Air conditioning on, car randomly stalls at idle

[opethmike]

So I've got a really strange one that just started cropping up a few weeks ago.

If I have the A/C running, the car will randomly shut off at idle. There won't be any surging, juttering, anything like that, it just shuts off as if I had turned the key.

While driving with the A/C, the car will keep running and driving (thankfully!), but I can feel a little surge whenever the compressor cycles.

No clue where to start searching on this one.

My car is Megasquirt EFI + idle control. I have made zero changes to the tune this year, so I am doubtful it is anything in there.

[Rich W]

Hi Mike,

I have seen the same type of issue on another DMC club member's car. Stock K-Jet PRV, 5-speed, engine stalls intermittently at idle, only when AC is on.

I have seen the engine stall while at idle (in neutral) or when coming to a stop (previously at speed) taking foot off gas pedal and depressing the clutch pedal.

This issue first cropped up during the July 4th parades, a little over a month ago, but the intermittent stalling problem still persists.

With the AC off, engine runs fine and does not stall ... idle remains smooth, either in neutral or with clutch disengaged.

I am planning to investigate the issue further after the Woodward Dream Cruise this upcoming Saturday.

Later,
Rich W.


So I've got a really strange one that just started cropping up a few weeks ago.

If I have the A/C running, the car will randomly shut off at idle. There won't be any surging, juttering, anything like that, it just shuts off as if I had turned the key.

While driving with the A/C, the car will keep running and driving (thankfully!), but I can feel a little surge whenever the compressor cycles.

No clue where to start searching on this one.

My car is Megasquirt EFI + idle control. I have made zero changes to the tune this year, so I am doubtful it is anything in there.

[Bitsyncmaster]

My guess to the problem would be the sudden current load when the AC clutch and fans kick on at once. You probably have a bad connection (maybe a ground) that drops to much voltage and you ECU just cuts out with that low voltage.

[David T]

My guess to the problem would be the sudden current load when the AC clutch and fans kick on at once. You probably have a bad connection (maybe a ground) that drops to much voltage and you ECU just cuts out with that low voltage.

I agree but it could be even more subtle than that. At idle the alternator can't out out as much so you wind up running on the battery. Even without the fans and compressor hitting hard on the electrical system, with the alternator unable to keep up with the load, the battery voltage will drop to a critical point and the ignition ECU will call it quits. Any bad connection will only make things worse.

[hmcelraft]

I have had that symptom before. It was after I had replaced the micro switch and I messed with the adjustment. The adjustment for the micro switch is important as to timing for turning on the idle speed motor. Turning it on too soon causes the engine to stall - to late and you get a run-up in the engine. I think I got that right? Anyway, after I adjusted the throttle movement everything went back to normal. Look on page D:04:15 of the Workshop Manual - Spec 3.

[hmcelraft]

So I've got a really strange one that just started cropping up a few weeks ago.

If I have the A/C running, the car will randomly shut off at idle. There won't be any surging, juttering, anything like that, it just shuts off as if I had turned the key.

While driving with the A/C, the car will keep running and driving (thankfully!), but I can feel a little surge whenever the compressor cycles.

No clue where to start searching on this one.

My car is Megasquirt EFI + idle control. I have made zero changes to the tune this year, so I am doubtful it is anything in there.

Oops - Megasquirt! Never mind

[dmc6960]

I had an issue like that back in 2010, that one was caused by low voltage causing the Megasquirt computer to shut down. I could see it in the last few frames of the log the voltage dropping during the stall.

Lately this year, I've been having a similar issue, but this one is not voltage related. I believe my OEM idle motor (I'm still on stock idle control) has gummed up and is not responding properly to input from the computer. Under a light load I'm getting oscillations I've never had before, and under a heavy load (AC on) it tries to stall entirely. It still seems to rotate freely by hand, but after 140,000 miles, it really could be at end of life. Of course, that is only a theory until I actually try to fix it.

[opethmike]

I think Mr. Reeve may be on to something here. While I was running the car yesterday, waiting for the stall to happen, I noticed that my idle voltage was only 13 volts. I didn't see the gauge in Tuner Studio drop when the car shut off, but after reading Jim's post, I examined my log closer and found that voltage did in fact drop, just about an entire volt, at the time of shut down.

