The impossible off idle stumble

[demetk]

In your guide, you mentioned in step 6. that "Setting the mixture screw at the base setting will now allow the ECU to run in closed loop mode and control the mixture properly". Does that mean I should set my AFM to the base setting now? I wouldn't think so because I just went through the trouble of adjusting it. I wanted to make sure. My AFM was originally set to 1, 1/4 turns out from all the way in. So I believe it was running a little rich. Now its about 2 turns out.

The "base setting" is the mixture screw setting that you finally achieved with the O2 sensor unplugged. So the "plugged in" O2 sensor voltage dithers now between .2-.8 volts every second or so?

[delorean175]

The "base setting" is the mixture screw setting that you finally achieved with the O2 sensor unplugged. So the "plugged in" O2 sensor voltage dithers now between .2-.8 volts every second or so?

Ok, I just wanted to clarify that's what the guide meant by "base setting". Yes it does bounce around between .2 to .8 volts.

[demetk]

The "base setting" is the mixture screw setting that you finally achieved with the O2 sensor unplugged. So the "plugged in" O2 sensor voltage dithers now between .2-.8 volts every second or so?

Ok, I just wanted to clarify that's what the guide meant by "base setting". Yes it does bounce around between .2 to .8 volts.

Cool. I'd be curious to see what voltages you're getting while driving. An air/fuel gauge is helpful here since it converts the bouncing voltages into something readable. I use one of these: https://www.summitracing.com/parts/sum-g2986-1 but you can also find them on ebay for cheap.

[delorean175]

I'm going to get me one of those gauges soon here. After making the adjustments to the AFM the car ran a little better. The stumble when revving got a little better but the dip in the RPM's after revving got worse. After a few drives to and from work it stayed the same. But after a nice long trip on the freeway it seams better. I'm not holding my breath because once in a great while the car would run fine, then go back to normal. It stumbles pretty bad when its cold but i'm not as concerned about that because I usually let it warm up to operating temp before driving.

[tn535i]

Maybe it just needs a good hot soak and Italian tune-up. If you generally drive like your grandmother or spend a lot of time in stop and go traffic the combustion chambers get pretty heavily loaded with carbon and it might be extreme in your case. This was the problem when my wife drove our old e30 for years and can cause off idle stumble. I've read that too and BMW used to blast the intake with walnut shells when really bad.

Nothing (short of replacing bad parts) has ever made my car run better in a short time than a hot soak and then a hard drive with full power pulls to redline. Add some rain and it's even better as moisture does as well as anything to clean out the combustion chamber.

[delorean175]

I have noticed it does run better when it has had some hard freeway miles on it, but its usually a very temporary fix. I do a lot of stop and go driving with this car, even on the freeway its stop and go. Traffic here is terrible. I go to work pretty early before the traffic starts. So I try to open her up every chance I get but its usually short bursts. I've heard of walnut blasting the intake for an N54 engine, i guess they are prone to getting gummed up inside. I had my intake off about 2k miles ago and it looked decently clean. I've read about some guys running methanol injection on N/A engines and one of the side effects is a super clean engine.


I currently have the gas tank out of the car because I've had a subtle gas smell in my car for what seams like forever. I've replaced all of the hoses (even the 3 vent hoses) and have had no luck fixing the smell. I'm checking for leaks in the tank. My car was running good when I parked it, so I'm curious to see what it will run like when I get it back together tomorrow.

[delorean175]

Hey guys, today my car was running fine when it was about empty. I filled it above half tank and it immediately started its hesitation thing. I'm stumped. Could this be fuel level related? Maybe a return line clog or something? I know some cars run poorly when they are around/ below quarter tank because the in-tank fuel pump has failed. But this is the opposite. I played around with the ICV today, and its a suspect again. I've cleaned it a few times but it never makes any change. The reason I suspect it is because the car stops doing its big dip in RPM's before it idles when the ICV is unplugged. It idles a little lower than normal but smoother. I may have to find someone local here who is willing to let me try out their working ICV.

