[ajm55]

Some background:

This is my first post in this section. Not what anyone wants from a first post but it is what it is.

I bought an n54 335i (2008) with 100,000 kms a few months back and in the period leading up to the engine failure described below, I did the following (or rather had the following done) in the way of repairs/ maintenance:

- New valve stem seals
- New OE HPFP
- New OE valve cover and oil housing gaskets
- One step colder NGK plugs
- Carbon clean

Performance Mods:

I then added the following performance mods:

- Catless downpipes
- BMS 500 hp kit (which includes a JB4 unit) with the 7.5 inch upgraded intercooler

Upgraded ECU and new injectors and coil packs:

After installing the performance mods, I downloaded a JB bef using the MHD App.

About 2 days after loading the bef, the mofsets in the (MSD80) ECU fried.

Out of an abundance of caution, I upgraded the ECU to an MSD81 ECU and at the same time installed new OE index 12 injectors and new OE coil packs

Engine Failure:

Within a week of installing the MSD81, new injectors and new coil packs, the motor was hurt while doing a WOT run in 2nd gear. I hadn’t reloaded the backend flash as I wished to do some testing on Maps 2 and 5.

We haven’t yet stripped the motor but when the incident occurred, a CEL popped up indicating a misfire on cylinder 6 and compression and leak down tests subsequently done indicate a major problem with that cylinder.

Data Logs:

I attach 2 data logs.

The first is a log of a 3rd gear run done on Map 2 on 24 September 2021.

There is no issue with the car on this run and this log gives insight into how the car should run on map 2 with the stock tune.

The second log is a log of the 2nd gear run done on Map 2 on 26 September 2021 when the engine damage occurred.

I’m no expert but it seems to me that just before 5,000 rpm the car starts running lean and that at about 5,000 rpm the ECU tries to protect the motor by commanding reduced boost (3.8 psi) while the JB4 targets 12.9 psi and the car delivers 12.7 psi. The problem is then compounded from there with the JB4 targeting (and the car delivering) high boost amid lean AFRs notwithstanding the ECU’s attempt to limit the boost and protect the motor.

I emailed the logs to Terry at BMS. He proclaims that nothing in the log appears problematic. He knows a lot more than me about interpreting logs, but in my layman’s view, his view is simply (and obviously) wrong.

As appears from the log, map 4 was engaged but way too late - long after the damage was done.

Going forward:

We are going to strip the motor and possibly rebuild depending on what we find, but unless we can identify and address the cause of the lean condition and reason for the failure of the JB4 unit to heed the ECU’s attempt to avert imminent danger, I’ll be risking the new motor

I’d appreciate the insights of the experts on the forum

[NoGuru]

Sorry to hear about this. Can you upload the logs to Datazap?

[CarAbuser]

That certainly looks like the JB4 messed up. You're correct in saying that the ECU wanted to target reduced boost but the JB4 just kept on regardless.

Pretty worrying that Terry said that's normal behaviour. One of his sales slogans is that the JB4 has additional safeties in place but this clearly shows it also has additional risks too.

The JB4 is only in control of the boost so once the DME started to target a lower boost value, the JB4 just lied and reported a lower boost value to the DME but kept targetting 14PSI anyway. That also explains why the DME is targetting such a lean AFR since it believes it's at a much lower load.

The reason for the DME trying to reduce boost like that could well be related to a severe knock event. I don't know if the load reduction would have saved it, but the JB4 just decided to send it anyway.

[ajm55]

https://datazap.me/u/ajm55/n54-335i
https://datazap.me/u/ajm55/n54-335i-24-sept-21

Thanks NoGuru. Not sure if I've done this correctly

[ajm55]

Originally Posted by CarAbuser

That certainly looks like the JB4 messed up. You're correct in saying that the ECU wanted to target reduced boost but the JB4 just kept on regardless.

Pretty worrying that Terry said that's normal behaviour. One of his sales slogans is that the JB4 has additional safeties in place but this clearly shows it also has additional risks too.

The JB4 is only in control of the boost so once the DME started to target a lower boost value, the JB4 just lied and reported a lower boost value to the DME but kept targetting 14PSI anyway. That also explains why the DME is targetting such a lean AFR since it believes it's at a much lower load.

The reason for the DME trying to reduce boost like that could well be related to a severe knock event. I don't know if the load reduction would have saved it, but the JB4 just decided to send it anyway.

