1989 635csi rough idle and slight misfire

[sbr2]

Since I've solved the starting issue in my other thread I figure I'd create this one to cover why my 6er is idling roughly.

After starting it will still idle at the appropriate RPM but you can feel the car shake a small amount for about 20 seconds and a CEL will come on and then go off, after it settles down it still kind of feels slightly lumpy but otherwise OK.

The stomp test determined the exact code to be 1222 - LAMBDA 1.

A little research on the web determines that the LAMBDA 1 code typically refers to the following issues with things I've checked in parenthesis:

•air leak
•defective Air Flow Meter ( AFM)
•bad engine temperature sensor (replaced with new)
•incorrect fuel pressure (replaced FPR, Fuel Filter - pressure test was a 42 or 43psi after the two changes)
•problem with injector(s) (I replaced the injectors and had them tested when I was chasing the no-start and they are good)
•defective evaporation system
•empty gas tank (not an issue)
•issue with spark plugs (spark plugs have been replaced and tested they are good)
•problem with valves (I performed a valve adjustment after I bought it and everything checked out there)
•etc.

So what could be causing an intermittent rough idle? When it is doing this if I put my foot into the gas it balks.

[MIKYZZ4]

Did you consider replacing the oxygen sensor on bank#1?
I've replaced all 4 sensors on my 2.5L-m54 engine(pre+post cat).

[MIKYZZ4]

Have you considered cleaning the throttle body also to prevent sticking?
Just trying to cover all the variables.

[sixerguy]

I've been working through a rough idle possible misfire on my e28 w/ the m30b34. No engine light or anything on mine just a little rough and stumbly, intermittantly sounds like a misfire at all rev ranges..

As I previously mentioned on your other thread old fuel, whether non-ethanal or 10% ethanal, can play havoc in trying to diagnose rough idle and stumbling performance. I had a full tank of 91 non-ethanal go bad after just a few months recently - that's NON ethanal - and the ethanal stuff is hydroscopic so even worse. Its amazing how much h20 can get absorbed into a tank in even just a few months. The car kept stalling just getting itt out of the driveway. I had to restart the car 5 or so times and keep the revs above 2k just to do a 1/4 mile test drive loop. A complete drain and fill totally solved the problem. Since then I've been putting in Stabil fuel stabilizer.

What brand FPR did you install and are you getting 42/43 at idle or just with the car off? I've got a brand new Peirberg OE on both the e24 and e28 and new bosch pumps on both and am getting exactly 43.5 when the car isn't running, but both cars are in the 37/38 range and I'm pretty sure in my case that's what's causing the stutter/misfire sounding idle on the e28 at least.

I just started a thread for my issues and uploaded a youtube video showing the fuel pressure and vac pressure at idle. It may be helpful to compare yours to:
https://www.bimmerforums.com/forum/s...new-everything
https://www.youtube.com/watch?v=dB5F2ITHBL8

I'd suggest smoke testing the pre intake and post throttle body and check vaccum.

You can back probe the O2 sensor with a multimeter and see what its doing - somewhat. An osciliscope is better - I don't have one yet, but soon...

I doubt the evap system would have anything to do with your issue. Its interesting its on your list though. If you haven't refreshed those lines they are for sure rotten.

Are you using WR9LS plugs? I doubt your issue is plugs, but good to rule out some off brand "upgrade".

Did you refresh the cap/rotor/wires and/or check wires and coil against spec? In my case I just did the plugs/cap/rotor and the wires/coil test fine but I'm going to replace the wires/coil as soon as parts are available.

In my case I just thoroughly cleaned the ICV but I'm not convinced its fully functional. I bought a cheap ebay chinese part which was total junk and confused the heck out of the ECU causing very high idle. I'll probably have the ICV out again at some point and either swap parts and/or clean it again. The ICV for sure has a profound effect on idle.

I guess it could be an AFM. I wouldn't start there, but at some point might end up there.

Have you changed the o-rings on the dipstick and the gasket on the fuel filler neck and oil cap?

That's all I've got - at least the car isn't leaving you chilling on the side of the road any more though!

[sbr2]

Did you consider replacing the oxygen sensor on bank#1?