[briansmith]

I thought I'd just bomb this thread with some more ideas.
Seeing as how you've employed the parts cannon a few times, and the engine management system seems generally in good shape despite the one issue, I thought that maybe you're both willing to try a number of things or to spend some money on diagnostics, or both.
My familiarity is greater with Motronic 1.0 and LH systems, but 1.3 isn't that foreign.
I'm spotting three areas of research desirable in your detail of the system's present behavior.
1 - While you don't yet know whether the stumble is misfire-related or simply air control related, you know that it's related to opening the throttle. For no money, you could eliminate the possibility of the ECU itself causing the stumble. Rig your throttle position sensor such that it is still connected, but that the ECU doesn't receive the signal that the throttle is against the stop. See if the idle performance and throttle bypass aren't causing problems in this condition with the throttle on the stop. See if the stumble characteristics, as you take the throttle off the stop to the area where it would normally stumble, change in any way. These tests should help you figure out if the transition between idle control system active and idle control system non-active holds potential fruit in tuning.
2 - Your indication that the system's idle performance is the same with the vacuum line to the fuel pressure regulator installed or removed is grounds for further investigation. Do you have a vacuum gauge? If it were mine, I'd be curious how much idle vacuum is being produced. Most likely, your hotter camshaft, and the valve adjustment thereon, are reducing the vacuum at idle. This makes the vacuum signal to the fuel pressure regulator weaker. This raises the fuel pressure maintained in the rail, likely necessitating a fuel trim adjustment to the stock map at idle, within the range of the ECU's adjustment capacity. IF you know that the problem isn't with area of concern number 1 above, AND you know that you've changed the fuel pressure being maintained in the rail at idle, then you know something that has changed from stock. The adjustable fuel pressure regulator available and fitting in the stock location would be at least a good diagnostic firing of the parts cannon, if not a long-term durable solution. Restoring the stock fuel pressure at idle removes one variable leading to area of concern 3. You might find that by adjusting your camshaft timing or by small increases in valve/cam clearance, that you can restore substantial amounts of vacuum at idle and restore a vacuum signal that puts the fuel pressure adjustment back within the happy zone headroom of the ECU.
3 - While you have confirmed good O2 sensor functional swings at idle, you haven't been able to test and report the readings that would be present during the moments the symptom presents itself. I've found the presence of an on-board fuel mixture gauge to be valuable when diagnosing these pre-OBD systems' driveability issues. The rabbit hole gets deeper here, but in the interest of brevity and utility, if you can confirm area of concern 1&2 aren't (or are no longer) related to your symptom's presence, a correlation between mixture reading and symptom will provide you better aim with the parts cannon and wayfinding the direction of further modifications.
Best of luck, this stuff can become more education than one ever wanted along the way to reaching system performance worthy of the money you've spent on it!

[delorean175]

