[drhousedk]

Hi,


My friend owns a 2006 E91 335d. We're experiencing some overboost issues, which makes the car go into limp mode (sorry for VVS - was needed to fit all info in the recording):

From top to bottom:
1. Ambient atmospheric pressure
2. Boost, actual
3. Boost, wish
4. Rail pressure

Unfortunately, the Delphi 2015.3 diagnostic tool can't put the charts "on top of eachother", so a on-top comparison isn't really an option, even the scalings are different.

It's overboosting quite a lot when pulling hard (.3 bar) - and we can't figure out why. The remap has been done by a highly recognized tuner, and the tuner is running the exact same map in his own 335d, with a similar setup. Overboost also happens when running a stock map, although not as much (0.15 bar approx), so limp mode isn't triggered. It seems like the boost is controlled fine for a brief period, and then the boost just rises.

(Update: Feb 11, 2017: On a test drive in frosty weather a boost pressure of 2.79 bar was registered from the DDE! Delphi 2015.3 used for diagnostics.)

Remember, to calculate actual boost, you need to subtract atmospheric pressure from the boost pressure.

Modifications done:

- DPF removed (downpipe - all sensors still present, in their respective positions)
- Swirl flaps removed
- EGR disabled in the DDE map

The following repairs has been done to trace down the issue:

- All vacuum valves replaced with brand new OEM parts.
- All vacuum hoses replaced with brand new OEM hose (even in the correct blue/red colors where specified). Replaced one by one and cross referenced with schematics to make sure locations are correct
- Turbo switchover valve (quite a b*tch to replace, actually), replaced with a brand new OEM part
- EGR and engine thermostats
- All six glow plugs including relay

Observations:

- Error code 428B and 447B are thrown
- Tried adjusting the wastegate controlling the exhaust switchover valve (no. 12 on this parts diagram: http://www.realoem.com/bmw/enUS/show...diagId=11_3963) - didn't do any good...
- Soaked IAT sensor, MAP sensor and MAF in isopropyl alcohol for 30 mins, no difference

What are we missing?


Troels,
Denmark

[335dlci]

Has the overboosting fault been happening after downpipe install?

[drhousedk]

Originally Posted by 335dlci

Has the overboosting fault been happening after downpipe install?

Seems to be limited to when the downpipe is installed.

Original DDE map (tested both with 100% stock and stock w/dpf removed) also gives overboost, just not by as much (not enough to go into limp). I've heard talk of the wastegate map needing to be remapped aswell due to the lesser backpressure - don't know if this holds up?

I also have a 330d (231bhp) with downpipe, mapped by the same tuner, where overboost isn't an issue. Actual and specified follows nicely.

[Torqu3]

I know this was an issue here in the States back in the day. I do believe you've got it right. It needs a new tune so the wastegate opens up.

[Yozh]

Certainly could be overboost related as code 428B is a negative deviation of the low pressure turbo control, too high of the charge pressure. BMW suggests looking at leaks, boost sensor as well as turbo switchover and wastegate to make sure not stuck. Says that torque reduction in effect and no DPF regen, but does not apply to your case. However, 447B is an O2 sensor code for lower oxygen concentration. I would not run out to replace the sensor right away, but it does play a role in the control strategy. Also, just because you have replaced all the vacuum lines and pressure converters, it does not mean that your turbo control system now functions correctly. You may still have leaks elswehere that will affect the control. In your case you are overboosting, but please check the wastegate actuator and it's control. Please run deisred vs. actual airmass. Also, if O2 sensor is past 200K km, may be it's time for a new one, also it it's been mishandled during removal/reinstall of the DPF/Downpipe.

[Deftronix]

Check out the work TDIwyse has been doing with a turbosmart external wastegate off the manifold egr port to negate this issue until it can be fully understood. Should at least keep the car from going into limp mode.

http://www.e90post.com/forums/showpo...14&postcount=3

[drhousedk]

Replying from mobile, sorry for the missing quotes. :/

We've vacuum tested the entire system (vacuum pump with gauge), and it seems to be within specs. Switchover (on pressure side) is replaced with a brand new OEM part.

Wastegate moves freely, as well as the switchover / regulator on the exhaust inlet on the big turbo.

We will replace the boost sensor on the manifold in the coming week, as well as the IAT sensor.

[drhousedk]

I've got a laptop with ISTA-D / ISTA+ at my disposal tomorrow.

