S54 EML question

[BimmerBreaker]

Some backstory, I was driving my S54-swapped "2.8" coupe aggressively awhile back and after hitting full throttle the car threw an EML light and went into limp mode. I pulled over and restarted - the EML light went away and the throttle response actually felt better than it ever had before. No check engine lights (note: my traction control is not working, so the light is always on so not sure if that went on or not). Over the next week the throttle became progressively more laggy - more "drive-by-wirey" before throwing another EML and going into limp mode again. Again, restarting it made it feel much better but I haven't driven the car since... I checked the throttle position sensor on the front of the engine as I read those are the most common culprit - a friend and I did some tests on it with a multimeter to confirm it was functioning properly. The previous owner had hose clamps on the intake manifold but those have been on there for a long time... can't imagine they'd be to blame?

Some nice ITB and CSL style box porn for ya'll







I took the manifold off today to do some servicing to it as it is aftermarket and I noticed a couple spots where there is a vacuum leak and I'm going to test the other TPS too, but if its the pedal I haven't found any way to test that part to ensure its bad before just buying a new one? Most research leads me to E46 land and their setups are slightly different. Is it possible to test the sensor and/or would someone with diagnostic software be able to tell me specifically what sensor is bad? I assume the pedal part is 13621407446 which isn't hugely expensive... part of me says to replace them all (the TPS sensors are part 13637840383 and it needs 2) as I'd be in it about $400 and that should clear up any throttle lag. But I'd prefer to isolate the problem first and go from there...





I am also curious what the two lines above are for. The one in the first image was just fished loosely into a hole cut in the intake manifold and the other one attached to a metal end that was RTV'd in place. I'd like to get something I can screw in from the back to give a firmer connection, with just a dash of RTV to seal it up. Same for the bottom - open to any ideas before I harvest things from ACE hardware to make it happen. They are vacuum lines that are out of sight so I'm not overly concerned. The second hose connection also broke off while removing, can it be replaced?

[Shoe Coupe]

Some backstory, I was driving my S54-swapped "2.8" coupe aggressively awhile back and after hitting full throttle the car threw an EML light and went into limp mode. I pulled over and restarted - the EML light went away and the throttle response actually felt better than it ever had before. No check engine lights (note: my traction control is not working, so the light is always on so not sure if that went on or not). Over the next week the throttle became progressively more laggy - more "drive-by-wirey" before throwing another EML and going into limp mode again. Again, restarting it made it feel much better but I haven't driven the car since... I checked the throttle position sensor on the front of the engine as I read those are the most common culprit - a friend and I did some tests on it with a multimeter to confirm it was functioning properly. The previous owner had hose clamps on the intake manifold but those have been on there for a long time... can't imagine they'd be to blame?

Some nice ITB and CSL style box porn for ya'll







I took the manifold off today to do some servicing to it as it is aftermarket and I noticed a couple spots where there is a vacuum leak and I'm going to test the other TPS too, but if its the pedal I haven't found any way to test that part to ensure its bad before just buying a new one? Most research leads me to E46 land and their setups are slightly different. Is it possible to test the sensor and/or would someone with diagnostic software be able to tell me specifically what sensor is bad? I assume the pedal part is 13621407446 which isn't hugely expensive... part of me says to replace them all (the TPS sensors are part 13637840383 and it needs 2) as I'd be in it about $400 and that should clear up any throttle lag. But I'd prefer to isolate the problem first and go from there...





I am also curious what the two lines above are for. The one in the first image was just fished loosely into a hole cut in the intake manifold and the other one attached to a metal end that was RTV'd in place. I'd like to get something I can screw in from the back to give a firmer connection, with just a dash of RTV to seal it up. Same for the bottom - open to any ideas before I harvest things from ACE hardware to make it happen. They are vacuum lines that are out of sight so I'm not overly concerned. The second hose connection also broke off while removing, can it be replaced?

that second hose that broke might be causing a vacuum leak enough for many issue to happen. It basically drains oil blowby from the airbox back to the oil pan. The material they are made of gets brittle and they break. This happened on mine. You CAN replace it but it's not cheap. What I did on mine was rescue the specific ends, cut out the brittle solid tubing and replaced it with equal size vacuum rubber hose. Fixed the problem and as far as I can tell, it fixed my misfire issue.

