1997 3.1 gp fuel sensor question

[rooneystang]

Hello,

I have a 1997 grand prix with a 3.1 in it, and the fuel level gauge reads full and the low fuel light is on all the time.


I have checked continuity on the black and white wire from the sending unit back to the pcm and it was fine (.5 ohms).

the weird thing is that on the purple (5v) wire it is reading 2.7 v all the time (should it not read 5v all the time)?. I have tried robbing 5v's from the tps to run back to the sending unit and nothing changed. I have tried to ground out the purple wire and still have not had a change in the fuel gauge. I have tried to bypass the sending unit and jumped the purple wire to black and white and still no change.

Should i not be able to get the fuel gauge to move one way or the other if i ground out the purple wire? should i not be able to get the gauge to move if I apply 5v to the black and white wire?


I am a little confused, i have seen one diagram online that shows the black and white coming back to the pcm and joining with a couple of other straight blacks and going into the pcm. Should the black and white wire not go back to the pcm seperately to measure the voltage drop across the fuel level sensor?


Any help or input would be greatly appreciated as this is driving me nuts

[dsmuts]

Your fuel sending unit could be bad.

[rooneystang]

but why can i not get the gauge to change when i apply voltage/ground to the wires?

[dsmuts]

That i don't know.

[rooneystang]

ps i can get the gauge to read completely full by unplugging the connector at the sending unit

[rooneystang]

Well i put a new cluster after following the flow chart on mitchell. That made it so I can at least get the gauge to read completely full or completely empty. I then decided to replace the level sensor, with no luck. I have bypassed the wires going to the level sensor by running wires straight from the cluster, they are 2 wires that come from the cluster a purple and a black with grey. The level sensor was reads the proper resistance through the entire range of motion.

The fuel gauge will begin to move up when you start the car (as they all do) but when it reaches half a tank it goes straight to empty.

Does anyone else have any other ideas or ways to try and diagnose this problem?

[Rico]

The only thing you haven't done is replace the harness to the sender, right?

Unplug everything and make sure all wires are open with respect to ground and to each other.

Make sure you have continuity from connector to connector on all wires.

Plug everything back in and make sure any grounds have continuity with the chassis (0.0 ohms).

This will rule out a wiring problem.

[rooneystang]

I have checked continuity trough the harnesses and it checked out good. There are no direct grounds to the level sensor itself, only the black wire that grounds the fuel pump itself. I ran two wires right from the cluster to the level sensor to rule out the possible to of a short to ground somewhere.

I am going to check and see if the clusters are different from year to year, maybe theres a pin that needs to be moved.

[Rico]

On the sender does the resistance increase when more fuel is added to the tank? Or is it the opposite?
Your supply voltage may be dropping out at higher current. When it drops to empty does the supply voltage still read 5 volts?

[rooneystang]

as you add fuel to the tank the resistance increases.

are you sure it is 5v that should be on the purple wire or is it 12v? I am starting to second geuss.

The black and grey wire does go to ground (i thought i came back to the cluster) and it is good. I put the black and grey straight to ground and ran a wire back from the cluster again with no luck.

The cluster I have in the car now is out of a 1998. The wiring diagram is different for the two cars. On the 97 the purple wire goes straight back to the fuel level sensor, and on the 98 the wire is org and runs to the pcm.

Perhaps this may be my issue?

[Rico]

You may be on to something. I know that the fuel pump signal (on a 2000 GP) is 12 volts right out of the PCM to the relay. I imagine the supply voltage to the sender resistor could very well be 12 volts intended.

Do any of your neighbors have a vehicle with the 3.1 in it? LOL I'll bet all of GM's are the same though.

Are you subscribed to ALLDATA? This is probably your best option unless someone with the knowledge chimes in.

[rooneystang]

I am subscribed to mitchell, and so far it looks like it is the cluster. Besides the fuel gauge going back to the pcm on the 98 the fuel gauge part of the cluster appears to be pretty much the same. I checked the voltage going out to the level sensor on the new cluster and it was only 1.7v. Mitchell doesn't say what the voltage should be, but i haven't found a lot of sensors out there that run on 1.7v most 5v and some 12v.

I would almost think that the supply voltage to the cluster would be bad, but nothing else is affected and works perfectly.

[Rico]

Here is a thought: Is the gauge supposed to do a calibration when the key is turned on?

When the key is turned on (when the gauge is working properly) does the fuel gauge go to full and back to zero for calibration? You might be seeing the calibration process only. With an incorrect supply voltage the calibration process may be incorrect. Try filling the tank all the way and see what happens. I'll bet after calibration the gauge goes back up to around half.