Is there any specific tests / calibrations / live data I should fetch, that might be able to help you guys give me a hint as what to look at? :-) And can ISTA produce graphs instead of just numbers? I got it listing MAF related values today on my own 330d, but couldn't see a logging / graph feature. Did I miss something?

[drhousedk]

Two new videos.

ISTA+ measurements (note that MAF actual is 40% higher than MAF target, and boost is also way above):

A video showing the wastegate actuation - seems to be just around the 5mm movement that BMW states is necessary:

Vacuum line is fine, holds vacuum perfectly.

Replacing IAT sensor and boost sensor didn't make any difference. Reset the MAF adaptations aswell.

Still throws the oxygen sensor error (but as stated by another user earlier in this thread, we're ignoring this for now)

Any suggestions?

[Yozh]

Try adjusting the preload on the wastegate so it opens earlier/more. Also, looks like you have replaced all the pressure converters, lines etc. Have you looked at the system vaccum and monitored it's values at various points of operation. I had a small leak from the wastegate pressure converter that was not picked-up by a vacuum gauge and I only found it by chance where I heard it from the parallel extension I was running into the cabin when monitoring vacuum behaviour. Small leaks could affect proper vacuum system functionality that will have a delayed response on components on control of things such as wastegate.

[drhousedk]

Originally Posted by Yozh

Try adjusting the preload on the wastegate so it opens earlier/more. Also, looks like you have replaced all the pressure converters, lines etc. Have you looked at the system vaccum and monitored it's values at various points of operation. I had a small leak from the wastegate pressure converter that was not picked-up by a vacuum gauge and I only found it by chance where I heard it from the parallel extension I was running into the cabin when monitoring vacuum behaviour. Small leaks could affect proper vacuum system functionality that will have a delayed response on components on control of things such as wastegate.

Haven't measured it at various points, would be quite a job. I've tried pulling vacuum in the lines, and it doesn't leak at all - observed it for two minutes. IMO a leak that affects operation this much should be severe at least to a point where a vacuum drop after two minutes would be present.

I'm not sure about wastegate adjustment, but at zero vacuum the wastegate should just about close - and not "lean" onto it's flange, I'd guess? I think there was some information in ISTA+ about this, might need to read up on that.

I'm not sure what to make of the higher MAF value - could it calculate boost from a mix of MAF and more - or is the boost shown, directly from the boost pressure sensor?

By the way, the ambient pressure and MAF value at idle has +46 mg difference - BMW (ISTA+) states that it should be within +/- 25 mg. Bad MAF, or just a coincidence? The ambient pressure was 1004 according to the measuring blocks, the MAF said 1050, and the weather station around 3 kms from here says 1009, so I guess the ambient pressure sensor in the DDE is okay.

[Yozh]

MAF measures air mass not pressure. Can you post what ISTA+ says. Boost control is from the boost pressure sensor that also controls fuelling. MAF, O2 and IAT sensors are interrelated and will affect fuelling as well. Throw an EMP sensor into the mix and you can see that they are all interrelated.

[drhousedk]

By the way, I've just spotted (on the video) that the charge air temperature seems to be reaching almost 90c... That sounds really, really wrong?!

[drhousedk]

Originally Posted by Yozh

MAF measures air mass not pressure. Can you post what ISTA+ says. Boost control is from the boost pressure sensor that also controls fuelling. MAF, O2 and IAT sensors are interrelated and will affect fuelling as well. Throw an EMP sensor into the mix and you can see that they are all interrelated.

The first video in post #9 is ISTA-D values at full throttle, all MAF related values though, but I hope it is of some help. I've just noticed that charge air seems to be reaching 90C. That doesn't sound right... Or does it?