So for shits and giggles, I put my spare battery on my charger over night, and installed it today. At idle, I found I was now, on the "new" battery, getting 14.4 volts at idle. I could see the voltage drop down to ~13.5 when the A/C compressor would cycle, but the car never stalled. Granted, I didn't run it for particularly long today, but this makes me think that the weak, "old" battery combined with the drop during cycling is what was causing the stall.

Still, that seems like a pretty brutal voltage spike from the compressor cycling. I cleaned every single f'ing ground on the car last year, so I doubt that it is a grounding issue. If/when I improve that voltage drop, I'll let everybody know the solution.

[mluder]

Slightly askew from topic... Anybody ever have an issue with an idle surge as you come to a stop?

My car will be running along just fine and as I come to a stop I get a slight idle surge to about 1200 RPMs and the it sits down into normal idle. More pronounced when the AC is running but it does it whithout the AC as well.

Thanks

[Bitsyncmaster]

Slightly askew from topic... Anybody ever have an issue with an idle surge as you come to a stop?

My car will be running along just fine and as I come to a stop I get a slight idle surge to about 1200 RPMs and the it sits down into normal idle. More pronounced when the AC is running but it does it whithout the AC as well.

Thanks

It depends how you drive your car. If you decelerate until the engine is below 775 RPM and then hit the clutch it would probably surge on most cars. It's funny how each driver will shift the car different from other drivers. I tend to step on the clutch at 1500 RPM or there about. I don't use the tach for my shifting, just the sound and speed.

[Jonathan]

Slightly askew from topic... Anybody ever have an issue with an idle surge as you come to a stop?

My car will be running along just fine and as I come to a stop I get a slight idle surge to about 1200 RPMs and the it sits down into normal idle. More pronounced when the AC is running but it does it whithout the AC as well.

Thanks

I had this idle surge on my car Steve, and it turned out to be related to the connection to the thermistor. I believe my thermistor is either not well connected or not connected at all via the electrical plug. I put a jumper on the two corresponding pins on the ECU connector to bypass it entirely and the idle surge went away. I plan on going in to reconnect the thermistor only when I really have a good reason to tear into that area (VoD). As I understand it, the function of that thermistor is only going to affect driving the car in cold weather, which I don't typically do. You might try this just as a test at least.

[mluder]

It depends how you drive your car. If you decelerate until the engine is below 775 RPM and then hit the clutch it would probably surge on most cars. It's funny how each driver will shift the car different from other drivers. I tend to step on the clutch at 1500 RPM or there about. I don't use the tach for my shifting, just the sound and speed.

I'm not quite sure when I engage the clutch on deceleration... I think I engage it pretty early. I'll pay more attention the next time I drive and see if that's causing the issue.

Thanks

[mluder]

I had this idle surge on my car Steve, and it turned out to be related to the connection to the thermistor. I believe my thermistor is either not well connected or not connected at all via the electrical plug. I put a jumper on the two corresponding pins on the ECU connector to bypass it entirely and the idle surge went away. I plan on going in to reconnect the thermistor only when I really have a good reason to tear into that area (VoD). As I understand it, the function of that thermistor is only going to affect driving the car in cold weather, which I don't typically do. You might try this just as a test at least.

Where is the thermister at? On the Y pipe? I replaced everything in there approximately 2 years ago but it's possible something may be faulty in the wiring.
Can you point me to which pins to jumper on the ECU so I can test?

Thanks

[Jonathan]

Where is the thermister at? On the Y pipe? I replaced everything in there approximately 2 years ago but it's possible something may be faulty in the wiring.
Can you point me to which pins to jumper on the ECU so I can test?

Thanks

Try looking through this:

http://dmctalk.org/showthread.php?8969-Engine-stops-instead-of-idle-when-warm&p=148325&viewfull=1#post148325

The wires mentioned are: "I made up a jumper and put it across the black/slate and black/yellow wires at the connector on the ECU. Started the car up and it went back to idling at 775. Shut the car off."

It might not be your problem, but it's a pretty easy thing to test to at least find out.