Thanks for your feedback. Yes, your assessment resonates with me. I interpret the log to indicate a disconnect between what the ECU was trying to achieve and how the JB4 dealt with that

[Saif2018]

Originally Posted by CarAbuser

That certainly looks like the JB4 messed up. You're correct in saying that the ECU wanted to target reduced boost but the JB4 just kept on regardless.

Pretty worrying that Terry said that's normal behaviour. One of his sales slogans is that the JB4 has additional safeties in place but this clearly shows it also has additional risks too.

The JB4 is only in control of the boost so once the DME started to target a lower boost value, the JB4 just lied and reported a lower boost value to the DME but kept targetting 14PSI anyway. That also explains why the DME is targetting such a lean AFR since it believes it's at a much lower load.

The reason for the DME trying to reduce boost like that could well be related to a severe knock event. I don't know if the load reduction would have saved it, but the JB4 just decided to send it anyway.

I always thought JB4 was safer due to the "Safeties" looks like it deserves to be in the dumpster.

[Saif2018]

Originally Posted by ajm55

Thanks for your feedback. Yes, your assessment resonates with me. I interpret the log to indicate a disconnect between what the ECU was trying to achieve and how the JB4 dealt with that

When you rebuild the engine, get rid of the JB4 and go with a reliable MHD custom tuner.

[NoGuru]

Target dropped once you let off the throttle but boost stayed and even went up.
I am not sure this is a JB4 issue. Could be a stuck wastegate.

[Bimmer_Bro]

Originally Posted by NoGuru

Target dropped once you let off the throttle but boost stayed and even went up.
I am not sure this is a JB4 issue. Could be a stuck wastegate.

I agree with the above. Boost should not climb once you have let off the throttle.

If you recently installed downpipes I would make sure the v band clamps are not coming into contact with the waste gate arms.
Also check your bov or DV's to make sure they are venting the pressure properly.

[CarAbuser]

Originally Posted by NoGuru

Target dropped once you let off the throttle but boost stayed and even went up.
I am not sure this is a JB4 issue. Could be a stuck wastegate.

This is the timeframe in question:
https://datazap.me/u/ajm55/n54-335i?...lo=5&zoom=0-28

DME_BT is DME Boost Target and ECU_PSI is what the DME is seeing as actual boost. The boost and target params are the actual values that the DME never sees.

[Bimmer_Bro]

Originally Posted by CarAbuser

This is the timeframe in question:
https://datazap.me/u/ajm55/n54-335i?...&zoom=0-28

DME_BT is DME Boost Target and ECU_PSI is what the DME is seeing as actual boost. The boost and target params are the actual values that the DME never sees.

Hmmm.....

[NoGuru]

Sorry, I am not a JB4 guy but I do see what you mean.

[CarAbuser]

Originally Posted by Bimmer_Bro

Hmmm.....

That's not relevant to what we are discussing. There's plenty of reasons the boost reading might spike after letting off the pedal but his wastegates getting stuck on his vbands is probably the least likely.

[Sgop335]

Op, it ran lean for sure. But in the damage log i notice after 2nd gear it was wot in 3rd as well with bank 2 afr pegged at 20.
When i had persistent cyl 6 misfire, i had bank 2 lean conditions as well perhaps shorter duration, not sure. Did finally fix it w/o damage.
Also were your injectors coded?

[gster109]

Originally Posted by CarAbuser

That certainly looks like the JB4 messed up. You're correct in saying that the ECU wanted to target reduced boost but the JB4 just kept on regardless.

Pretty worrying that Terry said that's normal behaviour. One of his sales slogans is that the JB4 has additional safeties in place but this clearly shows it also has additional risks too.

The JB4 is only in control of the boost so once the DME started to target a lower boost value, the JB4 just lied and reported a lower boost value to the DME but kept targetting 14PSI anyway. That also explains why the DME is targetting such a lean AFR since it believes it's at a much lower load.

The reason for the DME trying to reduce boost like that could well be related to a severe knock event. I don't know if the load reduction would have saved it, but the JB4 just decided to send it anyway.

This is hilariously sad... JB4 is doodoo.

[Dave92N54]

Sorry for your loss OP.

This is one of the reasons I never felt safe going with a piggy back. Best case, it runs pretty well, worst case, miscommunication and boom. Not that flashed cars can't blow, but they don't blow from the dme trying to do one thing, while the piggyback wants to do something else.