So I've ordered a replacement OEM o2 sensor to see if that will help, easy enough to swap out.

Have you considered cleaning the throttle body also to prevent sticking?

This has been done, the throttle body has a very prominent click when it returns to idle.

As I previously mentioned on your other thread old fuel, whether non-ethanal or 10% ethanal, can play havoc in trying to diagnose rough idle and stumbling performance. I had a full tank of 91 non-ethanal go bad after just a few months recently - that's NON ethanal - and the ethanal stuff is hydroscopic so even worse. Its amazing how much h20 can get absorbed into a tank in even just a few months. The car kept stalling just getting itt out of the driveway. I had to restart the car 5 or so times and keep the revs above 2k just to do a 1/4 mile test drive loop. A complete drain and fill totally solved the problem. Since then I've been putting in Stabil fuel stabilizer.

For sure something to consider, mine does not die on me like yours however. It just kinda runs rough for a while.

What brand FPR did you install and are you getting 42/43 at idle or just with the car off? I've got a brand new Peirberg OE on both the e24 and e28 and new bosch pumps on both and am getting exactly 43.5 when the car isn't running, but both cars are in the 37/38 range and I'm pretty sure in my case that's what's causing the stutter/misfire sounding idle on the e28 at least.

Delphi was the brand.

I'd suggest smoke testing the pre intake and post throttle body and check vaccum.

This is on the list to do but what makes me think it is not vacuum related is that in my experience a true vacuum leak where unmetered air is getting causes havoc at all points on the rev range and doesn't tend to go away.

Are you using WR9LS plugs? I doubt your issue is plugs, but good to rule out some off brand "upgrade".

Spark Plug - Bosch WR-8-LC+ (7909) are the ones I replaced the old ones with

Did you refresh the cap/rotor/wires and/or check wires and coil against spec? In my case I just did the plugs/cap/rotor and the wires/coil test fine but I'm going to replace the wires/coil as soon as parts are available.

These have not been touched yet.

In my case I just thoroughly cleaned the ICV but I'm not convinced its fully functional. I bought a cheap ebay chinese part which was total junk and confused the heck out of the ECU causing very high idle. I'll probably have the ICV out again at some point and either swap parts and/or clean it again. The ICV for sure has a profound effect on idle.

I guess it could be an AFM. I wouldn't start there, but at some point might end up there.

Have you changed the o-rings on the dipstick and the gasket on the fuel filler neck and oil cap?

I too picked up a chinese ICV for $30 bucks and and swapping it in when I was chasing the no start had no effect either positive or negative.

I swapped AFMs with a known working unit and it didn't seem to do much, but I've got a remanufactured one that will be here soon.

The o-rings and gasket on the dipstick have not been touched but I will take a look this week since it is a very easy thing to check and fix.

[sixerguy]

I just check realoem and it lists the WR9LS as the correct spark plug for your year m30b35 - same as the m30b34. You should start by swapping a set of those in. Changing the cap and rotor is pretty easy - ignore the Bentley's suggestion to remove the radiator and or accesory fan. In my experience you can get the cap/rotor out and back in with just removing the fan and shroud. Its a 30 minute job if that. Even just pulling the cap and rotor out to inspect is worth the peace of mind that they look in good shape.

I'll check on pricing/compatabillity for a Delphi FPR for mine - I can't imagine its anywhere near the cost of the BMW genuine one and I'd love to rule out FPR for my low fuel pressure at idle issue.

I'm not sure you are right about the affect of a vaccum leak being equal at all rpm ranges. Depending on where the leak is, its letting in a particular amount of air at all ranges, while the amount of metered air is changing at all ranges. So if you have an X size leak at 800 rpms and a Y amount of air being metered at idle - you still might have an X size leak at 5000 rpm but a 10(Y) amount of metered air - in that case the vaccum leak should be having less of an effect on air/fuel ration at WOT than at idle. Of course if its a leak that for some reason is changing in size based on rpms or some other cause like engine vibration then I suppose the leak could grow proportionately with the rpms. A massive crack in the intake boot might causee a leak like that....

Water in the fuel tank might cause stalling or no-start depending on the concentration. At lower concentrations it can cause misfires, stumbles, stutters, etc. Its a spectrum...

