Engine Dying at Idle

[gte619n]

Engine Stalling at Idle

Hey guys,

Here's my original thread WRT to my issue. Basically, I was getting a really high idle (1,250+) and what seemed like some very hot running. I inspected the throttle body boots and the AFM boot and a there were cracks-a-plenty. My AFM also looked like it had been messed with as the box itself had a thick layer of black silicon around it. I decided that I should swap over to the Miller MAF and install the silicon boots once and for all.

The conversion was a success. The car seems to idle much better around ~1,000 rpm. I drove the car up to Chateau Elan for SESF (70 minutes), stopping for gas, and everything seemed great; however, once I pulled into the parking lot the car begin stalling at idle. I could give the car more gas and it would rev up, but as soon as I let off the pedal it would stall. 20 minutes later, it was back to fine running.

I drove from Chateau Elan to Highlands, NC without issue (2 hours). I started and stopped the car all weekend and it was fine. Today, I drove back from Highlands to Atlanta in a MONSOON and the car was running great. I stopped to have lunch and get some new tires. I started home and as I'm pulling off I85, the car begins to stall at idle again. Luckly, I was able to pull into a Chik-fil-A. I had some ice cream and felt better.

I let the car sit for 30 minutes, hopped back in and everything seemed fine. I made it 2 miles in stop and go traffic and the condition returned. This time I pulled into a Wendy's. I had a frostie but I am not feeling any better.

I'm going to wait for traffic to clear and then limp the car home. It doesn't seem like air leaks to me, as everything looks good. What else could it be and where should I start? It sounds like jumpping the fuel pump and main relay is a good starting point, but I'm not sure where to find it on the M635CSi. Timing light on the coil to see if it stop sparking?

Symptoms seem similar to Sixracer1's post here, but I'm not sure how that would account for the intermittent conditions.

[gte619n]

Fellas,

After more consideration, could this be a dead or dying coil? I believe that this is the original part. Could it be crapping out when it doesn't have enough rpm to spark frequently? Would that make sense?

[gte619n]

Hey guys, I did some test readings of the CTS and the coil.

CTS looks like it checks out: 2.59k ohm at ~75 degrees and then 291 ohm after about 10 minutes of running.

Now, the coil is a little bonkers. I am getting 5.6k ohm (cold) and 6.0k ohm (hot) from 15 to the center. I think that's okay, but I was have A LOT of trouble getting a consistent reading between 15 and 1. When cold, I never did get a settled number with the meter bouncing all over the place. When hot, I got a bouncing number between 50k ohm and 0ohm. Sounds like the coil, while maybe not the problem, is certainly part of the problem?

Furthermore, in checking out the CTS, temp sender and temp time, I must have jiggled some contact and my temp gauge started working again. It was showing well above the center point, maybe 3/4 of the way up the gauge. Could that also be part of the problem?

When the car was cold, it started up and idled great. After running for a bit, the stall issue manifested itself again. So, at least I feel like I have a repeatable scenario.

[Brucey]

the coolant temperature gauge has nothing to do with the operation of the engine itself, it is entirely separate from the engine management. However, that you have had an intermittent connection there says plenty; it means that any and every connection in the engine management could be bad too, including those to the crank sensors; I would suggest checking and cleaning everything. If you are going to check components, it is a good idea to double-check the resistance readings at the ECU plug; this way if there is a bad connection you have a chance of finding it.

I can (sort of) understand why you might have addressed a running fault by replacing the AFM etc but IMHO this is a classic case of treating a problem that was different to the one you actually had, and then being surprised when the original problem didn't go away. [Just because the AFM cover had been off doesn't mean that the AFM was actually bad; it could just be a symptom that something was wrong and the AFM got the blame.... Also, on an engine with a working O2 sensor I think you can usually unplug the AFM at idle and the engine ought to tick over OK...]

The idle on these cars is controlled by the ICV. This can only open when the idle contact is made in the TPS. It also ought not open at all unless the engine speed is otherwise trying to dip below the set rpms (which vary with model between 650 and 750rpm IIRC). So if the idle speed is ever higher than that, it usually means you have too much bypass air from somewhere, probably an air leak.

However on a car with a catalyst, there is an additional system in play, being the O2 sensor. This will be trying to control the mixture strength, once the engine is warm. If the signal from this is bad (e.g. bad sensor, air leak into exhaust) it will be doing the wrong thing (moreso at idle with some faults) and causing the actual mixture strength to go out of range and this can cause the engine to stall.

So I would be looking for air leaks (still), if the ICV is working properly, looking at the mixture strength at idle, and what the O2 sensor is reporting.

