PCM Problem?

[Phil]

I have a P0117 code on my '95 3.8. The diagnosis points to a bad coolant sensor, but the reading from the sensor is intermittant. Further, the car sometimes runs really rough and occasionally surges or won't respond to throttle input. Also the dash battery guage fluctuates, dropping way down when it is running rough.

A temperature sensor can't cause all these problems, but the PCM seems like a common factor. I still have the factory PCM. Any thoughts would be appeciated.

Phil

[Rob B (shoebox)]

A bad temp sensor that the PCM relies on sure can cause drivability problems. The PCM needs that info for correct a/f ratio calculations. A bad PCM is a rarity. A bad temp sensor is not.

[Injuneer]

The question is.... is it running rough that causes the voltage on the dash gauge to drop, or is it the low voltage that is causing rough running? You need to check the alternator and the wiring at the battery.

From the description of intermittent problems, that may also point to a faulty ground or postive connection somewhere near the battery. Look for corroded battery terminals, corroded wire strands in the main wires, loose connections, etc.

[Phil]

I haven't had time to do much diagnostic work yet. It will be quite simple to check the temp sensor. I will check the resistance to confirm its function. For the voltage drop, I checked the connections and they look good. The guage drop is intermittent and may be hard to simulate.

If I find the temp sensor is good, then the diagnostic table in the shop manual says to replace the PCM. I guess that is final. I may have to do that before I can find the reason for the voltage drop.

Does this sound logical?

[Phil]

I had to replace the water pump today for an unrelated problem, so while I had the coolant drained, I took out the temperature sensor. The resistance on the sensor measured fine when I checked it at various temperatures. Also the connector to the sensor measure find for the 5volt feed and ground. The temp. sensor error was intermittant before, but while I had it off the car, it worked without failure for almost two hours.

Sounds more like the PCM all the time.

[Injuneer]

Your CTS resistance matched the table on Shoebox's guide?

http://shbox.com/1/4th_gen_tech2.html#ect

End of the harness connector looked OK, no frayed or bare wires?

[Phil]

Yes, I compared the resistances to the table in the factory shop manual and the response was fine. The connector and the sensor are clean and the connections look good. I checked the ground and the voltage at the connector and they were good.

That being said, when I put it back together, the code is gone and the scanner is returning good information. I understand that these sensors are quite fragile. I wonder if the act of taking it out and checking it caused it to work better. It is hard to know since the problem was intermittant before, but occured about 50% of the time.

I will drive it again and see if how it works. Perhaps it will give me more data to work with.

[sneitzel]

The temp sensor is maybe $15?

If it were me and I had already torn everything down to replace the water pump I would go ahead and replace the sensor anyway, just in case.

Just my opinion...

[Joe 1320]

I've had a temperature sensor bench test just fine, but had intermittant faults that caused the car to stall and not restart when warm.

[Phil]

Good call. I don't like to through parts at a problem, but I didn't realize how cheap these sensors are. I will replace it and see how it goes from there.

Thanks.

[FTA1995]

Just a thought... How often do failed or failing pcm's throw an "x" code for those older obd1
models? I remember seeing that one time on a cobalt at the dealership a few years back, but that's a much newer obd2 CAN type PCM.

I hope your PCM is ok, as was said previously, PCM failure is very very uncommon.

Also, have you or a shop performed any datalogging? That might point you in the exact direction of a problem.