That's a good point on swapping the AFMs - I might do that on mine to rule it out.

[sbr2]

I just check realoem and it lists the WR9LS as the correct spark plug for your year m30b35 - same as the m30b34. You should start by swapping a set of those in. Changing the cap and rotor is pretty easy - ignore the Bentley's suggestion to remove the radiator and or accesory fan. In my experience you can get the cap/rotor out and back in with just removing the fan and shroud. Its a 30 minute job if that. Even just pulling the cap and rotor out to inspect is worth the peace of mind that they look in good shape.

Just ordered them, lot of places are backordered on the WR9LS. I'll check the cap and rotor, anything in particular I should be looking for?

I'm not sure you are right about the affect of a vaccum leak being equal at all rpm ranges. Depending on where the leak is, its letting in a particular amount of air at all ranges, while the amount of metered air is changing at all ranges. So if you have an X size leak at 800 rpms and a Y amount of air being metered at idle - you still might have an X size leak at 5000 rpm but a 10(Y) amount of metered air - in that case the vaccum leak should be having less of an effect on air/fuel ration at WOT than at idle. Of course if its a leak that for some reason is changing in size based on rpms or some other cause like engine vibration then I suppose the leak could grow proportionately with the rpms. A massive crack in the intake boot might causee a leak like that....

When explained like that it does make sense, I'll double check things for a crack.

[RSheiman]

I have never seen an issue with the cold start valve but if it does not work or sticks, it can mimic some of your issues. Perhaps after the car is warm, disconnect and see if any change in idle. Tough to get too however.

[sbr2]

I have never seen an issue with the cold start valve but if it does not work or sticks, it can mimic some of your issues. Perhaps after the car is warm, disconnect and see if any change in idle. Tough to get too however.

It is my understanding that the 1989 635csi does not have a cold start valve

[RSheiman]

Good. As Dad would say "one less thing to deal with".

[bkats]

I'm going through something similar with my '89!

The '89 uses motronic 1.3, which is pretty good at controlling idle. If it's bad enough to throw a code, it means the ECU can't keep up. It can theoretically handle (very) minor deviations in fuel pressure and vacuum/air.

From that, I would lean toward something pretty substantial.

Obviously an air leak can do this, but you're only going to find out with a smoke test.
If your TPS not only clicks, but gives continuity across 4 and 6 at idle throttle position, I think you can rule that out.
If your ICV tests okay for resistance and free operation, probably not that.
I think we can assume CPS and CTS are not the cause.
Fuel delivery has at least some evidence of being good with regards to pump, injectors and FPR.
You can try testing the evap purge valve for operation and suction. Though I don't think a vacuum leak like that would go away.

As far as things that specifically control idle, that kind of just leaves the oxygen sensor. I don't think the AFM signal is even used at idle.
As far as indirect things - you can look at ignition components like wires, cap, rotor.


Good luck!

[sbr2]

Obviously an air leak can do this, but you're only going to find out with a smoke test.

You can try testing the evap purge valve for operation and suction.

As far as things that specifically control idle, that kind of just leaves the oxygen sensor.

As far as indirect things - you can look at ignition components like wires, cap, rotor.

Good thoughts, I think as I have discussed with SixerGuy these are the items I will be going over one by one to see if there are any changes. Got an OEM o2 sensor coming friday so that will be the first thing.

[bkats]

BTW, the B35 stock-spec plugs are Bosch 8-series (WR8LC) and the NGK 5-series equivalent. It's a colder plug, presumably to account for the compression and cam of the B35.

You'll find that a lot of the information here needs to be screened for the B35 / Motronic 1.3. The chassis info is pretty much right on for an E28, but the engine (and management) is going to be shared with the E34 535i and E32 735i. There are several potentially critical differences between the B34 and B35. Stuff that will have you confused if you don't keep it in mind. Vacuum, injection management, crank position sensor, etc.

So, lucky you, you'll need an E28 and an E34 bentley manual (or similar) if you really want a (more) complete reference.

[sixerguy]

I don't have a B35 engine and havn't done any kind of real research on correct plugs for this engine, but realoem does list the WR9LS as the correct plug.
update: I just checked realoem.com and for a production date 01/1989 635csi they list the WR9LS in one diagram and the W8LCR in a different diagram.

