3.4L van intermittent no-startproblem/fuel pump relay chattering

[CANADA43]

Anybody seen this problem before? I have a 1996 Pontiac Trans Sport 3.4L (without Passkey anti-theft) which on occasion won't start after being parked.

The fault only happens when the vehicle has been sitting, it does not die while being driven (so far). The engine cranks normally but does not fire. The peculiar and significant thing about the fault is that the fuel pump relay (driven directly by the PCM), chatters at about a 10 Hz rate. Instead of receiving a steady 12VDC to keep it energized, it is receiving a pulse stream from the PCM.

The first few times this occurred, I heard the relay "chattering " under the hood , and was able to start the van by unplugging the relay and shorting the connector terminals for the contacts together with a wire jumper ( these terminals connect "hot at all times" 12VDC from the PCM fuse (orange wire) through the relay contacts to the fuel pump +ve). The electric pump would run and I could start the engine easily. By the time I arrived at my destination, the fuel pump relay 'drive' would be normal and I could plug the relay back in and continue normal operation.

The last few times this problem occurred, the van would not fire -- no spark as long as the relay drive was pulsing. When it would finally go back to normal operation of the relay, the engine would start OK.

Last Spring in cold weather I had the van towed 2 blocks to a Pontiac dealer with this problem. The problem resolved before they inspected it; however after the Technician's inspection the Service Advisor said there was no fuel pressure at the test port and that the pump must be deffective (shorted), overloading the relay causing it to chatter. That diagnosis made no sense to me. They wanted to replace the fuel pump as the solution ($600). I got past the Service Advisor and asked the mechanic to show me the fuel pressure -- the car started on the hoist when I tried it, and fuel pressure was normal (>30psi). I left with just the relay changed, as they thought the original would have burned -up contacts from excessive current to the pump. (I cut it open later and found its' contacts to be perfect). While at the dealership I asked the S.A. if he knew of any TSBs, or other incidences of this problem; he said he had not heard of this problem before.

I had the same problem one week later, but not again all summer until this Fall's cold and wet weather. I had to have the vehicle towed home and it still wouldn't start the next day. As soon as I accessed the PCM behind the glove box and jiggled it the relay chattering stopped and car could start. However, later the same day, still in the driveway, the fault was back and no amount of jiggling or other manipulations had any effect. Later still I went out to test for spark, and the fault was gone. (Cold, wet weather all day).

Vibration seems to have little effect. I have checked the PCM connectors' security and pulled, twisted and wiggled the wires coming from it, as mentioned above, to no effect.

I suspect the fault is inside the PCM and is thermally induced by cold weather. The next time it happens I will try a warm hair dryer on the aluminum housing of the PCM.

I wonder if there is a fix for this (repair a cracked solder joint, etc.) other than replacing the PCM. I work in electronics, and will open the PCM up and inspect it before forking out for a new one.

Sorry for such a long dissertation, but does anyone know about this?

[CANADA43]

Well, the hair dryer worked this cool wet morning, after 4 minutes of heat the fuel pump relay stopped chattering and car could start.

Regarding the PCM, can it be operated on? Someone suggested that it would be hermetically sealed in potting compound, silicone, epoxy or similar material, making it impossible to work on.

Re: the ground connection; it's good on both the pump and relay; they won't energize at all without it. But hey-- I wasn't thinking about the ground for the PCM. A poor ground would certainly screw that up...

[CANADA43]

Just to finish up this issue on this old thread; shortly after last post in 2004, I opened up the PCM and checked it over - no faults found inside. PCBs are conformal coated and electronic components are sealed onto the boards. Bought a used PCM and it cured the problem totally. I am still driving the vehicle, and never experienced this problem again.

[LMP]

Good to know. Later better than never