I thought I'd just bomb this thread with some more ideas.
Seeing as how you've employed the parts cannon a few times, and the engine management system seems generally in good shape despite the one issue, I thought that maybe you're both willing to try a number of things or to spend some money on diagnostics, or both.
My familiarity is greater with Motronic 1.0 and LH systems, but 1.3 isn't that foreign.
I'm spotting three areas of research desirable in your detail of the system's present behavior.
1 - While you don't yet know whether the stumble is misfire-related or simply air control related, you know that it's related to opening the throttle. For no money, you could eliminate the possibility of the ECU itself causing the stumble. Rig your throttle position sensor such that it is still connected, but that the ECU doesn't receive the signal that the throttle is against the stop. See if the idle performance and throttle bypass aren't causing problems in this condition with the throttle on the stop. See if the stumble characteristics, as you take the throttle off the stop to the area where it would normally stumble, change in any way. These tests should help you figure out if the transition between idle control system active and idle control system non-active holds potential fruit in tuning.
2 - Your indication that the system's idle performance is the same with the vacuum line to the fuel pressure regulator installed or removed is grounds for further investigation. Do you have a vacuum gauge? If it were mine, I'd be curious how much idle vacuum is being produced. Most likely, your hotter camshaft, and the valve adjustment thereon, are reducing the vacuum at idle. This makes the vacuum signal to the fuel pressure regulator weaker. This raises the fuel pressure maintained in the rail, likely necessitating a fuel trim adjustment to the stock map at idle, within the range of the ECU's adjustment capacity. IF you know that the problem isn't with area of concern number 1 above, AND you know that you've changed the fuel pressure being maintained in the rail at idle, then you know something that has changed from stock. The adjustable fuel pressure regulator available and fitting in the stock location would be at least a good diagnostic firing of the parts cannon, if not a long-term durable solution. Restoring the stock fuel pressure at idle removes one variable leading to area of concern 3. You might find that by adjusting your camshaft timing or by small increases in valve/cam clearance, that you can restore substantial amounts of vacuum at idle and restore a vacuum signal that puts the fuel pressure adjustment back within the happy zone headroom of the ECU.
3 - While you have confirmed good O2 sensor functional swings at idle, you haven't been able to test and report the readings that would be present during the moments the symptom presents itself. I've found the presence of an on-board fuel mixture gauge to be valuable when diagnosing these pre-OBD systems' driveability issues. The rabbit hole gets deeper here, but in the interest of brevity and utility, if you can confirm area of concern 1&2 aren't (or are no longer) related to your symptom's presence, a correlation between mixture reading and symptom will provide you better aim with the parts cannon and wayfinding the direction of further modifications.
Best of luck, this stuff can become more education than one ever wanted along the way to reaching system performance worthy of the money you've spent on it!

I appreciate the reply. I wish I would have checked this earlier today while I was in the garage. The stumble issue when giving it gas isn't really there anymore. It has a slight hesitation for a split second but then revs normal. I figured that was normal throttle response for the m20. The issue I'm chasing is a big drop in RPM's after every time a give it gas. After giving throttle the RPM's should fall and return to idle but it doesn't, the RPM's drop below idle, then bounce up above idle for a second, then return to normal. Since making adjustments to the air fuel ratio the hesitation when giving throttle went away. Or became extremely subtle like I described above. I'm still going to buy that air fuel ratio gauge demetk recommended. I forgot to ask if it requires a wide band o2 sensor. I can tell when the car is going to doing its usual thing when I start it in the morning. Sometimes the idle is kind of bouncy. 25% of the time its will idle a bit better and run normal. I'm absolutely stumped when It runs normal. I'm not sure how an issue like this can be somewhat intermittent.

I did mess around with the car today and try some things. I swapped in another ECU which was a 153 (motronic 1.1 i believe). It ran the exact same way. I adjusted the throttle stop screw 1.5 turns in which was just enough to keep the TPS from clicking and turning on the idle control system. It did not really have a stumble issue but the idle was high. While running I unplugged the ICV and it was better (as described in a recent post) but It still did have a stumble. While running I unplugged the temp sensor, and it ran kind of normal. I have verified continuity between the plug and the DME plug. So the temp signal is reaching the DME. I also tried cleaning the ICV again today. I also tested the resistance between the 3 pins and it tested ok according a thread on r3vlimited. The m20b25 ICV's are different and have a rotating flap instead of a flap you can push in. After cleaning it, no change. While running I also unplugged the AFM. It did still run, but not well. I will test out your suggestion number 1 tomorrow. I will have to acquire a vacuum gauge. I did a recent valve adjustment, but it didn't really seam to change anything, only the car feels a little more (I'm not sure how to describe it) crisp.


My car has been having this dip in the RPM's thing for a long time. The subtle misfire is a relatively new issue. I tested the ignition system thoroughly. I suspect the injectors, they are pretty old. I was eyeballing some rebuilt/ flow matched 17lbs injectors. I want to figure this issue out first. My over all goal is to go with megasquirt, then forced induction. But I want to iron this issue out before going with MS. I have learned so much by owning this car, its been 8 years so far . Its been a long and informative journey.

[demetk]

I'm still going to buy that air fuel ratio gauge demetk recommended. I forgot to ask if it requires a wide band o2 sensor. I can tell when the car is going to doing its usual thing when I start it in the morning. Sometimes the idle is kind of bouncy. 25% of the time its will idle a bit better and run normal. I'm absolutely stumped when It runs normal. I'm not sure how an issue like this can be somewhat intermittent.