[BimmerBreaker]

That hose only broke during removal, was fine when the car was being driven. There was a small vacuum leak from the hose on top, but I cant imagine it'd have been substantial enough to cause a problem regarding the throttle system. I'd imagine a vacuum leak that severe would at least throw a CEL before throwing an EML, no?

[Randy Forbes]

Okay, first of all, seeing that c/f airbox made me decide that I hate you, so any "help" from me may not be trustworthy...

I don't know what you're using for diagnostic software, but considering the depths you get into, you should have some decent. I'm completely spoiled by Autologics, and there is a full function test of the DBW system that not only verifies movement at 0 & 100%, but points in between as well as response time in ms. I'd run that before throwing parts at it (likewise, a smoke test of the inlet tract for tracking down any breaches in sealing.

In summary, I have never had to replace a TPS or pedal assembly, and I've seen a few (I do recall several cases of the pedal assemblies being replaced under warranty when the cars were relatively new, but that too may've just been parts chucking and not bothering to figure out what was wrong).

[BimmerBreaker]

The CF airbox is definitely what sealed the deal regarding me buying this car I had a laptop with diagnostic software but I never learned to use it and traded it to a friend awhile ago. The extent of my current diagnostic software is an OBD2 scanner. But between this and the S52 I'm thinking some proper software may be in order - a quick scan of the Autologic website doesn't give me an option to buy - do you have a recommendation on a particular unit (and maybe where to buy it?).

I would think that it was some sort of fluke the first time and the result of a loose sensor the second time but it seemed there is some learned memory in there - after restarting it it cleared the code and the throttle really felt better than it ever had. So I'm thinking something, somewhere in the system is amiss. However, this may be due to a vacuum leak that had grown and affected throttle response? Its never thrown a CEL (knock on wood) but maybe its a small enough leak that the system is just learning to introduce some lag into the throttle and then flooring it made the vacuum leak more substantial, triggering something? The reason I took the intake manifold off was to seal it up again where the two halves meet, and now with the box off I've noticed the bottom was just a constant minor vacuum leak. My current plan - especially knowing how easy it is to take off - is to seal it all back up and see if any symptoms resolve.

Guess I'll just find any nipples that those hoses will attach to and throw em on?

Follow-up question as well - to remove the front throttle position sensor I had to remove the engine lift bracket. The bracket threads into the thermostat housing though, so I'm wondering if I need to bleed the system or if it will self-bleed? Can't imagine a lot (if any) air got in as I just unscrewed and screwed back in, but still something I'm curious about. I know S62's self-bleed (to an extent, anywhoo), does it's I6 bretheren do the same?

[Randy Forbes]

With regard to the t/stat hsng, I just keep thumb pressure on it while removing replacing the long bolts. If you didn't let any water out, I doubt you let air in, nor would it require bleeding (though if you didn't have heat at idle__but then, how would you know...?).

The Autologics is an expensive tool, and they just released a completely new architecture/software. IF you were to find a used one__I was VERY lucky__you'd want one of the "blue boxes" or even one of the older black ones, and it would need to have the BMW-specific software loaded. I talked to a US/Florida rep about the new ones (which handle all brands at a cost of around $12-13k) and he readily admitted that for our older models, the blue box is a better choice, with more functionality/testing, etc. It was a stroke of luck, but I found a legitimate one on ebay, loaded with only the BMW software (+ generic OBD-II) for $6900.00; I had already resigned myself to buying a new one ($14k including one__1__manufacturer) and just got lucky. Sure, there are NOW much, MUCH less expensive routes to take, but for me, I'm smitten with the Autologics intuity__like zero learning curve whatsoever.