[Yozh]

Do you still have ISTA to play with? If not, looks like you have the cable. Why don't you download the wonderful TestO and then you will be able to log anything you want and graph to compare. I have noticed that is that video from post #9 (even though it is hard to see) that you boost does not start to deviate until higher into the boost levels and when LP turbo comes online fully. Your MAF deviates high all the time. I wonder what your rail pressure compares like desired vs actual. By all means the boost desired and actual and then MAF desired and actual may not follow exactly, there is always a bit of deviation, how much is the question. On my car I get a bit of deviation, but rail pressures are spot on. I am not sure on the IAT temperature, I know at full fuelling runs some get it up to 70C, but there are too many variables like intercoolers, ambient tempereatures, etc. You do know there are three air temperature sensors. One for the ambient (behind the bumper), one in the MAF and one on the cold side of the charge pipe right before the throttle plate. You can compare the three and see what they report. I know that on failing MAFs its the temp sensor that is acting up. I have not found the parameter for the wastegate yet, but it would be interesting to see where the DDE is requesting EMP bleed. Be also interesting to see what EMP is being recorded at these full fuelling runs. Do you have any leaks in the EMP sensor feed? Or any exhaust leaks especially at the manifold or turbo sides. With new downpipe fitted, I would check.

EDIT: also, there is a thread called "log requests" that is specifically to compare parameters. May be useful to request what you are interested in, but you should run TestO. I have data but no full fueling, need to drive out of the city to get it stretch it's legs.

[drhousedk]

Originally Posted by Yozh

Do you still have ISTA to play with? If not, looks like you have the cable. Why don't you download the wonderful TestO and then you will be able to log anything you want and graph to compare. I have noticed that is that video from post #9 (even though it is hard to see) that you boost does not start to deviate until higher into the boost levels and when LP turbo comes online fully. Your MAF deviates high all the time. I wonder what your rail pressure compares like desired vs actual. By all means the boost desired and actual and then MAF desired and actual may not follow exactly, there is always a bit of deviation, how much is the question. On my car I get a bit of deviation, but rail pressures are spot on. I am not sure on the IAT temperature, I know at full fuelling runs some get it up to 70C, but there are too many variables like intercoolers, ambient tempereatures, etc. You do know there are three air temperature sensors. One for the ambient (behind the bumper), one in the MAF and one on the cold side of the charge pipe right before the throttle plate. You can compare the three and see what they report. I know that on failing MAFs its the temp sensor that is acting up. I have not found the parameter for the wastegate yet, but it would be interesting to see where the DDE is requesting EMP bleed. Be also interesting to see what EMP is being recorded at these full fuelling runs. Do you have any leaks in the EMP sensor feed? Or any exhaust leaks especially at the manifold or turbo sides. With new downpipe fitted, I would check.

EDIT: also, there is a thread called "log requests" that is specifically to compare parameters. May be useful to request what you are interested in, but you should run TestO. I have data but no full fueling, need to drive out of the city to get it stretch it's legs.

EMP? Tried Googling it, but had no luck. What is this - sorry if I'm missing something obvious.

No exhaust leaks - we checked that very thoroughly, as the first fitting was done without exhaust mounting paste - so it was leaking a bit. We when removed the downpipe, checked for sooty areas, used mounting paste, and everything is nice and clean now, no soot deposits around any mounting spots.

Rail actual and desired are spot on (within a few bar).

I can borrow ISTA+ again any time, it's my buddy's car though, so I don't have it at my disposal every day - and he's not very good with a computer, so I need to be there.

I will see if I can get TestO going - I've tried it before (on another car) without much luck - that's probably my bad.

[Yozh]

EMP = Exhaust Manifold Pressure.

That is good that rail pressure is in line. I would lean towards boost control. Something does not line up to keep boost in check. Set-up TestO, it's easy and you will love it. Can check where your EMP is. I'll try to find the parameter for wastegate control. I also thought, you can run a test plan in ISTA to see how quick your turbo switchover operates. I know you have checked all of that, but may be worth it. I'm no expert, but in these issues, I think the sequence goes like this: vacuum pump, leaks, vacuum valve mechanical operation, vacuum valve holding vacuum, electrical connections, electrical integrity of EPDW and EUV valves, vacuum integrity of EPDW and EUV (vacuum leaks), rest of the system vacuum leaks (engine mounts, EGR (EU), etc), boost leaks, exhaust leaks, electrical sensors (boost, MAF, O2, IAT, EMP) functionality through logging, tune.

[drhousedk]

Originally Posted by Yozh

EMP = Exhaust Manifold Pressure.