Your original BMW owners manual should list the correct plugs - my one for my '86 lists it in there somewhere, I'm sure. A little searching the internet and you should be able to find an '89 635csi manual in pdf form. Anyway definitely double check!

Evidence of a worn cap and rotor is mostly down to carbon build up or signs of arcing/scorching/burns at the gold contact points. I guess really severly worn parts might be cracked or melted or worse. Compare yours visually to a photo of a new one and it should be obvious. There's probably also a resistance test - but if it looks at all junky and is already out of the car, they are not terribly expensive and should probably just be replaced.

Hidden damage to ignition wires may be possible to see at night, evidenced by visible sparking.

Yep, no Cold Start Valve on the B35 - controlled instead directly by the ECU modifying the injector pulses. I had one die on my e24, replaced it for a bloody mint - should have just left it for dead as the car started fine without it.

[Bert Poliakoff]

I have a slight miss at idle My thoughts are a sticking Injector It runs flawlessly at speed That is the one thing not mentioned here OR distributer and points as has been mentioned

[sbr2]

Seems like we all have some sort of idle issue!

https://drive.google.com/file/d/1LCL...ew?usp=sharing

The above link is a video where it is running rough, for everyone to compare.

[sixerguy]

These cars never really had the smooth kind of idle we've come to expect from newer prod. year engines. I remember that Lexus commercial with the wine glasses on the hood demonstrating how little vibrations the car generated as compared to other luxury brands - that was pretty much directly targeting cars like ours. That being said, from the video the idle doesn't sound great, but it may be exascerbated by a stone cold engine and an O2 sensor in open loop.

Bert has a good point that we should always consider fuel injector health when looking at idle issues. its a shame that the best way to check out an injector is by pulling it - which is a bit of a PITA. I swapped in new injectors on both my m30b34 cars after bench testing both sets and finding they were several in each set were either somewhat clogged and producing less-than-ideal spray patters and/or producing too much variance for fuel flow per pulse. The new set of bosch injectors I swapped in, actually the spec you'd use on your engine, were all bench tested an were all within 5% of eachother at several pulse widths whereas the old sets had as much as a 20% variance IIRC. I tried cleaning the old ones and ran several tanks of Techron through the system and the old ones didn't improve. The Bosch OE injectors are not crazy expensive compared to the BMW branded ones.

When I smoke tested my car(s) I didn't see any smoke coming from the injector/FPR o-rings, but once I had the old injectors out it was the clear the old o-rings needed replacement. Injector o-rings are an inconvenient refurbishment, but can be a vaccum leak source and if you're really trying to get the vaccum as tight as possible should be replaced.

The first thing that has to be sorted before any other idle diagnosis is vaccum. In my case on both cars I've refreshed all gaskets, hoses, tubes except for the intake manifold gasket which is just such a PITA to get to.

I watched your video and see the engine is stone cold. A video with a warmed up engine would be useful too. Does that stumble/surge smooth out after a minute or two? That idle definitely is not great. It would be good to also get a longer clip of the engine bay sound and exhaust sound. I'm not sure how helpful a longer clip will be - but it might be...

Doing a smoke test, vaccum test, fuel pressure test, monitoring the O2 sensor and other stuff mentioned before will all start to fill in the blanks as to what conditions are linked to the problem. I can suggest the harbor freight vaccum gauge and fuel pressure gauges. I tested both against better/more expensive units and at least in my case the HF stuff is accurate. I''m using the Ancel s3000 smoke tester - its a nice unit. I think I posted my results in a video linked above - it would be great to compare your numbers and see what's happening on the gauges.

[sbr2]

I watched your video and see the engine is stone cold.

I had actually unplugged the dash temp sensor to test some things while I was doing that, so it's not actually pegged cold it's hot. It exhibits brief revving at stone cold.

[sixerguy]

Bert - if you're up for it, pull those injectors and test them! My e24 ran like a champ across the whole country, yet after pulling the injectors they were clearly not peak.