[RSL]

Originally Posted by ajm55

Some background:

This is my first post in this section. Not what anyone wants from a first post but it is what it is.

I bought an n54 335i (2008) with 100,000 kms a few months back and in the period leading up to the engine failure described below, I did the following (or rather had the following done) in the way of repairs/ maintenance:

- New valve stem seals
- New OE HPFP
- New OE valve cover and oil housing gaskets
- One step colder NGK plugs
- Carbon clean

Performance Mods:

I then added the following performance mods:

- Catless downpipes
- BMS 500 hp kit (which includes a JB4 unit) with the 7.5 inch upgraded intercooler

Upgraded ECU and new injectors and coil packs:

After installing the performance mods, I downloaded a JB bef using the MHD App.

About 2 days after loading the bef, the mofsets in the (MSD80) ECU fried.

Out of an abundance of caution, I upgraded the ECU to an MSD81 ECU and at the same time installed new OE index 12 injectors and new OE coil packs

Engine Failure:

Within a week of installing the MSD81, new injectors and new coil packs, the motor was hurt while doing a WOT run in 2nd gear. I hadn’t reloaded the backend flash as I wished to do some testing on Maps 2 and 5.

We haven’t yet stripped the motor but when the incident occurred, a CEL popped up indicating a misfire on cylinder 6 and compression and leak down tests subsequently done indicate a major problem with that cylinder.

Data Logs:

I attach 2 data logs.

The first is a log of a 3rd gear run done on Map 2 on 24 September 2021.

There is no issue with the car on this run and this log gives insight into how the car should run on map 2 with the stock tune.

The second log is a log of the 2nd gear run done on Map 2 on 26 September 2021 when the engine damage occurred.

I’m no expert but it seems to me that just before 5,000 rpm the car starts running lean and that at about 5,000 rpm the ECU tries to protect the motor by commanding reduced boost (3.8 psi) while the JB4 targets 12.9 psi and the car delivers 12.7 psi. The problem is then compounded from there with the JB4 targeting (and the car delivering) high boost amid lean AFRs notwithstanding the ECU’s attempt to limit the boost and protect the motor.

I emailed the logs to Terry at BMS. He proclaims that nothing in the log appears problematic. He knows a lot more than me about interpreting logs, but in my layman’s view, his view is simply (and obviously) wrong.

As appears from the log, map 4 was engaged but way too late - long after the damage was done.

Going forward:

We are going to strip the motor and possibly rebuild depending on what we find, but unless we can identify and address the cause of the lean condition and reason for the failure of the JB4 unit to heed the ECU’s attempt to avert imminent danger, I’ll be risking the new motor

I’d appreciate the insights of the experts on the forum

If you're going to run a JB4, don't ever do it without a BEF and if you have to flash anyway, screw adding JB4's mess into it. If JB4 was left on 4/2 with no BEF, it probably made it worse. It at least used to be that JB4 ran the higher of it's target or DME target, which sucks when the DME cuts boost to save your bacon and JB4 says, what? JB4 also runs higher targets on the preset maps if BEF is selected AFAIK.

It was obvious something wasn't meshing well on the log from days earlier and might've been avoidable if it was noticed and something done about it. Trims are tapped out and on both logs, average ignition is through the roof.

On the previous log, it was at least in the 110-120 load range, which is tuned for maybe 5-7psi on a stock bin, not the 14-15psi JB4 is running. When DME cut boost on the catastrophic log, it dropped to only 80-90 load for 14-15psi, which is even leaner AFR and higher timing targets. To compound the issue, the DME is calculating even less fuel with no drop in true fuel needs, which increases trims to try to make up for the larger difference. That's why trims are high on the first, but completely pegged on the second.

No idea why JB4 didn't safety until the second time you tried to get into. It probably should've hit lean safety during the pull, but obviously DME tried to save you LONG before any of JB4's highly touted garbage kicked in. It was a combination of DME trying to act, JB4 ignoring everything and running in low and lower boost tuned areas for not low boost.

Just a random stock bin to show AFR and timing targets at 80-125 loads while you're running no BEF at 14-15psi. Tuning in the stock bin at loads that would be 14-15psi without lying to the DME are wildly different and much, much safer.

EDIT: stupid attachments

[ajm55]

Many thanks for all the feedback. This is much appreciated.

