is enough fuel to start floating the sender before the short happens
Because it will read empty with the float dropped down
-s

Otto,
How much effort does it require for you to manually move the float arm up and down? Should be quite loose. If not, the float won't have enough buoyancy lift the arm. Just spit balling here.

Yes, it's very loose and free moving.

Otto.
Do these steps
Power on

Disconnect sending wire at sending unit
Gauge Should go to full on gauge and beyond.
Put wire on metal should go to empty
Test with ohm meter from sending post to ground. It should have some resistance if it’s zero something is grounding the sending wire at sending unit side

Watch this. Substitute 30 for 90 ohms when listening
Fuel Gauge & Sending Unit Troubleshooting

He never confirmed there was fuel in the tank yet
Everything working correctly is still a possibility
-s

Yes, the a tank is 3/4 full.

If you read the threads he spells it all out. I really looks like he is getting a full ground on sending wire at sending unit. When it gets installed.

Oh kk I missed that part
Yeah it seems the issue is at the tank end agreed
The ohm meter is you friend and best bet now
Best of luck
-s

Thanks. I'm going to try and get at it tomorrow if I have time.


Do you mean that the wire on the sending unit stud is grounding out somehow?

I think one detail people are suggesting is that somehow the sending stud on the tank unit is grounding out when installed
Like maybe it has a manufacturing defect or some how part of it on the underside is contacting the tank or something
The idea being that it is the most likely area you should focus on now

Ohm out the sending unit sending wire stud to the sending unit body (ground) with it outside of the tank
And move the float arm up and down to confirm it works properly
Then with the meter connected up to the sending unit sending wire post place but dont screw the sender into the tank and repeat the measurement while giggling the unit

At this point the ohm meter should be reading about 20-25 ohms perhaps with 3/4 full of fuel
If you see that ohm meter drop to 0 ohms intermittently at this point you will have confirmed the sending unit is somehow shorting out when it’s being placed into the tank hole

The next step would be to carefully examine the sending unit bottom and top side and try to figure out where its shorting

Can you post a photo of all sides of the sending unit?
There is more than one style and that might help us help you
-s

Personally I would be hesitant to test for a shorted out sending unit by putting it into a tank of gas while there is power to it.
A spark at the moment the sending unit contacts the top of the tank might make for very bad day.
Might be a better idea to test for a short between the post on the sending unit and the top flange of it where the bolts go through on the bench without any power to it with an ohmmeter.

An ohmmeter is not equivalent to circuit powered on
It is well under the intrinsically safe threshold for deliverable power during the test
There is a whole industry and research field behind this detail
You will not be able to generate a spark or ignition event with a good meter on ohm mode during this type of test
Intrinsic safety is a real thing and the tools are safe to use in this situation
-s

2ManyTrucks, I don't know if your post was a response to my post above, regardless of whether it was or not, I am having difficulty understanding what you are attempting to convey.

"Intrisically Safe" is a technical term.

Putting the sending unit in the fuel tank and checking for it being shorted by using an ohm-meter is safe because ohm- (and multi-) meters are designed to be "intrisically safe".

In this context, that means that they do not use a high enough voltage or current to spark under any circumstances.

There is a whole field of study and set of design principles dedicated to how to make intrinsically safe tools for various applications.

Edit: there are almost certainly more criteria to being instrisically safe that just non-sparking (such as low current to avoid heat production, etc). My intent was to address your specific concern, not give a comprehensive list of applicable factors.

It sounded like there was safety concern around possible electrical discharge

I was just trying to be clear about using an ohm meter won’t be dangerous in giggling bits around to trouble shoot a intermittent short in this scenario
It should be impossible to make an ignition spark based on electrical discharge


Other sources of sparks would be more likely
For instance I was disconnecting my under truck fuel line the other day (like a few months ago now)
And of all the places to ever see an actual spark fly off a wrench
My tiny open end box wrench slipped and I saw a big yellow spark fly 6 inches from the tool just as I was loosening the fuel line hold down clamp on the frame, I couldn’t help but just have a nervous laugh out loud.
-s

I disconnected the wire to the sending unit stud and connected a jumper wire to the stud and the other end to the ring connector. Fuel gauge immediately registered accurately.

So, I'll either leave that stupid jumper in place (not really) or I'll scrub up the stud and ring with a brass brush and see if that helps.

Just for giggles, throw a random ring terminal on the stud and tighted it down, then check it with the jumper.

Another bolter reported an issue where tightening the ring terminal nut on the sender caused an internal short due to a defective sender. You could be seeing the same thing.

I wire brushed the threaded stud and the ring connector. I also used emery cloth on both.

I then installed an external star washer over the stud, prior to installing the ring connector. It just sat on the base of the stud before. Then I installed the ring connector secured by it''s nut with a captive star washer. All is working now.

For future reference, is an "always empty" gauge reading usually attributed to the hot wire side of the circuit and that an "always full" gauge reading can usually be attributed to a ground issue? This assumes that the gauge and sender are known to be good.

Oh I think I understand why we were all confused now
I thought the only way to get empty reading was a full short or close to it to simulate the sender with float down
But your saying if the sending wire is open circuit completely then that also registers as empty at the dash
Maybe because 0 current flows and there is no magnetic force to raise the needle to full position?

Thanks for updating us on your progress
-s

What you guys are missing is the gas gauge actually has two coils and three connections. A positive that goes to the battery, a ground and the sending unit wire. The coils are acting against each other. One coil is pulling it full while the other (attached to the sending unit) is trying to pull it empty. If you ground the sending unit wire it will power the coil and pull the needle empty. Lift, it will go full. When the sending unit float is down, it reads zero ohms, so there is no resistance (full voltage on coil), pulling the needle empty. When the sending unit float is up, it reads 30 ohms which is enough resistance to negate the effect of the sending unit coil allowing the other coil to pull the needle full.

Credit for the attached diagram is by Jon G.