Mercedes-Benz Forum banner

Idle stumble and race (very infrequently)

6K views 74 replies 9 participants last post by  dolucasi 
#1 ·
Engine in question is M103, 1989, 2.6L

Everything is in order from engine performance point of view:

(1) Idles normally ~650RPM (manual transmission, in neutral)
(2) Accelerates normally
(3) Fuel efficiency is normal
(4) Vacuum gauge is pegged all the way to the stop during idle

Has a rebuilt Fuel distributor that is about 20K miles ago.
ICV has been replaced and the behaviour has not changed at all.

Here is the issue that is a nuisance. I'm afraid it is a symptom that is going to get worse over time.

Issue is an idle "stumble" issue. About once a day, the idle jumps to 1100RPM from 650RPM, then corrects and stumbles.
This can happen 2 or 3 times in a row and the engine my stall the third cycle. But sometimes it just self corrects and this does not happen again the rest of the day. If there is a stall, there is no issue restarting. One quick strike and the engine is running again. During the stumble it appears Air Flow Meter drops to let in more air on its own. So it appears emissions related not ignition related.

Additionally during warm start if the car was sitting for more than 15-20 minutes if I crank the engine unless I give it gas, it will start and immediately stall. Second strike is always a perfect start. Cold start no problem, hot start no issue at all, ome quick strike and the engine starts.

Don't know if the two are related at all, probably not.

What can this be? Both are somewhat annoying but not detrimental at the moment.

Any help would be appreciated.
 
See less See more
#2 ·
Two quick suggestions:

Regarding the idle jump, check the AFM’s potentiometer … and regarding the warm start issue, check the fuel system’s holding pressure.

I don’t know if there are proper instructions here at BenzWorld for testing these things ... I intend to create detailed instructions when I have more time. You could, however, have a look at the following threads:


H.D.
 
#10 ·
H.D. Here is the AFM pot readings.

1-3 - 4.2K, never changes when the AFM is pushed all the way down
1-2 - 1.5K initially, then when moved about 1-2 mm jumps to ~3K, another 1mm or so back to ~2K then steadily climbs up to 19K about a 2/3 in and the last 1/3 movement drops down slowly to 8.5K
2-3 - 5.5K initially, then when moved about 1-2 mm jumps to ~6.7K then another 1mm or so back to ~5.5K then steadily climbs up to 16K a 2/3 in and the last 1/3 movement drops down slowly to 4.5K

The only anomaly I see it the early jump for the first 1-2mm of movement.

Could this present a problem?
 
#7 ·
Hi Jayere, I know the OVP is always the first potential problem people point at but this does not have the appearance of an electrical problem.

The OVP is the original, never changed, fuses in tact. I do have an extra one that I can plug in and will do so just to rule it out though. That is an easy check, thank you.
 
#13 ·
Thanks H.D. I do not see an obvious screw/nut/bolt/latch/etc that fastens this Pot to the AFM housing.
I hope I do not have to remove the AFM/FD for this.

I could not find instructions on removal neither on the forum nor in my Haynes repair manual.

If it is not too much trouble could you point me to a document/posting?
 
#14 ·
Sorry dolucasi, I don’t know if there are procedures about the POT replacement at BenzWorld. I haven’t had time to browse through the forum yet. My procedures only exist in my head. I’m thinking about creating an extra thread about the POT, like for many other things too, but currently I don’t have enough time for that.

You don’t have to remove the AFM/FD for the POT replacement … but the fuel pressure regulator. The POT's screws are hidden under plastic caps which can be levered out with a small screwdriver.

H.D.
 
#16 ·
Yes, the screws appear to be behind the plastic tabs that I could not pry off last night. Now the problem is sourcing a good quality sensor. Seems not available at parts stores and Bosh does not make them anymore. I'll try the dealer this afternoon but I heard they will only sell you the entire MAF housing with the sensor. Sounds like an expensive proposition.

Thanks for the quick reply
 
#15 ·
my 103 had a random stumble at idle that sometimes would stall when you gave it gas to roll, this turned out to be air leaks at the injector seals after I fixed lots of other problems. a co2 or other inert gas test can help find these sorts of air/vacuum leaks (basically flood the suspect area with said inert gas, if the idle changes, there's likely an air leak).
 
#17 ·
Injectors are nearly new, replaced at the dealer ship. Hopefully they know what they were doing. I'll try H.D.'s MAF sensor replacement first if I can source the darn thing without having to go to a cheap aftermarket brand.
That needs to be replaced anyway from the diagnostics.
 
#20 ·
that part is easily tested with a multimeter. it has three pins, lets call them (top to bottom) 1,2,3 (I forget if thats how its actually numbered, that's not important here). remove the air cleaner, unplug the connector from this pot. leave ignition OFF throughout this test.