As for this threads topic, I really don't have enough to go on to even hazard a guess. The one (1) car I had here with the c/f airbox had the fitting(s) on the underside dis-bonding, so I epoxied at least one, maybe two back in place. Always start with the simple stuff (and isolation) first, then when there's no other diagnostic choices left, swapping in know good parts can point you in the right direction (ideally, from a know-running car, not the dealer's parts counter)

[Itsablurr]

In summary, I have never had to replace a TPS or pedal assembly, and I've seen a few (I do recall several cases of the pedal assemblies being replaced under warranty when the cars were relatively new, but that too may've just been parts chucking and not bothering to figure out what was wrong).

Certainly advocate a complete diagnosis up front, as you recommended. However, in the case of EML lights going into limp-mode, fixed upon re-start, both my E46 M3 and former S54 M Roadster were caused by a faulty TPS, in particular the one that resides at the front of the throttle bodies. Browsing through m3forum as well shows this fault, and typically this sensor position to be fairly common with age.

[BimmerBreaker]

Certainly advocate a complete diagnosis up front, as you recommended. However, in the case of EML lights going into limp-mode, fixed upon re-start, both my E46 M3 and former S54 M Roadster were caused by a faulty TPS, in particular the one that resides at the front of the throttle bodies. Browsing through m3forum as well shows this fault, and typically this sensor position to be fairly common with age.

My research led me the same way but my front TPS tested good. If it comes back after sealing and reinstalling the airbox, I may just go ahead and replace it.

With regard to the t/stat hsng, I just keep thumb pressure on it while removing replacing the long bolts. If you didn't let any water out, I doubt you let air in, nor would it require bleeding (though if you didn't have heat at idle__but then, how would you know...?).

The Autologics is an expensive tool, and they just released a completely new architecture/software. IF you were to find a used one__I was VERY lucky__you'd want one of the "blue boxes" or even one of the older black ones, and it would need to have the BMW-specific software loaded. I talked to a US/Florida rep about the new ones (which handle all brands at a cost of around $12-13k) and he readily admitted that for our older models, the blue box is a better choice, with more functionality/testing, etc. It was a stroke of luck, but I found a legitimate one on ebay, loaded with only the BMW software (+ generic OBD-II) for $6900.00; I had already resigned myself to buying a new one ($14k including one__1__manufacturer) and just got lucky. Sure, there are NOW much, MUCH less expensive routes to take, but for me, I'm smitten with the Autologics intuity__like zero learning curve whatsoever.

As for this threads topic, I really don't have enough to go on to even hazard a guess. The one (1) car I had here with the c/f airbox had the fitting(s) on the underside dis-bonding, so I epoxied at least one, maybe two back in place. Always start with the simple stuff (and isolation) first, then when there's no other diagnostic choices left, swapping in know good parts can point you in the right direction (ideally, from a know-running car, not the dealer's parts counter)

Some coolant did get out as I wasn't aware that this was a pass-through to the coolant passage (poor design imo). Should I bleed the system, if so - maybe just crack the bleed screw, let it run till it doesn't bubble anymore? I don't see the need for a full bleeding procedure for what was likely a small amount of air if any that got into the system? And seeing the price of the Autologics I'll keep an eye out for a good deal on one. Given that I have no experience with these sorts of things, something easy to use like you describe the Autologic would be great for me.

I'm thinking I will test the rear TPS I have access to now with a multimeter, and then seal up the CF box with new fittings for the vacuum lines and go from there. If the code comes back, I will then diagnose further. Hoping it may have indeed just been some odd problem due to a minor vacuum leak becoming more significant at higher loads (still - wouldn't this throw a CEL?) but its strange it had not done that until now, then within a week threw two codes. The bottom of this airbox had no fitting at all for the lower vacuum line and it was just fished into the intake manifold. I'll epoxy a fitting into place for that, epoxy the other fitting as it came loose, and then go from there. Hoping that will resolve whats going on.