That is good that rail pressure is in line. I would lean towards boost control. Something does not line up to keep boost in check. Set-up TestO, it's easy and you will love it. Can check where your EMP is. I'll try to find the parameter for wastegate control. I also thought, you can run a test plan in ISTA to see how quick your turbo switchover operates. I know you have checked all of that, but may be worth it. I'm no expert, but in these issues, I think the sequence goes like this: vacuum pump, leaks, vacuum valve mechanical operation, vacuum valve holding vacuum, electrical connections, electrical integrity of EPDW and EUV valves, vacuum integrity of EPDW and EUV (vacuum leaks), rest of the system vacuum leaks (engine mounts, EGR (EU), etc), boost leaks, exhaust leaks, electrical sensors (boost, MAF, O2, IAT, EMP) functionality through logging, tune.

I'll give TestO a whirl as soon as I get to the car again. Are their any specific scenarios you'd like me to log? Like 3rd / 4th gear pulls, or? Due to some unique geographical advantages, we're able to do 200+ kph pulls aswell (although limp most often sets in at around 150 kph).

Leaks should be ruled out - I've tested the system as a whole, and each and every endpoint (even engine mounts) seems to hold vacuum just fine. All valves are replaced as well. Might need to check how much vacuum the pump is producing, although braking seems fine / well assisted even after a series of hard braking. Boost and IAT sensors have been replaced. MAF temperature seems fine, responds well if we duct some cooler air to it while idling.

The oxygen sensor error was, btw, "implausibly low oxygen concentration". Don't know if this tells you anything.

I've just gotten a hint on a german E91 forum (only response yet though), that DPF removal, on a 335d, will always produce overboost due to the wastegate being only 11mm, thus relying heavily on the DPF to reduce boost. I really don't buy into that, as a lot of people are running downpipes, with no real issues to my knowledge. Also from the same tuner, using the same basic map (with minor adjustments depending on the engine's overall condition and health).

I really appreciate you taking the time to help out here - it's quite the nightmare when you don't have that much experience with newer diesel and especially the setup on the 335d.

In that aspect, I love my 330d for being relatively more simple. :P

[Yozh]

You Europeans are lucky to have great roads. In North America there is a motto that "speed kills", where really it just stupidity kills. I have only 100kph roads around here that I can do 120kph and then get problems if get caught speeding.

I thought that is addition to all the boost, MAF values it would be useful to log Exhaust Pressure, Lambda Equivalency Ratio and Exhaust Gas Temperature (need to pick the right sensor as there are three in (US) but only two in (EU). I'm thinking that you need to eliminate the possibility of exhaust pressure being read low and your O2 (lambda) sensor functioning properly. On the basic level, if O2 reads rich it should be defuelling, but if EMP is on the low side it will affect boost control and keep that on the high level without wastegate bleed or late wastegate bleed. Also, there is a second DPF pressure sensor, I wonder how is that handled within the tune. Would be interesting to log that one to see what values it outputs. Otherwise, than the last thing to consider is that the new tune did not consider a better flowing exhaust and did not adjust expected boost values. I apologize for so many guesses in so many different directions, but this is just from my limited experience of trying to diagnose issues with my car that ended up being my O2 (lambda) sensor.

[drhousedk]

Originally Posted by Yozh

You Europeans are lucky to have great roads. In North America there is a motto that "speed kills", where really it just stupidity kills. I have only 100kph roads around here that I can do 120kph and then get problems if get caught speeding.

I thought that is addition to all the boost, MAF values it would be useful to log Exhaust Pressure, Lambda Equivalency Ratio and Exhaust Gas Temperature (need to pick the right sensor as there are three in (US) but only two in (EU). I'm thinking that you need to eliminate the possibility of exhaust pressure being read low and your O2 (lambda) sensor functioning properly. On the basic level, if O2 reads rich it should be defuelling, but if EMP is on the low side it will affect boost control and keep that on the high level without wastegate bleed or late wastegate bleed. Also, there is a second DPF pressure sensor, I wonder how is that handled within the tune. Would be interesting to log that one to see what values it outputs. Otherwise, than the last thing to consider is that the new tune did not consider a better flowing exhaust and did not adjust expected boost values. I apologize for so many guesses in so many different directions, but this is just from my limited experience of trying to diagnose issues with my car that ended up being my O2 (lambda) sensor.

Fun fact, the roads in Denmark where the general speed limit on the motorway is 80 mph, the asphalt is a lot better and a lot smoother, compared to the german autobahns, where there in some places are no speed limit at all (although 75 mph is recommended).

I've been on the autobahns a couple of times, and it is quite a peculiar feeling overtaking a police car when going 140+ mph, without having to think about license, tickets etc. etc...