It simply connects to your existing O2 sensor wire.

It's possible that drop in rpm is ICV related. To test, replace the ICV with some type of 1/2 plumbing ball valve and adjust it to 700 rpm. Then take it for a drive.

[briansmith]

I'm still going to buy that air fuel ratio gauge demetk recommended. I forgot to ask if it requires a wide band o2 sensor. I can tell when the car is going to doing its usual thing when I start it in the morning. Sometimes the idle is kind of bouncy. 25% of the time its will idle a bit better and run normal. I'm absolutely stumped when It runs normal. I'm not sure how an issue like this can be somewhat intermittent.

I will have to acquire a vacuum gauge.

If you've got the scratch, and you want to use gauge indications while driving to help with diagnostics and/or tuning, you may be better off with 270deg sweep gauges than the simple old school bar of leds style gauge. I've used one of these, it's a stepper motor gauge and works either with a traditional O2 sensor or with a wideband. It also has user defined color and warning color change and playback and it might even be able to switch on and off a remote relay, I can't recall right now. You probably need less capability than all this, but the wider sweep would be welcome. While figuring out where/how you'll mount it, create space for other gauges as well, such as stepper motor vacuum and fuel pressure.
https://www.jegs.com/i/ProParts/722/300128/10002/-1
Alternatively, Glowshift is an economy supplier of stepper motor gauges, also available in the 2 1/16" size, they're a bit noisy but they get the job done.

The somewhat intermittent symptom is very possibly not something that is sometimes failing and sometimes not, but rather some condition present that you're not monitoring. In one experience I've had, an engine was simply producing less vacuum when the weather was cold and humid, and idle was stumbly. I found a threshold value over which the car would become smooth. I found a fuel pressure at which I could extend that threshold lower. I then made cam timing adjustments and ecu idle speed changes such that even in the cold humid weather the car would produce vacuum in excess of the threshold at idle. Worked wonderfully and made the modified engine feel like a factory tune.

My car has been having this dip in the RPM thing for a long time. The subtle misfire is a relatively new issue. I tested the ignition system thoroughly. I suspect the injectors, they are pretty old. I was eyeballing some rebuilt/ flow matched 17lbs injectors. I want to figure this issue out first. My over all goal is to go with megasquirt, then forced induction. But I want to iron this issue out before going with MS. I have learned so much by owning this car, its been 8 years so far . Its been a long and informative journey.

8 years is a good haul. Suspicion can be reason enough to replace parts, but falls a bit short of a diagnosis. A 25% reduction of the symptom via some better-than-stock matched injectors could theoretically make your issue harder to diagnose rather than cure it. BTDT. Fun hobby, but not one that is precisely aimed at a cure.
I'm not sure what necessitates getting the factory engine management system working nicely before making a major system change to MS. That said, having things working correctly before piling on new modifications is usually a good way to go, I understand. Motronic 1.3 ought to easily accommodate slightly worn or mismatched injectors on this mildly modified engine. Despite the modifications being mild, enough have been made on your car at this point, in my opinion, that in conjunction with the declining reliability of replacement parts available and installed heretofore, you're too far away from a stock factory setup to consider "newly replaced" as definitive indication of "good." You may need more data than part types and replacement dates to diagnose and solve the problems here. It seems unlikely you'll find someone with your same set of conditions and modifications who has "been there and done that" and can prescribe the magic bullet. The extra data available with additional instruments may be helpful.

[delorean175]

Hey guys! I think I may have figured out the issue. I bought a vacuum gauge from Harbor Freight and hooked it up with a T way on the hose that pulls vacuum for the FPR. I was getting 12Hg of vacuum at idle. I saw on a Pelican Parts article this was too low. I knew it had to be a vacuum leak despite all the poking around I've done plus a primitive smoke test. I checked my timing just to be sure something wasn't off, all was good.

