Help us help you. By posting the year, make, model and engine near the beginning of your help request, followed by the symptoms (no start, high idle, misfire etc.) Along with any prevalent Diagnostic Trouble Codes, aka DTCs, other forum members will be able to help you get to a solution more quickly and easily!
I recently purchased a 2009 Cobalt, 2.2L that had a new fuel pump/ module installed and still wouldn't run. The people didn't want to invest any more money on it. (They had also installed 4 tires and all new brake and fuel lines) My first check was the fuse and relay. I have the test base for the 4 prong relay and saw that It had 12V constant power for the fuel pump side. It did not turn the pump on when the key was turned on. I used a test light and a multimeter to verify ground to the relay switch circuit. I then applied power to the fuel pump delivery side and the pump didn't run. That sent me under the car. I did find that the connection under the vehicle had been probed and was spread out causing no connection to the pump. Once that was fixed the car started. A whopping 25 minutes and it ran. The problem I currently have is that the relay is staying latched as long as the key is on. It will turn off if the PCM harness is disconnected and I don't see any stray voltage in the PCM wire to the relay when it is unplugged. As a test I used the multimeter and I see 6.5V to the latch circuit at the relay, (12V when the relay is unplugged but not enough juice to light a test light) yet the scan tool data says the fuel pump circuit is off. If I put a test light on the latch circuit it turns the relay off ( I assume robbing the power from the circuit). I am thinking I have a bad driver in the PCM that has shorted to power. I hope this all makes sense. Can anyone tell me if I am on the right track? I hate buying PCMs and having to rely on dealers to program them.
So the relay control circuit is power side switched, by the PCM, and permanently grounded? If so the relay may be held on by driver leakage or bias voltage. Though the lack of voltage on the control wire with PCM unplugged isn't conclusive. Could be a high resistance short to another PCM powered circuit in the harness, PCM plug, or fusebox.
Another thing is is the relay be correct OEM one? It's possible the relay has been swapped with one that requires less power to remain latched.
The diagram I have is for a 2008 cobalt but I have chased all connections very specifically using it. I do believe that the PCM controls power to the relay coil and the relay has a permanent ground that is terminated under the coolant bottle. I'd like to say that I know the relay is correct but I can't. It is supposed to be the same as the A/C relay and it is, but I can't say they are both right. What is the best way to discount a bias voltage from other PCM powered circuits?
What is the best way to discount a bias voltage from other PCM powered circuits?
Either de-pinning the control wire at the PCM, if it's not too difficult, or cutting the wire a few inches from the PCM are the easiest ways to narrow down where the voltage is coming from. If you decide to cut the wire check the PCM plug and socket for dirt and corrosion first.
Regards the relay you could try swapping in the AC relay. Looks like the correct relay has a resistor in parallel with the coil which could have failed.
I did try swapping the relays and it made no difference. Is the resistor the arrow in the coil diagram that points toward the fuse? I was afraid you'd say to de-pin the wire from the PCM connection. Those are some awful tiny pin release tools that are needed. I found a place that sells reman pcm and they claim they fix these units all the time for fuel pump circuit failures. I will try to isolate that wire when I get home and respond with results.
That's just showing the ground also goes somewhere else inside the fusebox. I was concerned the relay case might have an arrow, AKA diode, symbol on it.
I have de-pinned the fuel pump relay wire from the pcm plug. The wire doesn't have any power on it when the key is on. The pcm terminal is still showing battery voltage with a DVOM but not enough power to light a regular test light. Guess I'm ordering a pcm.