BTW If the coil were failing the engine probably wouldn't run well at highway speeds; it sees much greater voltages at part throttle.

Also NB; the oil heats up far more slowly than the O2 sensor; after the engine has been running for five or ten minutes the O2 sensor should be working but the air demand at idle will still be high because the oil is still cold. If the idle is good under these conditions it usually indicates that the ICV is OK, the O2 sensor is working etc. This in turn suggests that you have an air leak that is screwing the idle when the air demand of the motor is smaller, because the air leak will comprise a greater proportion of the total air demand.

hth

cheers

[gte619n]

Brucey,

Thank you for the detailed response. I appreciate the folly of the parts spray-n-pray. Two of four intake boots were cracked and the AFM boots also had a nice split in it, so I thought that was my air leak. While we were in there, I went and switched to the MAF. Afterwords, the thing ran great, but now this.

This car is a 1985 M635CSi with the M88. Being a non-cat car, I thought that it did not have an O2 sensor? Also, in 1985, I thought the crank sensor was not actually used by the ECU? I defer to your expertise on this but if you could clarify I'll go check those parts.

Given the measurements of the coil, it seems like it's bad. If not bad, it's certainly not good. The issue being temp dependent seems to point me in the electrical direction. That being said, I'm still a bit shell shocked from my former e-type, so it could just be my confirmation bias.

So my strategy now, and any advice here is 100% welcome, is:

1. Contact cleaner on EVERTYTHING.
2. Replace coil, distributor cap and rotor.
3. Repair CTS and Temp Time connectors, as they air missing the metal retainer.
4. Replace vent hose from valve cover to plenum. The PO is using this for vacuum for a non-existent evap system. (Could be air leak?!)
5. Spark plugs

For additional air leaks, where would the remaining culprits be located?

[wattsmonkey]

Short response:

Take the airbox off and have a poke around the hoses and the idle control valve.

Rambling response:
The idle was running very high on my M635 after rebuild. Went brilliantly on part/full throttle, but revs would rise alarmingly at idle.

All sorts of reasons for this, with no "magic bullet", but finally sorted now (fingers crossed).

I had a small air leak from the cam cover gasket (can't believe this was enough to affect idle speed much) and another from the central bypass screw o-ring.

Idle should be pretty high according to BMW specs: 850 to 900 rpm - there are online guides to setting up throttle bodies which help with this.

It sounds like the extra pipework on the cam cover to airbox hose is job number one after the electrical cleanup you've already identified.

Careful spraying of something like soapy water over hoses might reveal where there is a leak, although lots of pipework is awkward to get at and not very soapy water friendly (the hoses to the idle control valve, for example).

If you've got cracked hoses though, it's a good bet they are all past their sell-by date. The airbox/trumpet join has only a really skinny o-ring in each of the trumpets; there's an o-ring in each adjustment screw, the bellows between afm and airbox leads a hard life...

On top of that, the throttle bodies are easily knocked out of sync, which is likely to cause a high idle. Easy to check with airbox removed.

It's entirely possible the idle control valve needs a clean too. You have to call it an "air slide valve" or "auxiliary air valve" on the M though!

Once obvious stuff sorted, there's still the adjustment of idle with the central bypass screw to play with.

I still want to set my throttle bodies up properly by making a manometer to adjust the individual bypass screws, but the fates seem to be against me on this one!

Good luck getting it sorted.

[gte619n]

This morning before work, I replaced the coil. Condition persisted after 10:16 of idle. Will continue troubleshooting whenever the parts get here.

[Brucey]

....This car is a 1985 M635CSi with the M88. Being a non-cat car, I thought that it did not have an O2 sensor? Also, in 1985, I thought the crank sensor was not actually used by the ECU? I defer to your expertise on this but if you could clarify I'll go check those parts.....

If it is a non-cat car then (most likely) you very simply have an unmetered air leak. As the engine warms up the amount of metered air becomes a smaller proportion of the whole, the mixture goes lean and the engine stalls. There are other things it could be, but that is easily the most likely.

You should have three cranks sensors on that motor. The one at the front of the engine isn't used unless you plug diagnostic equipment into the diagnostic plug.

For additional air leaks, where would the remaining culprits be located?

anywhere and everywhere. Test everything. Rob's experience tells its own story; this was on a motor that had been entirely taken apart and every part examined or renewed before reassembly; yet still there were air leaks...