I did another smoke test. This time I googled a (better) cheap homemade setup. I used my wife's airbrush compressor set to a few PSI. I used a pickle jar, a few pieces of fish tank hosing, and a bunch of cone shaped incense. I sealed up the intake boot by the AFM and connected up my smoke line. I saw a little bit of smoke coming out from the spring on the throttle body and from the ICV. I must not have caught it the first two times because I didn't have enough PSI.

I did some research and found one post on e30tech or r3vlimited (i can't remember which one) about a guy taking about his throttle body being the source of his vacuum leaks. It took him a long long time to figure it out. I went on offerup and found a guy parting an e30 with 200k miles on it. He wanted $30 for the TB so I went for it. I installed it and hooked up my trusty vacuum gauge. It fired right up and pulled 17 ish Hg of vacuum at idle. It seams to be running (dare I say?) normal. I'm currently doing some other work underneath the car so I haven't been able to drive it yet. I'll post what happens when I do. I'm trying no to get too excited.

[tn535i]

Good news is you probably don't have any vacuum leaks. But you should check that also.

In my first response to something Kyle said I think. Well anyway, you've now proven the value of simple things like gauges rather than guessing.

I hope it turns out this was the root cause and you have it solved.

[catso]

Tenacity and the incense! A combination that not only solved your problem but will give your engine great karma in the future! Well done!

[briansmith]

Great progress, and the events and results show the value of measurements in addition to visual inspections and versus blindly replacing parts.
I rather suspect still that the fact that there was a vacuum leak played less of a role in the problematic symptoms than did the resulting vacuum signal to the fuel pressure regulator resulting from the 12" reading. Engines can be tuned to operate without stumble with 12" of idle vacuum, although that's a more difficult task with the stock engine management components. The 17" reading is well within the range that I have encountered as being totally workable on stock engine management systems of the vintage. It also seems to be a reasonable resulting reading for an engine with the collection of hard parts involved. A 300+degree duration camshaft and ITBs or DCOEs and a race header might be a different story.
Thanks for posting the follow-up!

[tn535i]

I agree you should get at least 15' Hg on an engine running well with no big problems and no big cams. My old gauge has a scale with green/yellow/red and I think somewhere in that 15-17 range is the green yellow border. In the old days I could generally set my ignition advance nearly ideal by maximizing vacuum at idle.

I would have doubted the FPR could play a role but maybe that's an interesting thought. I understand the purpose of it is to bump fuel press up as soon as you lose vacuum (off-idle or anytime throttle opens a bit) almost like an accelerator pump on an old carburetor. It's a way to add some fuel even before the computer has a chance to re-determine what it wants to do. The computers in our cars are snails pace slow compared to newer cars, especially at idle.

[briansmith]

I would have doubted the FPR could play a role but maybe that's an interesting thought. I understand the purpose of it is to bump fuel press up as soon as you lose vacuum (off-idle or anytime throttle opens a bit) almost like an accelerator pump on an old carburetor. It's a way to add some fuel even before the computer has a chance to re-determine what it wants to do. The computers in our cars are snails pace slow compared to newer cars, especially at idle.

Processing power aside, I think the oxygen sensors couldn’t provide a suitably quick actionable signal near idle for the computer to process. The carburetor analogue of the accelerator pump seems appropriate to partially explain the function. I didn’t expect the fpr to have such an impact either until I had a vacuum level sufficiently far from stock to try to tune around.

[delorean175]

I appreciate the replies you guys. So another update to the update. So I replace the throttle body with a lower mileage one from Craigslist. That made things better but not 100%. It idles really rough/ bouncing around on cold mornings. Since the new throttle body I decided to take a leap of faith and install some rebuilt/flow matched m50 fuel injectors (the newer style with the 4 hole design). I did this because I had a slight misfire before, I knew it was due to high mileage/ dirty injectors. It ran, but weird at first. Lots of stumbling. The idle wasn't quite right but it ran really smooth everywhere but idle. My ass Dyno definitely felt a difference between the old and new injectors. The m50 injectors are a nice upgrade from the stock m20 injectors. If you have a 173 dme, no chip is needed, it just needs time to adjust. I don't know about the 153 and older dme's. My old TB was sitting around, I decided to take it apart and investigate the bad vacuum leaks. It turns out there is a spot for an Oring. I decided to replace the Orings and do some modifications to the TB while I had it apart. You can see the mods below in the pics.