[BimmerBreaker]

Got it all buttoned up, no vacuum leaks now - drove it to a gas station and throttle response was not great. Babied it around locally until oil temps were proper as I was going to give it the beans and see if I could replicate the EML. Well it threw an EML at about 1/2 throttle before it got up to proper temp. Shut it down, restarted. EML was gone but throttle response was still bad (prior to today when this happened, restarting made the throttle response very sharp). Drove it another block or two and got another EML under very light throttle - this time accompanied with a CEL. Actually a bit of a sigh of relief as now I have something to actually diagnose - to an extent anyways. There are two codes showing on my code reader, the other is an EVAP system code - car has had this since I bought it, happens when it overfills as the gas filler won't automatically click off when full - never comes on when filled just a tick under full. Just hadn't driven the car in awhile and forgot about that "feature"



So I have a P1637 code. Doesn't indicate circuit A, B, bank 1 or anything on my code reader. Some research tells me this may be the TPS on the front near the oil filter housing. I may just go ahead and start with replacing that, unless anyone else has other advice?
Doing too much research is giving me a headache. Some people are reporting that they have to sync the throttle body motors (book says 8 hours!?) but only when accompanied with a P0121 and P0221. Not finding a lot for just a lone P1637.

First I will remove the sensor harness and clean the contacts... maybe this is just a poor connection. One can hope. Open to any advice here before I start wantonly throwing money at the car

Edit: OEM TPS sensors appear to be about $50 at AutohausAZ right now. I'm going to buy 2. I also noticed that one of my vacuum lines has again popped off so I'm thinking I will go back to the stock intake box for a bit and see if I can track down the problem before putting the CF intake back on. If the TPS does not cure it, I will look at the pedal and if that doesn't fix it, will get it to a shop as maybe it needs the throttle actuator synced. I fiddled with the throttles a bit when the intake was off because making vroom vroom noises while opening the throttles seemed like the right thing to do.
I also bought a used pedal potentiometer on eBay. So I will have all the pieces, hoping this fixes it...

[BimmerBreaker]

Front TPS did not fix it, will revert to OEM airbox and replace the throttle actuator TPS while I'm in there. If that doesnt fix it ill replace the pedal potentiometer, if that doesnt fix it, its off to a shop and probably will need the throttle actuator recalibrated

edit:



Good news everyone!

The TPS sensor on the throttle actuator itself appears to have fixed the problem. Throttle felt very tight and direct after replacing it and it revved freely whereas before it felt a bit laggy and hesitant to rev even without the EML light on. Because it was the TPS on the throttle actuator itself, I'm thinking I may go ahead and get the throttle actuator and throttle bodies synced as I'm thinking a problem may have manifested in that region. Couldn't hurt and my local indy doesn't charge a lot for the job.

For anyone diagnosing a similar problem in the future, all my research online led me to first replace the TPS in the front of the engine by the oil filter housing, then the TPS under the intake manifold, then the pedal potentiometer (or for E46's I believe its the complete pedal assembly). Glad I did it this way and I think that was the correct diagnostic route to take - the TPS near the oil filter housing appears to fail most often and is extremely easy to access (note: do NOT try to remove it without remove the engine lift bracket - you are very likely to strip the lower screw).

Am going to run my multimeter (well, essess' multimeter...) over the bad TPS to see what sort of readings it gives to help anyone diagnosing TPS problems in the future.

Cat on my car for posterity's sake



edit: All the TPS sensors tested within about the same range that the E46fanatics thread indicated a healthy TPS would operate within. I am going to clean the contacts and install them back to see if it may have simply been a contact issue.

[E.Hands]

I seem to be getting the same symptoms and EML that resets on restart. I appreciate your insight and for sharing your experiences. Do you still feel it was the lower TPS?

[BimmerBreaker]

It was definitely the lower TPS, I have been daily driving the car and it has not come back and throttle response is very sharp. I would start with the upper one by the oil filter housing and if that doesnt fix it then replace the lower one under the manifold, if none of those fix it, look towards the pedal.

[BimmerBreaker]

Well it was definitely not the lower TPS... at least that wasn't the whole story...

