I had a nightmare with a Chevy Monte Carlo. It was a 2003 model 3.8 liter engine with only 50,000 original miles. The problem was the engine speed would go to 3,000 RPM. But it was intermittent, and could happen anytime in the bay or on the highway. The owner was afraid to drive it because it had unintended acceleration.
The local Chevy dealer installed a new ECM and reflashed it. Then he claimed the problem was a vacuum leak. Not! The first thing I did was to install an aftermarket ECM and reflashed it. No change.
Watching the data stream when it happens, I saw the ECM make the idle control counts go up and up. Then it would peg at 255 counts, with the idle controller fully open. I also noticed that the radiator cooling fans would come on and pulse their speed.
The data for RPM showed actual engine RPM, and all other scan data is normal. Looking for the problem by default I unplugged the following items with no effect: Alternator, A/C compressor, cooling fans, TPS, and the automatic transmission harness. But when the coolant temp sensor was unplugged, the engine stops racing and immediately stalls.
The coolant temp data reads normal, so I installed a resistor to substitute for the coolant sensor. That made no difference. Expecting there might be a short, I opened up and examined every inch of the engine wire harness and found nothing. Just in case you’re thinking, there is no remote starter or alarm, nor add-on electrical accessories.
Frustrated after wasting hours and hours trying to nail the cause, I emailed my problem to the fellow members of the MotorWatch Technical Committee in hopes of getting help.
Mike answered, “What is the TPS reading while this is going on? Sounds like I would also look at the MAF sensor, it helps calculate engine load.” I checked his suggestions and found nothing wrong.
Mark’s son Brian is a GM tech, and he had a lot of suggestions: “This engine has the plastic upper plenum which caused problems with the intake melting around the EGR port inside causing an internal vacuum leak. Check the data stream while idling and when it goes to 3,000 rpm. Most of the time a GM product will show you the problem through the data stream. Look closely at the short and long term fuel trim. This will let you know if it’s a problem the computer is recognizing through the oxygen sensor. This engine also has MAF sensor problems that can result in idle fluctuation, lean, or rich conditions. I would also monitor the TPS value, and with the engine off make a full sweep of the TPS/gas pedal to verify no jumps, or voltage drops. In 2 separate cases I have had a similar problem. And when looking at the current misfire status on each cylinder, it will show a misfire on every cylinder on bank 1 or 2. This ended up being a upstream oxygen sensor. The vehicle didn't display any drivability concerns related to a misfire and didn't feel like it ran bad or had a misfire at all. I also I would check to see the status and correlation of the CKP & CMP. It may or may not have thrown a code for this.
Bob asked if a "Known Good" ECM had been substituted. “I know you have tried factory remans, and I am doubtful that two remans could be identically bad, but I still have to ask.”
Bob continued, “I was on a 3-week internship program at the Buick Powertrain Engineering plant in 1991, and worked with the Driveability Task Force, a crew of their best techs, who would visit unfixable cars (yes, on site) and take everything with them that could be replaced. After some preliminary electronic diagnosis, they would just start substituting parts until the car was fixed and then go backward until the symptoms reappeared, then forward until the car was correct. At that point they would take the bad parts and go home for factory analysis of the replaced items. I never asked if there were cars that were truly unfixable. When you do not have a factory subsidy, this is a tough (expensive) way to go.”
“I know you disconnected the A/C compressor but is it turned off at the control? Has the switch been disconnected? This is the AC anticipate signal--but it shouldn't make it go to 3,000 RPM. Could the combination of PSPS and AC anticipate be a clue? Is the radio/sound system off, and disconnected? Has the intake manifold ever been removed to determine if there is some kind of weird crack that opens with heat? This doesn’t seem like it would correspond with high IAC counts, really just the opposite.”
“Have you tried driving it with the IAC disconnected? Is the high IAC count a cause or an effect? Could the brake booster have a massive periodic leak? Is it normal on this model car for the engine to stall with the CTS disconnected? Don't think so, but gotta ask. What about harness routing? Could any previous repairs be responsible? Suppose you’ve checked and rechecked every possible ground? I know this doesn't go along with high IAC counts, but could the 255 IAC counts by itself cause 3,000 RPM? Might be interesting to manually put IAC at max, then disconnect. Could cruise control coming on be a possibility? Glad it's not my nightmare!”
Systematically I checked each of Bob’s suggestions and still found no clue as to the cause.
Then Ken offered his explanation, “When the Body Control Module starts to command the cooling fan to cycle on, it also sends a command on the data bus to the ECM to increase the idle speed. Each time the BCM cycles the fan, another command gets sent to the ECM, with the result of the ISC counts ratcheting upward until they reach 255.”
I never suspected the problem could be a faulty BCM. The strange cycling of the fans turned out to be the clue since the BCM is responsible for the cooling fan operation. This explains why the engine would suddenly surge to 3,000 RPM. Since it was still covered under the emissions warranty, the BCM was replaced at no charge by the Chevy dealer.