1) measure the resistance from pin 1 to 3 on the pot. I forget the correct value, its something like 3K or 20k ohm if you get a reasonable resistance here, continue. if its open circuit, then you need a new pot.

2) measure the resistance from pin 1 to 2. its either most of the value of step 1, or its near zero (I'm dumb, I forget details like this and am too lazy to look it up right now). slowly depress the air plate, and the resistance should gradually change in proportion to the plate. what you do NOT want to see are places where it jumps to open circult.
.
3) measure the resistance from 3 to 2. this will be the 'opposite' of step 2, depress the plate, it should change gradually the other way from step 2. again, any abrupt jumps to open circuit are bad.


when I got my 300E 2.6, pin 2 on the pot was open circuit. I bought the aftermarket replacement pot, and it was STILL open circuit, so I ended up replacing the whole air flow meter assembly ($$$), which was probably stupid, I should have instead gently bent out the contact fingers in the AMF that the pot presses against. however, my car runs great now.
 
#21 ·
You probably missed post #10 on this LCG. I've done the tests and it appears according to H.D. most likely the problem is the "discontinuity" or "jump" in the resistance readings right at near closed throttle. H.D. believes this could cause the gyrations that I'm seeing at idle.

I'm tempted to just go to the Pick-and-Pull and test one that passes the test and snatch it for $20 just to make sure this is the issue.

I may still go out and purchase a bosch for $225 though.
 
#23 ·
Visited the pick and pull today on this matter to:

(1) Practice on removing the AFM Pot without having to remove any fuel lines. Seems doable, though one screw is somewhat blocked with a special tool I can do it.
(2) Take measurements on a few of these units in the lot. And if I found a good one to just remove it and try it on my car.

No luck, all 3 units checked were jumpy on the low end and all were actually worse then mine. So this avenue is shut down.
I'm also guessing this may not be the root cause of the idle stumble, but since mine is starting to age it is worth replacing.

One question I have is that the part numbers are different from year to year and model to model (260 vs 300, etc).
The one I found on AutohausAZ is a different part number than mine and they claim it is the incorrect part number for my car. The form factor for all of then appear to be the same however.

So I'm heading to the dealer I guess....
 
#24 ·
... I'm also guessing this may not be the root cause of the idle stumble, but since mine is starting to age it is worth replacing ...
Right. … As I said in post 12, on your car the POT resistance jump is, besides other possible causes, a typical cause for “idle jumps” (in your case from 650rpm to 1100rpm). … For the “idle stumble” there can be additional causes.

H.D.
 
#32 ·
Also neither the port from the thermovalve or the one from the intake really has any significant vacuum on them during idle. Actually almost zero vacuum.
If the engine is revved both develop some vacuum.

I think this is how it is supposed to work but if not let me know if this is not all normal.
 
#38 ·
H.D. Here is the data you had asked for:

40c - 900rpm 2 to 1 0.85V 3 to 1 4.86V
60c - 750rpm 2 to 1 0.68V 3 to 1 4.86V
80c - 700rpm 2 to 1 0.55V 3 to 1 4.86V
85c - 700rpm 2 to 1 0.54V 3 to 1 4.86V
85c - 1000rpm 2 to 1 0.70V 3 to 1 4.86V (achieved by pushing the throttle)

85c - 1000rpm 2 to 1 0.88V ICV disconnected. During disconnect engine revs up to 1000. If plugged back in the engine stalls (the throttle plate moves too)
85c - 700rpm 2 to 1 0.46V ICV plugged in, EHA unplugged.

Your Thoughts? Thanks in advance H.D.
 
#40 ·
Hi H.D, can you also comment on the behavior when EHA and ICV are unplugged? Is all that expected normal behavior as well?

I think I will observe behavior with the purge valve bypassed a few more days. Since the occurance is very infrequent it is not a big problem at the moment. Basically, when it starts (the jump and stumble) if I rev the engine it recovers and everything is normal.
Until it happens a day or two later.
 
#46 ·
Hi H.D, can you also comment on the behavior when EHA and ICV are unplugged? Is all that expected normal behavior as well? ...
The engine’s reving up with the IACV plug disconnected is normal. ... Since the IACV is controlled by the ECU, the engine behaves in the same way if the OVP relay’s supply voltage to the ECU is cut off, unless the duty cycle is adjusted to a value significantly higher than 50%.

Idle speed not further changing when (with Lambda control in action and IACV disconnected) you additionally disconnect the EHA plug shows that the duty cycle is close to 50% (at that rev).

If the overall idle speed change is the same as the one that you see when the infrequent idle jumps happen, it can not be excluded that these idle jumps are caused by an infrequent problem with the OVP relay. ... But with a definitely existing resistance jump on its output track the AFM potentiometer is more likely the culprit.