Anyway, so I drove around for a while on the TB from CL and it was ok. But I decided to install the old rebuilt TB and see how it runs. That was when I realized why the car was running funny. I forgot to plug in the TPS when I installed the craigslist TB. I plugged in the TPS and it runs exactly how It did before. Now the idle is smooth when cold and wavers when hot. I hooked up the vacuum gauge again and its bounces around between 12 and 15 Hg because the idle wavers. When I unplug the TPS the idle goes up slightly and it smooths out, but it runs kinda funny. The vacuum goes up to 15Hg ish. I guess the vacuum leak in the TB wasn't that big of a deal after all. Over all I think I can feel a slight difference in power up top near redline due to the mods I did. It still has that post rev bounce and a wavering Idle. But it is better, so I did some good. I'm going to do one more smoke test to see what I see. If I don't find anything I'll start figuring out my options. I'm heavily considering a low mileage m50 swap to fix my problem or getting an e60 550i. Probably the latter.



Cleaning out the old Oring in the needle bearing.


The new Oring fitting in the little gap I was talking about. The Oring fits in the other side the same way


The sized Oring that fit best


The shaft back in place


Also I decided to delete one of the useless elbows that I had capped off and cut threads and put a set screw in the hole, I used a 1/2" x 20 thread tap. The set screw is visible in a later pic.


The lower profile metal rod, the screws used to protrude out about 2mm. I would assume that would cause some minor turbulence.


The screw head side of the throttle plate. Also take note of the brass inlet length. It sticks out a bit more than I would like.


Throttle open


Shorter brass inlet (now capped off after installed on the car)


I decided to paint it while I had it apart. I decided to sand the high spots, kind of like people do to the valve covers, and intakes




All assembled, note those springs are a pain to wind when you're trying not to scratch up the fresh paint. The fatter/shorter spring takes 2 full winds, and the skinnier/taller one takes 3. I did try 2 winds but the gas pedal was way too soft for my liking. Also, make sure to wind the short/fat one first, then put the metal piece (not sure what to call it) on and lightly thread on the 11mm nut. Don't tighten because you will want some wiggle room. Then work on winding the tall/skinnier spring. The tall spring can be wound while the metal top piece is secured down with its 10mm nut, but its more difficult.


I also replaced the TPS screws with Allen head screws in hopes to be able to adjust the TPS without removing the TB. I'm able to adjust both of them now without removing the TB. Usually the right side screw is inaccessible with the TB installed. The i intakes have fatter runners and prevent you from accessing that right side screw.




before


after




Thats all I have for now folks. I'll try to make some improvements tomorrow. The paint is subtle but I figured since I have it apart might as well make it lo0k snazzy I hope you guys enjoy!

[catso]

Great write-up and a nice clean motor!

[delorean175]

Thanks man! I have been trying to keep it clean. lol After the rebuild It was completely oil leak free for about 15 k miles.


As a small update: The modifications to the TB seam to make a slight difference. Initially the new Orings didn't feel like they did much but after the car is good and warmed up it idles really nice and the stumbling is gone. But I have to warm it up all the way and drive it about It about 10 minutes to get to that point. It cold starts very quick and idles high (like its supposed) then it comes down, then about a min later it idles rough and the idle bounces. Then It smooths out as it gets to operating temp, but still stumbles if driven. Then about 15 mins into the drive its rock solid. I'm not sure what to suspect, for now I can live with it. I'm puzzled as to what it could be, It was probably a combination of things. Now that I've eliminated the TB as a source of vacuum leak I'm thinking it's something that has to thoroughly heat up though. In a perfect world I'd figure out what it is, but until then it is the way it is. I'll also do another smoke test when its cold. I recently did a warmed up smoke test and it didn't really reveal any leaks. I also did a vacuum test from cold to hot. The Hg definitely increases as it heats up. If I ever find out what it was I'll make sure to post the results.