Ok. Definitely backprobe the O2 sensor when hot and make sure voltage is oscilating between .1 and .9 volts - that'll indicate at least that the O2 sensor is doing something. It won't, unfortunately confirm that the ECU is getting the signal. You can probe the ECU to get that signal, however - I've never done it, but if you suspect an O2 issue its an option. If the O2 sensor hasn't been changed w/i 50-60k miles, it should be changed. Soak it and heat it before even attempting to remove it.

Try unplugging the ICV while engine running and see if there's an idle change. I'm thinking my ICV on the e28 is stilly somewhat sh*tty after having cleaned it and am going to put a scope through the intiake tube to see what its doing visually while running. I can't really tell from your video if that idle is surging - a telltale of an ICV issue. I won't swear to it, but I think I noticed on my most recent tests that the ICV plays a more active role on idle once the O2 sensor goes into closed loop. E.g. once the engine is warm and the O2 sensor is heated up and transmitting, so affecting fuel trim, the ICV plays a more active role.

So if that video was warmed up...a video showing the cold start would be helpful. I.e. is the idle better cold than warm? Does the idle start surging once warm, but not until?

I like those seats in your car by the way - its hard to get a good look in the video, but they look like M sport seats.

[sbr2]

o2 sensor replaced, it seems to have calmed down the shaky idle. Will retry with cold/hot start conditions to confirm.

Try unplugging the ICV while engine running and see if there's an idle change. I'm thinking my ICV on the e28 is stilly somewhat sh*tty after having cleaned it and am going to put a scope through the intiake tube to see what its doing visually while running. I can't really tell from your video if that idle is surging - a telltale of an ICV issue. I won't swear to it, but I think I noticed on my most recent tests that the ICV plays a more active role on idle once the O2 sensor goes into closed loop. E.g. once the engine is warm and the O2 sensor is heated up and transmitting, so affecting fuel trim, the ICV plays a more active role.

So the issue I was having was the reverse - it would cold start fine and then you could notice it switch over to the closed loop where the idle issues would present. Also unplugging the ICV caused the idle to immediately jump.

I like those seats in your car by the way - its hard to get a good look in the video, but they look like M sport seats.

Thxs, https://www.ecstuning.com/b-recaro-p...000~dk/?pdk=Aw <- that's the ones I got. All the electric seat functions work except for headrest and heated seats.

[sixerguy]

Good news on the O2 replacement making a difference. I may have been wrong on the ICV's role when cold vs. warm. I did a little researching and it suggests the opposite of what I thought. The ICV plays a more active role, apparantly, when the engine is cold.

I scoped my ICV a few days ago, through the intake boot's evap purge valve, and the ICV is not moving very much at all. It has the slightest jitter, but is essentially open at 50%. The valve moves freely by hand, and vibrates on key-on, but doesn't seem to make full open/close adjustments like I've seen in some videos. I'm not sure if that's an indication that the ecu thinks it doesn't need to make adjustments, or that the ICV is not working properly.

While I was examing the ICV I removed its mount from the valve cover - and then started the car while forgetting to re-torque that valve cover nut. It was a good reminder that the valve cover can be a source of vaccum leaks! Fyi I use the Victor Reinz valve cover gasket that has a small bead of red gasket maker pre-applied to the composite gasket material and have been happy with that brand.

So the e28's ICV doesn't make much of a difference when warm and unplugged, its subtle. I think I've done a good job on vaccum leaks, so that might be why. I took a video with the scope, fuel gauge and vac gauge all hooked up at the same time and will post it to youtube when I get a chance.

Did those Recaro seats bolt onto the floor mount points easily? Do they include the track? Do you loose any headroom vs. stock seats?

-Drew

[sbr2]

Did those Recaro seats bolt onto the floor mount points easily? Do they include the track? Do you loose any headroom vs. stock seats?

They did not come with the seat tracks, but they bolted in easily. The headroom question I am not the best source because the original seats were pretty collapsed and I sat in them once or twice.

While I was examing the ICV I removed its mount from the valve cover - and then started the car while forgetting to re-torque that valve cover nut. It was a good reminder that the valve cover can be a source of vaccum leaks! Fyi I use the Victor Reinz valve cover gasket that has a small bead of red gasket maker pre-applied to the composite gasket material and have been happy with that brand.