Here is a good picture showing where the idle speed thermistor is:

30022

[mluder]

Try looking through this:

http://dmctalk.org/showthread.php?8969-Engine-stops-instead-of-idle-when-warm&p=148325&viewfull=1#post148325

The wires mentioned are: "I made up a jumper and put it across the black/slate and black/yellow wires at the connector on the ECU. Started the car up and it went back to idling at 775. Shut the car off."

It might not be your problem, but it's a pretty easy thing to test to at least find out.

Here is a good picture showing where the idle speed thermistor is:

30022

Sorry, to bug you on this... I saw further on in the post about how you made the jumper but I'm not exactly clear... How exactly did you do it inside the connector and would it be easier to cross the wires outside the connector? I could understand doing it outside - essentially remove some sheathing and clamp a wire across. How would you do it inside the connector?

Thanks

[Jonathan]

Sorry, to bug you on this... I saw further on in the post about how you made the jumper but I'm not exactly clear... How exactly did you do it inside the connector and would it be easier to cross the wires outside the connector? I could understand doing it outside - essentially remove some sheathing and clamp a wire across. How would you do it inside the connector?

Thanks

No bother at all. I meant to take a picture and post it after I was finished and forgot, so thanks for the reminder :)

Here's a couple photos:

3003030031

There are actually three different lengths of wire to do it this way. I too didn't want to do it in such a way that I couldn't put it back the way it was. I tried to wiggle the wire connectors out the back of the connector but that didn't seem like it was going to happen, so I did it with a set of jumpers between the two ends of the connector.

Step 1: Put in an old taped episode of MacGyver on your VCR... just kidding.

... the back of the ECU connections are the male spade ones and the ones on the inside of the connecter are the female ones.

You want to continue the other two wires to connect like they were so that's where I made two wires with a male end on one end and a female end on the other. I made them with enough slack to give myself some room to work. All it was then was the male end of the wire plugged into the connector (with the wires off the other side going to the engine) and the female end of the wire plugged onto the back of the ECU. I used the crimp on connectors with some generous rubber/plastic end covers so not to risk them touching each other.

Part of the challenge with then getting the actual jumper wire to snug in beside each of those female connectors on the ECU end was that they can't be so wide or bulky that it won't physically fit. I think that was why I used a couple different style crimp connectors, because it got too tight. The actual jumper would then be two female ends and go from corner to corner as you can see where that black/spade and black/yellow wires would be connected if the plastic connector was still in place.

Does that make sense? If it isn't tight, or as another test, if you yank the whole connector off before you do any of these jumpers (or after too I guess), you should see your car spike to exactly 2,000 RPM. That seems to be the default idle if the entire connection to that ECU is lost. More troubleshooting info that test is for basically.

[mluder]

No bother at all. I meant to take a picture and post it after I was finished and forgot, so thanks for the reminder :)

Here's a couple photos:

3003030031

There are actually three different lengths of wire to do it this way. I too didn't want to do it in such a way that I couldn't put it back the way it was. I tried to wiggle the wire connectors out the back of the connector but that didn't seem like it was going to happen, so I did it with a set of jumpers between the two ends of the connector.

Step 1: Put in an old taped episode of MacGyver on your VCR... just kidding.

... the back of the ECU connections are the male spade ones and the ones on the inside of the connecter are the female ones.

You want to continue the other two wires to connect like they were so that's where I made two wires with a male end on one end and a female end on the other. I made them with enough slack to give myself some room to work. All it was then was the male end of the wire plugged into the connector (with the wires off the other side going to the engine) and the female end of the wire plugged onto the back of the ECU. I used the crimp on connectors with some generous rubber/plastic end covers so not to risk them touching each other.

Part of the challenge with then getting the actual jumper wire to snug in beside each of those female connectors on the ECU end was that they can't be so wide or bulky that it won't physically fit. I think that was why I used a couple different style crimp connectors, because it got too tight. The actual jumper would then be two female ends and go from corner to corner as you can see where that black/spade and black/yellow wires would be connected if the plastic connector was still in place.

Does that make sense? If it isn't tight, or as another test, if you yank the whole connector off before you do any of these jumpers (or after too I guess), you should see your car spike to exactly 2,000 RPM. That seems to be the default idle if the entire connection to that ECU is lost. More troubleshooting info that test is for basically.

Ahhh... I get it.

Thanks for the pics... I'll give it a shot when I have some free time and report back.

Thanks again.