05 GP Fuel gauge problem

[Firekatt43]

First of all, I'm new to these forums so Hi everyone!!

So, I just recently picked up an '05 Grand Prix...either a base or GT, I'm not sure which. 3.8L V6 non-supercharged. Anyways, when I got the car, I was told the fuel pump was bad. The gauge showed just under 1/4 tank but still well above empty. I could hear the pump run, but no pressure at the rail. I ordered a new pump assembly, complete with sending unit. When it arrived, I went to pull the old pump out and found that the tank was totally empty. I thought wow, bad sending unit then. So, I replaced the old pump assembly with the new one and added gas. However, it seems that my gauge is reading almost 1/4 tank more fuel than it has. Basically, when the gauge reaches 1/4, it's almost empty. The gauge works fine through the entire range...except being off by that 1/4 tank....i.e. at 1/2 a tank, it's really about a 1/4 tank. I'm at a total loss on what this could be and how to fix it. Everything else works just fine, all the other gauges and warning lights. Anyone have any thoughts? Thanks in advance!!

Oh, when turning the car on, the needle sweep goes from the actual empty mark, to full then back to empty, then reads the fuel. With the car off, the needle rests on empty as it should.

Edited July 29, 2023 by Firekatt43

[ManicMechanic]

Sending unit. They are known for going bad, usually it has to be replaced with a whole new assembly.

[55trucker]

If replacing the pump (that you state included a fresh sending unit) displays the same result  then I'd say you have a wiring issue upstream.

The way the circuit works is that a constant voltage is applied to the cluster gauge, the sending unit circuit from the tank intersects the gauge circuit to vary the voltage past the resistors in behind the gauge.

The sending unit is a potentiometer, as the fuel level varies the float arm moves & current going thru the sending unit coil windings varies from high to low.

With a full tank the circuit resistance is high & when the tank is empty the resistance is low. This change in current is applied to the gauge which moves the needle across the gauge face.

If you have a constantly *high* reading then you more than likely have a resistance issue in the circuit wiring somewhere upstream of the tank.

 

 

 

 

[Schurkey]

Does an 05 use balanced-coil gauges?

Based on his description of operation, he's got stepper-motor gauges.  They're controlled by the computer.

The stepper-motor gauges are infamous for failing, providing "crazy" readings.  In my experience, they provide "crazy" readings ONCE, and then they never move again.  Others have had multiple crazy readings.

But merely reading 1/4 tank too full isn't "crazy", just wrong.

 

I'm guessing wiring problems between computer and sending unit, or sending unit and ground.  Verify with the real, live Genuine GM service manual.

[Firekatt43]

Thanks for the replies!!! Yes, my car has the stepper motor gauges. If I recall correctly, the sending unit send it's info the the PCM and then the PCM passes it on to the cluster? I wish that I actually had a GM service manual for this car, at the very least, a wiring diagram. I'll get out and start tracing wires tomorrow morning though.

Common sense is telling me that it's more than likely the resistance in the wire from the sending unit to the PCM....possibly a bad/loose connector somewhere in between.

Edited July 30, 2023 by Firekatt43

[55trucker]

4 hours ago, Schurkey said:

Does an 05 use balanced-coil gauges?

Based on his description of operation, he's got stepper-motor gauges.  They're controlled by the computer.

The stepper-motor gauges are infamous for failing, providing "crazy" readings.  In my experience, they provide "crazy" readings ONCE, and then they never move again.  Others have had multiple crazy readings.

But merely reading 1/4 tank too full isn't "crazy", just wrong.

 

I'm guessing wiring problems between computer and sending unit, or sending unit and ground.  Verify with the real, live Genuine GM service manual.

What I posted was a simplified description of the circuit,

looking what I posted is easier than trying to decipher this.......not everyone can outright understand electrical schematics.

 

 

these are the actual fuel sending schematics for the '04 Grand Prix etc

 

[Firekatt43]

That is a huge help!! Yes, I can read schematics with no problem. Looks like the sending wire goes through 4 connectors on it's way to the PCM....counting the connector at the pump unit and the PCM itself. the part I didn't know was that it grounds inside the PCM. I'm wondering if it's possible to create a ground closer....before the PCM...mainly for testing?

[Schurkey]

That schematic shows the gauge is driven by the computer.  Almost certainly a stepper-motor style, not "balanced coil".  I've had stepper-motor gauge failures in my Trailblazer, and they're not hard to replace.  But this doesn't seem like a gauge failure.  Seems like high resistance in the sensor side.

Therefore, the problem is most-likely in the fuel level sensor circuit, which starts--and ends--at the PCM.

If the fuel tank pressure sensor is working OK, then most-likely the problem is on the part of the circuit between PCM and the actual level sensor in the tank, although it could be between the level sensor and the splice where the pressure sensor joins the ground-side of the circuit.

[Firekatt43]

I want to thank everyone for their input as well as for the wiring diagrams. After doing some wire chasing, it turns out that I had a connector, that the signal wire passes through, that was corroded and not making good contact. After cleaning up the connector and adding a generous amount of dielectric grease, the fuel gauge is working fine now!!!