I appreciate all the help you guys have given me over the past months.

[catso]

A few years ago I was having a similar type of stumble. At about the same time I noticed the exhaust system was starting to deteriorate so I replaced it from the cat back, but this time with pipes and muffler from a 535i. It's a bit bigger and seems to produce a bit more power. The flange that joins the exhaust system had to be modified, but that was all. The old adage seems to hold true: more air in, more exhaust out. YMMV.

[Mike W.]

I just read this from start to finish, but might still have missed something, and other things are unclear.

1. Fuel pressure. It doesn't change when disconnecting the vacuum line? It should bump up. That's a big deal. The concept is when you have high manifold pressure, it's kind of pulling fuel in, so you need less pressure. When you open the throttle, less vacuum, it needs more pressure to keep the fuel pressure/manifold pressure ratio the same.

2. O2 sensor. I would expect it to need a multi wire one, which you have put in. My 535i didn't run right with a single wire sensor I put in for testing. Subtle, but definitely not right. More importantly, it should be going up and down, around half a volt, with it connected. If it's not up and down it's not closed loop running off the O2 sensor like it should. But... hook up a meter with jumpers, plant the meter inside the car and drive it around. Safely of course, but look at what it's doing when you stab the throttle and it bogs. If it goes low as I would expect, that's confirmation it's running lean. Actually disconnect it just sitting in the driveway. Especially at idle, it should be close to .45V. Long ago I had an E21, much different FI, but it was coldblooded. I don't recall the numbers, but the O2 reading was low on open loop. You want it to be close to correct with the O2 disconnected, in an ideal world the O2 provides only a very subtle change to things. I tweaked things a bit and brought it up and it was just fine. If nothing else makes sense, hook up a remote gauge so you can watch fuel pressure when you're driving around. Looking at fuel pressure again, which FPR are you using? I would guess, and I'm not up on M20s, that the i engine management needs higher fuel pressure than the eta.

The O2 sensor can make significant adjustments to compensate for things, I suspect that's what's going on, why it runs ok, but is a problem in transition.

[delorean175]

A few years ago I was having a similar type of stumble. At about the same time I noticed the exhaust system was starting to deteriorate so I replaced it from the cat back, but this time with pipes and muffler from a 535i. It's a bit bigger and seems to produce a bit more power. The flange that joins the exhaust system had to be modified, but that was all. The old adage seems to hold true: more air in, more exhaust out. YMMV.

Very interesting. My exhaust has been untouched since I bought the car. It is in bad shape, and has been on the to-do list. It could be a source of my problems, or not helping Anyway. I do want to get a 535i dual exhaust. One factor is cost, I would love to buy a brand new one along with OBX long tube headers. One cost savings would be to buy one without a cat. I've done a little looking around so far. It looks like no one sells cat less exhaust systems. I will have to look into it some more though.

I just read this from start to finish, but might still have missed something, and other things are unclear.

I appreciate you taking the time to read my thread. So I have done some tests since my last update. So as of right now, if I do a vacuum test I get about 15Hg of vacuum. I have done a lot to trace and fix vacuum leaks. My car runs different when the FPR is unplugged now. I can definitely tell the difference in gas mileage with it unplugged. I think my TB vacuum leak was causing the issue before. It seams to be better now. The 02 sensor is pretty new, my old one wasn't bad either but I swapped it out Anyway. Its a 4 wire 02 sensor. I switched to m50 fuel injectors now so I will check out what my voltages are at the 02 sensor. My car got 26.5 mpg 2 weeks ago on a trip up to a ski resort and back. When it's warmed up all the way and driven for 10 mins the idle bounces and it stumbles after I rev it. After this point it runs and drives really good.

I tested my idle control valve, it seams to be fine at first. iirc It passed the ohms test, the flapper moves freely, and it makes the audible sound its supposed to when the key is in the on position. But one thing I seem to get conflicting information about is the flap inside. Some say its supposed to be spring loaded while others say it isn't. I'm buying another ICV from a good running car to see if its spring loaded. If it is indeed supposed to be spring loaded but has a broken spring. I could see how that would make the computer attempt to find the correct position for the flapper, but never find it because the air flowing could push it around more than intended. I'll see what happens when I swap out the ICV.