BTW the coils of these cars routinely measure weirdly for resistance on the HT side; I think there is a pressure contact inside the coil that corrodes slightly. This of course has no effect whatsoever on the way the coil works; the spark jumps all kinds of gaps in the dizzy etc. so a bad contact inside the coil itself is neither here nor there on the HT circuit. As previously noted when the HT circuit is bad the typical outcome with these cars is that you get a part-throttle misfire, rather than a bad idle. But you know that now....

cheers

[gte619n]

So the installation of the Powerfold intake boots increases the distance from the throttle body to the pleum by ~1.5". This makes the replacement OEM vent hose too short.



In short: ALJKGKJALDGKIJSLDKIGJSD!!

So I'm not very happy with the Miller Powerfold boot to begin with. The length of the boots is not consistent and, instead of being a badass black, they're more like a dull grey. I'm going to swap them out for OEM replacements. When I'm doing that, I will check all the various hoses and such that's in that area.

While the plenum is off, is there anything else that I should specifically clean or check?

Thanks!

[TN_M6]

This stalling... Is it only when the engine is hot? And is it only idle that is affected and runs fine off idle?

If so my first suspicion would be the TPS (Throttle Position Switch) signal. Bad solder joints inside it or just slightly out of adjustment are common and may open up with some heat. Use your meter to check the voltage at pin 1 when it's acting badly or stalls, or just unplug it once running and see if acts the same way. It would confirm the TPS is not providing the proper signal to control idle to the DME. It should go from 5v off idle to 0 volts when closed.

The idle speed should be mostly unaffected by unplugging the TPS when warm but should drop cold 'IF' the throttle bodies are set right and idle air balance is correctly set up. It's a whole 'process' to get it right on the S38 and the first step is to set the T/B's all to spec with the intake bellows out of the way. That may or may not get you at a good start point. But if the air balance is right a warm engine should run about the same speed with the TPS plugging in or not, maybe speed up slightly unplugged.

[gte619n]

@TN_M6

The condition appears after ~10:15 with the car running while parked. Start car, wait. After ~10 minutes, it dies at idle. During the fault condition, you can rev the engine and it kinda goes but it's herky/jerky and not very good.

I will check the TPS while the plenum is out of the way. It does make the clicking sound on contact, so I believe mechanically it's fine.

[TN_M6]

This sounds like something else related to warm up then or a bad connection somewhere as it warms up. Did you determine if you have an O2? If so I would try unplugging the O2 sensor to let it run open loop and see if that makes any difference. IF there is an O2 sensor that is good condition you could monitor the voltage to see if it is going rich or lean while stumbling. It's the AFM, CTS and O2 (if there) more or less in that order that determine fuel addition/mixture. And even if one of those last two is bad and you can creep it up to about 3500 rpm (I think about there) it should go open loop and run OK again to redline if the air meter is OK. Above some rpm it starts to ignore certain inputs but I cannot remember exactly when/what/how.

[gte619n]

@TN_M6

No O2 sensor, as it's a euro car. Just swapped the AFM over to the Miller MAF setup.

[TN_M6]

I've read some things on mye28 where people have had some issues with the Miller MAF but might be a different scenario altogether than yours. Does that eliminate or change the original intake air temp sensor? That may fine tune a/f ration but I'm not sure that could have profound effect on it.

I guess I would focus on the CTS since besides the MAF (fuel delivery) it is the one thing that really impacts a/f ratio AND changes significantly with temperature. If you simply had a bad connection or broken wire that was intermittent it might run fine cold but then badly as it warms up. I also once had a cheaper CTS that would open up when it got hot which was hard to diagnose. Actually why not check the wiring and just through a new $20 blue Bosch CTS at it. One of the few parts I would replace just because. And another diagnostic option is to bridge the CTS input (harness connector) with a 300-400 ohm resistor and see if it runs fine as it simulates the warm engine condition to the DME.

[gte619n]

Hey guys,

I'm waiting for more rubber bits, but I was wondering if the engine has a thermal cutoff switch? Could the engine be overheating?

E

[hornhospital]

No thermal shutoff. It'll cook itself if you let it when something goes wrong with the cooling system.

[gte619n]

@hornhospital

Thanks, good to know.

[gte619n]

Well, fellas, just got the coupe back from the mechanic and it's running great.

The issue? Plug wires. Seriously. New set of $800 BMW plug wires and the thing is running great.

Now I'm not saying all the air leaks and crusty connectors and frazzles wiring and bad air meter wasn't part of the problem, but the wires were the thing that put it over the edge. Ron told me that these high voltage ignition systems, especially in the M88, can cook the silicon wires.

Has anyone else experienced this?

Thanks for everyone's help on troubleshooting the car. Great to be bombing around in it again!