I replaced the gasket on mine after performing the valve adjustment with one that in trunk of the car (still sealed of course) and there is a small oil leak from the gasket now. It did not have the red sealant so I will get the one you recommend in the near future.

I scoped my ICV a few days ago, through the intake boot's evap purge valve, and the ICV is not moving very much at all. It has the slightest jitter, but is essentially open at 50%. The valve moves freely by hand, and vibrates on key-on, but doesn't seem to make full open/close adjustments like I've seen in some videos. I'm not sure if that's an indication that the ecu thinks it doesn't need to make adjustments, or that the ICV is not working properly.

The scoping is pretty cool, I also have no idea whether that indicates fault with the ICV or the ECU not adjusting. On my end I pretty much tested resistance to spec and cleaned it.

[sixerguy]

Thanks for the seat info. Without a buzz-cut my hair brushes against the roof on my e24 with its stock Recarro sport seats. I've got a bit more headroom in the e28 with its comfort seats, but I'm not a fan. So what did you do for rails? Were you able to re-use the orig. ones, or buy that seperately?

Yeah, definitely swap that valve cover gasket. Though I normally love the service from FCPeuro, they tend to mangle valve cover gaskets when they ship them, either putting them in a box where other items in the box crush the gasket, or forcing the gasket into a box that is a smidge too small. I once installed one of these slightly bent gaskets, an Elring one I think, and it was immediately apparent there was a problem. My former Emeryville, CA BMW mechanic had used a Victor Reinz one on the e24 and they did great work. I've been happy with them so far - and make sure to order them as a single item purchase from FCP so that they don't put it in the same box as say 2 gallons of Coolant.

Here's a video showing a cheap Chinese ICV vs. the orig. Bosch one - and how a scope is set up to look at the ICV's valve while its connected. I thought I did a good job cleaning the e28's ICV, but after looking again at this video it still looks pretty dirty. I'll probably have it off again.
https://youtu.be/kJ3as5i8tos

Here's a video of the the ICV on a scope, the fuel pressure, and the vaccum pressure all being monitored as the car warms up - this video is pretty inconclusive - for some reason I'm getting very low vaccum when connecting the gauge to the intake nipple I was using to connect the crankcase pressure tube as compared to the vaccum shown when doing a t-junction on the FPR. Also, even though I say in the video that the ICV isn't moving at all, IIRC it atually was moving, just so slightly I couldn't see it while holding the camera at a distance. I'm going to redo this test and try to get a better result:
https://youtu.be/ufxYei5yjrY

Also, this e28 starts misfiring audibly as it warms up. In researching it, I found a video which suggested measuring the injectors' resistance both cold and warm - on the test car in that video, the car was ODBII and conveniently throwing a misfire/injector code - and when measured for resistance the injector on the cyclinder in question produced a big variation in resistance when the car was warmed. God, how I hate taking off the injector C-Clips on the M30B34 - but I'm eager to give the new Bosch injectors I just installed a resistance test when warm. I only tested them cold for resistance, so am VERY curious!

[sbr2]

https://drive.google.com/file/d/1M1y...usp=drive_link <- video of a warm start after swapping out the o2 sensor with a new unit.

I'm cautiously optimistic that the warm start rough idle and bogging down is solved. I've recreated the exact situation that would cause the issues in the first video I linked. It now starts up easy and (I'm guessing) goes into the o2 sensor loop where it idles nearly flawlessly.

I've got to rebuild some trust with this car but this is a good start.

God, how I hate taking off the injector C-Clips on the M30B34 - but I'm eager to give the new Bosch injectors I just installed a resistance test when warm. I only tested them cold for resistance, so am VERY curious!

Bruh, I fiddled around with the outer most injector c-clips and just gave up and removed the intake manifold completely to give myself more access.

[bkats]

Cautious congratulations, I guess?

I know what you mean about trust. Every slight miss or rumble is going to get you for a while.
Definitely a tradeoff on ease-of-repair and ease-of-diagnosis.
For all their complexity, modern cars have pretty definitive diagnostics (with the right equipment), and much less condition-based running problems.
With cars like ours, unless you have a fresh sensor array, fuel equipment, and fresh ignition; it's hard to feel certain you're going in the right direction.

Happy driving!