The EML came back about a week ago, putting me into limp home. Limped it home, left it. Restarted it and it was fine for a few days, no EML at all. Was on my way home Thursday to pack a small bag for Dorkfest and it threw another EML and again limp mode. I have a friend with INPA and the full diagnostic suite installed and when we plugged in this car a few weeks ago it had some shadow codes for the pedal potentiometer. Figured with the new TPS sensors that's all it could be anyways... so when the EML came back with increasing regularity I ordered a pedal potentiometer. Good thing it got here today because the car has been in limp mode with the EML light every time I've started. It used to clear on reset and would pop back on a few minutes into a drive. But today on my commute it was on from the get-go. Not fun to be passed by Geo Metros in an S54 powered Z3... And on the way home it threw a CEL as well with the accompanying codes for pedal potentiometer... P1221 or something



The part got here today and after fiddling - a true understatement, if anyone has to do this job I hope you have small hands and an offset 10mm wrench - I got the new one in. I didn't clear the codes I just went and started it. EML popped on for a few seconds then went away. Revved it, EML came back. Cleared the codes, EML and CEL went away... revved it, just the EML came back. The pedal felt more responsive even in limp mode so I decided to drive it thinking it might need some calibration or something. Prior to this it had no modulation at all in limp mode, full throttle gave about 10% throttle, anything else was a dead spot. Drove it and it started to get a bit better. Cleared the EML a few blocks into the drive, pedal felt great again, very responsive... drove it more, gave it some hard revs (had been avoiding driving it too hard recently as driving it hard seemed to cause the problem more often than not). All seemed OK and EML did not return as of yet.

So - knock on wood - now that I've replaced everything in the EML system I'm hoping the EML is gone for good. If it comes back, I'll look into getting my throttles calibrated. That is something that should be done when replacing the lower TPS based on my research, but I haven't found a shop nearby I trust to do the job - my normal shop says its not something they usually do. But maybe I can DIY it as there are some guides on it out there.

Sorry for the tl;dr I just know the more info one provides the more helpful it may prove to someone in the future looking at this thread to diagnose their own problems. My S54 has over 150k miles on it FWIW.
And just as a note S54 Z3's share their pedal potentiometers with E39 M5's (do note though that when replacing the potentiometer you need to verify you put the arm on in the same orientation as what came out of the car. I used an E39 M5 potentiometer and it was rotated from where mine was and new units may not come with an arm pre-installed at all)

[Chamberlin]

Sure do love my S52 cable throttles!

Glad you finally made some headway! Calibration shouldn't be too bad though, usually consisting of rotating the pots until the proper voltages are seen, or swinging the pots from zero to 100% while in some sort of calibration routine mode.

[BimmerBreaker]

Sure do love my S52 cable throttles!

Glad you finally made some headway! Calibration shouldn't be to bad though, usually consisting of rotating the pots until the proper voltages are seen, or swinging the pots from zero to 100% while in some sort of calibration routine mode.

Yup this job (and whole process) made me miss the S52's quite a bit, but once running well I do at this point prefer the S54. Maybe I'm just saying that because my own S52 hasn't run in so long...

I'm sure you are right about the calibration. The only problem I foresee is this is an engine swap so a few things here and there are "off" when looking at it in INPA. Will look into the basic steps of calibration so I can get to the calibration menu and see how everything looks.

[Randy Forbes]

Sure do love my S52 cable throttles!

Glad you finally made some headway! Calibration shouldn't be to bad though, usually consisting of rotating the pots until the proper voltages are seen, or swinging the pots from zero to 100% while in some sort of calibration routine mode.

Sounds very analog. When I've put different engines and/or induction on S54 models, the Autologics diagnostics that I use finds and sets the end points (caveat: the individual butterflies, if disturbed have to be set manually), it also checks the activation and response times; it's all very simple and quick.

If your guy knows his way around the INPA stuff, should be able to fly right through it (I might have an old laptop I got off ebay with DIS/INPA loaded as VMs, but it was so cumbersome compared to the Autologics, that I never bothered to familiarize myself with it).

[BimmerBreaker]

After many months of the EML being absent after replacing the pedal, the EML came back after I recently did some work on the car. Thinking maybe vacuum leak related I took a close look at everything in the intake system. There were a few cracked throttle body boots, so replaced them all. Ensured lower vacuum lines were seating properly and installed the OEM intake back on to remove vacuum leaks from my CSL box as a possible cause of the EML.

No change. Its way worse than before, it won't clear on resetting the engine. Have tried battery reset and the throttle reset procedures. Have replaced both engine bay TPS and pedal potentiometer. I have another pedal potentiometer I am putting in tonight. We have not had a chance to calibrate the throttle bodies yet, we don't really know how, but we will try giving it another look in INPA tonight.

