Fuel gauge mistery

[grisoni]

Before I changed sending unit, gauge would read pretty accurate until about half full, where it would bounce around for a bit and then go zero.
New unit reads about a quarter with full fuel, 12v going into gauge-7v coming out as well as at sender.
If I take power directly from battery it will read full when connected to sender....

[majicwrench]

I would take unit out and play with it out of the tank. Move it to full, what does gauge read?? Move it to empty etc. I always check em before I stick em in. NIce thing about spiders super easy to yank out. You do need to ground it when out.
"If I take voltage directly from battery" not sure what this means. Are you connecting battery voltage somewhere?? Why??

[grisoni]

Yea, gauge out and 12V applied from battery and a jumper to sender, then it shows correct.
thought it might be wire from gauge to sender, got some resistance but I wouldn't think its enough to read 1/4 only about 30 ohms

[majicwrench]

Still don't understand why you are jumpering things. What does it do w the factory wiring hooked to it?? That is what matters. Hooking 12v to the sending unit (if that is what I sorta understand you are doing) is not a proper way to test anything.

[grisoni]

I am starting to think its the gauge, it reads about 130 ohms across, can't find what it should be.
jumping things trying to eliminate wiring from equation, it could be that I get 12 V but very little amperage and thus can't power the circuit

[azruss]

The 2 terminals in the sending unit are not a power in and power out. the tank is the ground, so one lead is for the gage and the other for the idiot light.

[spider2081]

What you did is very unsafe. Under the right conditions the fuel tank could have exploded!!!

Power for the fuel gauge is applied though a fuse to the gauge where the current is limited to the point where no spark or over heating can occur in the tank. You applied an unlimited unprotected current into the sender in the tank.

As stated the tank and sender need to be grounded to function properly. So you can't test the sender out of the tank unless you ground the senders mounting ring.

With the sender in the tank and tightened down. Disconnect the gauge wire (lt blue/white) from the sender the gauge needle should peg beyond full. Momentarily grounding this wire should make the gauge read empty. This tests the gauge and wiring. If the gauge reads properly doing these checks the problem is the sender, improper positioning during install, or connecting the wires to sender improperly.

Are you sure you positioned the sender in the tank properly so the float can move through its full travel

Are you sure you have the sender and warning light wires on the correct terminals??

[grisoni]

I'm not a complete idiot, at no point did I apply straight battery power to the sending unit. I do understand that it is a variable resistor.(My kid built a transformer that allows me to adjust voltage 0-19, so no fuel tank was endangered. it was safely done on work bench)


I applied +12v to the gauge on the y/b terminal (power IN to gauge on schematic) , ran a wire from pink terminal (power out from gauge) to the sending unit and a jumper to -12V
basically creating the circuit outside the jalopy. And it works as advertised, but not so much installed in the car.
Fuel sender to Tank is grounding good, wire out from gauge to sender reads about 3 ohms, which is a little on the higher side but should not give me a 3/4 lower reading

Some of the older gauges had a heating element inside that expanded and contracted (VW VDO gauges for example) that made the needle move. Newer ones have dual magnets I believe. Maybe its just tired (shows 130 ohms) have not found what resistance it should be

[azruss]

if everything is working on the bench, but not in the car then you may still have a grounding issue. I would jumper from the tank to a good chassis ground and see if that solves the issue.

[spider2081]

Thanks for the explanation.
If I take power directly from battery it will read full when connected to sender....
This sentence concerned me.

The tank ground is often a problem but I think a poor ground makes the gauge read higher than what it should.

If you have everything connected by your schematic in the car, and then disconnect the pink wire from the gauge or the light blue/yellow from the sender what does the gauge indicate??

[grisoni]

It indicates Zero, that's what is dumbfounding me.
I will try to find another gauge and see what it says.

The strange thing is that hooked up to the car the pink wire is down to seven volts,
hooked to the bench test it's 12, I guess it could be an internal fault that feeds to one of the other pins when connected

[majicwrench]

IF gauge works on bench but not on car, it's not the gauge at fault.
Just grabbed a 12ft jumper lead, w a fuseholder and two aligator clip, it had .4 ohms resistance. 3 ohms is a lot.

[spider2081]

When the pink wire is disconnected the needle should point beyond full. For the gauge to read 0 in the car with the pink wire disconnected and 1/4 connected does not make sense to me.

On my 1981 gauge.

Red: low fuel warning light to sender
yellow/black: low fuel warning light and has a short jumper to terminal B terminal B is gauge and low fuel light power in.
pink: terminal S Becomes light blue/yellow at gauge connector and sender
White/ black: is gauge ground and internal dash light ground, connects to base of low fuel warning light
White: dash light

If the pink wire terminal is shorting to case nut next to it , that could cause your problem
The gauge ground needs to be connected for the gauge to indicate properly.
Does the panel (dash) light in gauge work?
Does the low fuel light in gauge work?

[spider2081]

I am sorry to say I have been giving you the wrong info. I read this in the Fiat Electrical Diagnostic Manual:
Check it's sender by disconnecting it. If its pointer goes to maximum, the sender is probably broken. That statement is what I was referencing in my posts.
After my last post I was talking with another Fiat friend and he questioned if that is correct.
We both went to our cars and found the following. My tank is full and the gauge reads full. Disconnect the forward wire on sender light blue/yellow and the gauge goes to less than empty. ( you stated 0 which would be correct)
Ground the wire and the gauge reads more than full.
He checked his car and found the same.
This is opposite of the statement in the diagnostic manual.
Again I am sorry for adding to your problem.

Please let us know what you find

[grisoni]

Well the dang thing is working now, not sure why. Tried another gauge from my donor mobile with the same 1/4 tank indication, and when I put the original back..blammo it worked...gotta love that. we'll see how long it lasts.
Now my turn signal just quit tho