[demetk]

If you're running m1.3 then the ICV should be the 3-pin ICV, correct? It's not spring loaded like the 2-pin ICV. Remove power from a 3-pin ICV and the flap should not move.

With that cam you have I doubt if you'll see more than 15 in vacuum at idle.

Have you ever set the throttle plate to spec?

[delorean175]

Yes I am running m1.3 and it is a 3 pin ICV. I'm glad to get confirmation its not supposed to be spring loaded. Whats weird is If i unplug the ICV when not completely warmed up the idle is smoother and more stable. When heated up all the way, unplugged and it idles high. I don't know if that means anything or not.

When I reassembled the TB I didn't know what specs to set it to. When the set screw was taken out of the TB the throttle plate would slightly catch on the inner walls of the TB. I set it to be barely open, just enough to prevent the throttle plate from sticking. After trying that out and finding it didn't change anything I started moving the set screw in 1/4 turn at a time and re-adjusting the TPS switch. A few adjustments inward didn't really seem to change anything. I guess I'm stumped again. lol

[tn535i]

I wonder if some of this is not simply a result of the new cam and your pictures brought to mind a new thing or two. First come back to a few very fundamental things...

No offense to Mike above but you should NEVER get a steady O2 reading of 0.45 volts. With these basic O2 sensors it is either hi 0.9 v indicating rich or swings lo 0.1-0.2 v indicating lean or passing between hi and lo. Look again at the O2 voltage at idle and see where it is and it should be swinging back and forth. Also unplug the O2 sensor from the DME and see what it does. You want it to default slightly rich and if it doesn't you adjust the bypass screw on the AFM until it does that at warm idle. If you then plug the O2 back in it should start to control and swing hi lo.

I have probably already said it but now that you've worked on all this unplug the TPS and see what the idle does. With a 3 wire ICV it should go to it's neutral position (50% open/50% close time commanded I believe) and that should generally mean the idle speeds up a little, maybe 100-200 rpm. If NOT the T/B is too far closed and the idle control doesn't have the range needed from cold to warm operation AND the ICV is moving more than you want it to every time it goes off-idle back to neutral position. IME the ICV needs to be a little bit closed on warm idle with no load which means the idle will increase if you unplug the TPS as if it is off-idle. You also need to get the TPS idle signal to drop out just as soon as you bump the throttle open. Don't simply trust the click look at the 5V on pin 1 (I think it's pin 1 idle 5v, pin 2 ground, pin 3 WOT 5v).

Test again the O2 voltage off-idle. It should still swing plugged in or go hi if you unplug the O2 sensor. If not close the bypass a bit more until it goes hi (rich) with TPS unplugged.

Seems like you figured many things out but I'm guessing you still have an idle control range and off idle transition issue and could be too lean at idle, maybe too rich. Then again maybe the new cam and low manifold vacuum means it will just be persnickety no matter what.

Another thing on these cars I do not like is that vacuum generator on the brake booster. I think it plays with the manifold vacuum if not working right. I don't recall exactly how it works but a check valve is in there somewhere and they can go bad. Be careful with brakes and even the booster but you might try plugging it and see what it does to manifold vacuum. For example I think if bad it acts like a vacuum leak across the T/B.

That's all I can think right now.

[delorean175]

tn535i: I appreciate the response. I will run your suggested tests and post what I find. I am going to be replacing the brake booster hosing and simplifying it. Some of the e30 guys have deleted the 2 (idk why the hell 3 where necessary) extra vacuum feeds and make the booster pull from 1 inlet on the TB. The m20b27's had 1 brake booster vacuum feed, I don't know why the b25 had 3. lol I want to copy this guys setup in the pic below.

[JeanHaynes]

I actually just bought the same sensor just the Bosch 15717, so my sensor will have a 13" wires instead of 18". Which I can make work. It arrives Saturday. And the pigtail I was going to get from the local autoparts store. Thanks again for the help! Hopefully this will do the trick.