[dannyzabolotny]

So I went over to BimmerBreaker's house the other day and we plugged in my laptop. We initially tried INPA, which kept bugging out, so we switched to PA Soft and were able to clear the EML. After that, the EML hasn't come back. I'm guessing it just needed to be reset when the new pedal and TPS were installed way back when.

[BimmerBreaker]

Yes PAsoft was the only thing that was able to connect and reset the adaptions. I'm not ready to call it "fixed" yet but it hasn't come back in a week and throttle response is better than it's ever been... keeping my fingers crossed

At least I have the S52 to drive if the S54 bugs out again...

[Randy Forbes]

Yes PAsoft was the only thing that was able to connect and reset the adaptions. I'm not ready to call it "fixed" yet but it hasn't come back in a week and throttle response is better than it's ever been... keeping my fingers crossed

At least I have the S52 to drive if the S54 bugs out again...

Autologic diagnostic software and the S-54 get along well too

[BimmerBreaker]

The S54 has resumed daily driver duties (my X5 just ate it's rod bearings om nom nom)

And today this stupid EML came back on. But I noticed something, it seems to be related to voltage. The EML came on today idling at a red light, I think once the fan kicked on to cool it down is when the EML came on. The fan kicking on maybe drew just enough power that the throttle sensor didn't receive proper voltage for a second. The codes I got in the past were for throttle pedal low voltage... or insufficient voltage or something. I don't think I got a P1637 since I replaced the TPS sensors, I'm getting a P0121 now (thats the throttle sensor low voltage code) - but I'm getting it alone. Research is indicating that a P0121 alone may not indicate a throttle position sensor problem - those are accompanied, usually at least, with the P1637 (or related codes) I had before.

The electronics in the car are very sensitive it seems, the battery is OK, so I'm thinking maybe this is an alternator issue at this point. I've had other suspicions related to electrics in this car in the past, included suspecting a weak alternator for various reasons (heck, rolling down the windows causes the headlights to dim). And because the EML has not come on under the circumstances it used to (used to come on during high throttle... I drove the car hard earlier and couldn't get an EML... but it came on at idle now... which it wasn't doing before) I am thinking this is a separate issue to what I was chasing before

I have managed to get this car hooked up to INPA so maybe I should try to force the EML while INPA is logging. Though I think the alternator needs to be replaced anyways, so I will do that first and go from there

edit: After further postulation, I am going to buy one of those fancy phone app OBD2 things so I can see what the voltage is at idle and in different conditions. This should determine if my theory that the alternator is on it's way out is accurate or not before I spend the money on the alternator. Plus it seems like a handy tool to have

[Mr Bingley]

Put the battery on charge and bring it to full--then see if the symptoms abate for a period. Are you sure the battery is good?

[BimmerBreaker]

Positive the battery is good. But it is interesting you mention that... I used to run a different battery in this car, which is now in another car. This S54 is now running an actual BMW branded battery and the EML has not been on with this battery once (in only a few drives though, to be fair). Until today. And last night I accidentally left the dome light on in that car... started ok, but now you have me wondering if that overnight drain from the light, coupled with a weak alternator, while being pretty early into the drive (engine was not even fully warmed up yet, so had only been babying it on surface streets) all came together just right to trip a low voltage EML...

My buddy has a torque app thing I am going to borrow. I'll get to the bottom of this

I will also do as you suggested and pull the battery to charge it out of the car and see if symptoms are maybe alleviated tomorrow. I may also try to start stressing the electrical system at idle to force an EML, that may confirm my suspicions as well...

[cleanerPA]

Even if the battery is good, with a bad alternator, you're going to run down the battery and end up with low voltage issues. I'd test the alternator output before going much further.

[BimmerBreaker]

According to the cluster test the alternator will sit at 13.5 at idle, up to maybe 13.8 when revving hard and I'll see 12.6 when it first dips down into idle. This is with a very healthy battery, I am keeping it charged every night off the car. Since doing so I have not had the EML once (knock on wood)