H.D.
 
#41 ·
Seems like we may have done it now guys. After the last tinkering with plugging/unplugging EHA valve and idle control valve taking readings on the AFM pot etc. I may have tripped the "check engine" light.

It could also be independently tripped but I can not turn it off.

I actually tried to read a code but was not able to. With ignition on engine off pushing the button by the port near the battery (CA version) for 3 seconds I thought triggers a code reading process. Nothing happens.

Also, holding the button down for >10 seconds I thought resets this. No luck.

Meanwhile the charcoal purge valve is still unplugged with the 2 vacuum ports plugged. The One to the canister is not plugged.
Not sure if this would effect the warning light.

Engine is running just fine, no idle issues except the occasional stall (one a day maybe?)

Any help would be appreciated on the check light.
 
#43 ·
It's my understanding that the CA models with the test button only show emissions related codes. If you want to pull the rest of the engine codes, you go to the standard diagnostic port with you blink light code reader and count the blinks.

I was told if you disconnect the battery leads and touch them together for five to ten secs, you've wiped out all the code faults and can start over.

You should take a pic of the purge valve so the membership knows what that critter looks like and can identify it by sight, wherever it might be on their cars.

Kevin
 
#48 ·
Well guys, RayH might be correct in his suspicion. I have not had a idle race/stumble or stall in the last 4 days.
Nothing has been replaced so far, however we have plugged and unplugged several connectors (ICV, EHA, AFM).
And now I can not get it to repeat the misbehavior.

Quite to the contrary the car is running better. One thing that always bothered me was that once a few days or so the car would also stall when coming to a stop. Mind you it is a manual transmission so coming to a stop when the clutch is depressed and the rpm's are dropping from say 1500 to idle, it would just drop like a rock down to 500rpm or so and the it would recover. But every now end then it would drop like a rock and stall. I always attributed this to the hick-up that coincides with a clutch action coming to a stop.

Now, after all the poking and prodding the rpm's never drop like a rock, it comes down to about 900-1000rpm and takes 4-5 seconds to ease down to 750. Not sure if this is by design but I much prefer this behavior.

My Engine is super clean including connectors but after 28 years even in the dry CA weather, I assume some of these connectors may oxidize.

So, I'll wait till the misbehavior comes back and when it does, I will clean one connector at a time. I'm suspecting EHA but it could also be the others I suppose (like the ICV, which increases the RPM when unplugged and stalls when plugged back in)

I'll report back in a few weeks.
 
#50 ·
Thanks for the suggestion sheperd, since I have one ICV out of the car, I can play with this but it looks super clean inside. However, this is the time to see the operation of the unit out of the car. I have a variable power supply that I can use as well.

-Cheers
 
#58 ·
Thanks H.D., you seem to have the answers to all the questions....
I'm happy that my now fully restored car (inside out) is operating near 100%.
Well, it still needs new valve guides but that is another story.

As a final check I will:

(1) replace the AFM pot with a better unit to observe cause and effect
(2) replace the ICV with my original that has been flushed with carb cleaner and electrically verified as an experiment.
(3) I will wait for the OVP replacement as everything seem to be OK for the time being.

I will report any findings of course

Cheers!
 
#59 ·
Well Folks, after 3 weeks of no stall, it happened again today.

So I decided to go in reverse order. You get your wish Jayere, I replaced the OVP today. Let's see if it has any positive effect.

Will report the finding when/if I have the next stall.

-Cheers
 
#60 ·
After the OVP change (not a new one) the spontaneous stall is still happening. It appears more often know, not sure if it is coincidental.

Just as a reminder, the idle race up to 1100 rpm and then stall is not happening anymore after we fiddled with a bunch of sensors.

Now it is just a spontaneous stall out of nowhere. I was observing the vacuum today with the car idling. It is around 17 in. Hg. And all of a sudden the vacuum jumps to ~20Hg and it stalls instantaneously. The RPM's do not move up at all, it just dies. According to the Mityvac the vacuum is held at 20 in Hg after the engine stalls. Quick key strike and the engine starts right up, like nothing is wrong.

This is sort of what happens when coming to a stop and the clutch is depressed as well, the engine speed just crashes down to zero.

Any clues??
 
#65 ·
DC mode is good enough. … Particularly check if it gets stuck within the range of 3.5 - 4.5 V … and maybe you can simulate the heat under the hood, for instance with a hair dryer.

And while you‘re at it you might as well check if within the range of 5 - 6.5 V its opening degree is about the same as at 0 V.

The IACV should always “rattle“ when it‘s shaken, also when it's dry (and warm).
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top