I actually think the EU car has got only two temp sensors (pre and post cat/DPF), and only one pressure sensor. There's only one pipe with a tube coming from it running to a pressure sensor. I'm pretty sure the pressure sensor reads 0 at all times (that's from free memory, and might very well be absolutely wrong).

Furthermore there is the O2 sensor, which is throwing a code after just a few minutes driving, which I find odd.

This is identical to the pipe we're running:



What was the symptoms when your car had the issues?

[Yozh]

You know, I always wondered why european tuners like Ecotune built their pipes with little feed tubes to the pressure sensor, now I know. You only have one pressure sensor, the one before DPF, but on the downpipe. But I do not know what control strategy it employs, is it for turbo control or just for DPF regeneration control of both. On US cars there are two pressure sensors, one just the exhaust pressure sensor right off the manifold for turbo control, and one DPF differential pressure sensor for DPF regen control with two feed tubes, one before and one after the DPF. I would be very curious to see what values are reported by your only exhaust pressure sensor. It seems to me it should be fully functional.

Concerning your O2 sensor, I think you pulling the values and equivalency ratios is in order.

My car had issues because of a plugged SCR cat and an bad tune by Ecotune. I feel that elevated exhaust pressure induced damage to O2 sesnsor that is only rated to 4bar by Bosch.

BTW, I tried a wastegate parameter, but it did not report any variable value, still searching.

I have done a quick pull in third/fourth gear but only to 145kph before my wife would get uncomfortable (she is good at holding the laptop though).

Here is some info:

First graph is of exhaust pressure, requested and actual boost, you can see where it shifted from third to fourth. X axis is RPM. Second graph is EMP to Boost ratio. You can see that my actual boost is a bit higher than expected, but is only a few hundred hPa (3psi) difference and does not throw a limp mode.





Last summer when I was in Germany I did fully enjoy the roads in a rented MB GLE350d, except it's brakes were off an A-classe and it would not do more than 200kph even in Sport mode. May be it was somehow restricted as a rental.

[drhousedk]

Originally Posted by Yozh

You know, I always wondered why european tuners like Ecotune built their pipes with little feed tubes to the pressure sensor, now I know. You only have one pressure sensor, the one before DPF, but on the downpipe. But I do not know what control strategy it employs, is it for turbo control or just for DPF regeneration control of both. On US cars there are two pressure sensors, one just the exhaust pressure sensor right off the manifold for turbo control, and one DPF differential pressure sensor for DPF regen control with two feed tubes, one before and one after the DPF. I would be very curious to see what values are reported by your only exhaust pressure sensor. It seems to me it should be fully functional.

Concerning your O2 sensor, I think you pulling the values and equivalency ratios is in order.

My car had issues because of a plugged SCR cat and an bad tune by Ecotune. I feel that elevated exhaust pressure induced damage to O2 sesnsor that is only rated to 4bar by Bosch.

BTW, I tried a wastegate parameter, but it did not report any variable value, still searching.

I have done a quick pull in third/fourth gear but only to 145kph before my wife would get uncomfortable (she is good at holding the laptop though).

Here is some info:

First graph is of exhaust pressure, requested and actual boost, you can see where it shifted from third to fourth. X axis is RPM. Second graph is EMP to Boost ratio. You can see that my actual boost is a bit higher than expected, but is only a few hundred hPa (3psi) difference and does not throw a limp mode.

It's actually quite interesting, sollwert of your boost is around the same as on my buddys car (2.1 bar in euro currency ), but yours can keep the boost at 2.4 bar, which, I think you are absolutely correct, wouldn't cause limp mode. We only see limp when boost has been around 2.8 bar. No wonder - that's 40 psi...

Thank you very much for your logging, I'm positive that's going to be very helpful! Has TestO got EMP built-in a DDE7 template, or? I'm sorry for the many questions.

Originally Posted by Yozh

Last summer when I was in Germany I did fully enjoy the roads in a rented MB GLE350d, except it's brakes were off an A-classe and it would not do more than 200kph even in Sport mode. May be it was somehow restricted as a rental.

Probably, IIRC the GLE350d should be able to do 225 kph.

BMW's On Demand vehicles certainly aren't restricted further than what the factory does. July 2016 I went to Münich, where I rented an M4 droptop. It went straight to 270 kph (where you could feel the limiter easing off the throttle. That's power for you!)