RSL and CarAbuser, your input is at a technical level that may well exceed my comprehension, so I ask whether I am correct in understanding that you surmise as follows:

a) The trigger was probably an underlying issue (unconnected to the JB4) causing the car to run lean

b) Once the lean condition manifested, the ECU's attempt to protect the motor from harm was overridden by the JB4

c) Whether the ECU would have saved the motor but for the intervention of the JB4 is uncertain but what is certain is that the JB4 did not help but rather exacerbated the position.

PS: I forgot to mention in the originating post that the turbos had also been refurbished and fitted with larger compressor wheels. Not sure if that is relevant but I mention it for the sake of completeness.

Also, the new injectors had been coded by a mate. I thought that the coding may have played a role because no decimals were used but my mate explained that the torque pro tool used by him to do the coding obviates the need to manually introduce the decimals, so that theory went out the window

[CarAbuser]

Originally Posted by ajm55

Many thanks for all the feedback. This is much appreciated.

RSL and CarAbuser, your input is at a technical level that may well exceed my comprehension, so I ask whether I am correct in understanding that you surmise as follows:

a) The trigger was probably an underlying issue (unconnected to the JB4) causing the car to run lean

b) Once the lean condition manifested, the ECU's attempt to protect the motor from harm was overridden by the JB4

c) Whether the ECU would have saved the motor but for the intervention of the JB4 is uncertain but what is certain is that the JB4 did not help but rather exacerbated the position.

PS: I forgot to mention in the originating post that the turbos had also been refurbished and fitted with larger compressor wheels. Not sure if that is relevant but I mention it for the sake of completeness.

Also, the new injectors had been coded by a mate. I thought that the coding may have played a role because no decimals were used but my mate explained that the torque pro tool used by him to do the coding obviates the need to manually introduce the decimals, so that theory went out the window

I don't understand all the JB4 logging parameters so I'm not sure how much useful information is in the logs, probably not enough to pinpoint the blame. The values for timing and fuel trims are nonsensical to me but if RSL said the trims were high already then the problem could have just been caused by leaning out.

I would be confident in saying the JB4 had a hand in the destruction though. The DME was clearly trying to implement protection but the JB4 was blocking it. A lot of the diagnostic routines in the DME rely on airflow and load values to pick up on fault conditions, the JB4 just gets in the way of that.

The injector coding seems suspect though. I wasn't aware you could do anything useful with Torque Pro. Was it Pro Tools maybe? He is right about the last digits of the injector coding, I think the injector flow values are stored in a 16bit register so when you enter the entire 6 digit calibration value there is some loss of fidelity so the last couple of values don't always match up with what you enter.

[RSL]

Yeah, JB4 logging sucks on top of everything else. It's slow, have to do weird math to get some real values and it can't show negatives, so 25 on JB4 trims is 0% and 50 is 34% (max). 0-24 is negative.

In the most simple terms, JB4 just controls boost and DME has no idea how much is really being run. The entire problem almost certainly revolves around running that much real boost on bigger turbos at such a low load area in a stock bin.

Unless the injectors were completely questionable, the lean and trims could just be attributed to the above. The DME has coding limits for injectors. If he tried to input values completely out of range (like a missing decimal), it shouldn't accept them. I haven't used that app, but if it says no decimal needed and the values took, it's probably accurate.

2 things are clear: DME boost dropped and JB4 boost didn't. We may never know why for either, but if DME dropped its target and load itself, there would almost certainly be codes. If you don't have auto clear on, I'd try pulling them, but there's probably a mess of them at this point, if there are any.

It's combination of factors between settings, how the JB4 works, what was being run, etc. It's hard to point to one thing as a fault. All we can really see is what was going on leading up to it.

[ajm55]

Originally Posted by CarAbuser

I don't understand all the JB4 logging parameters so I'm not sure how much useful information is in the logs, probably not enough to pinpoint the blame. The values for timing and fuel trims are nonsensical to me but if RSL said the trims were high already then the problem could have just been caused by leaning out.

I would be confident in saying the JB4 had a hand in the destruction though. The DME was clearly trying to implement protection but the JB4 was blocking it. A lot of the diagnostic routines in the DME rely on airflow and load values to pick up on fault conditions, the JB4 just gets in the way of that.

The injector coding seems suspect though. I wasn't aware you could do anything useful with Torque Pro. Was it Pro Tools maybe? He is right about the last digits of the injector coding, I think the injector flow values are stored in a 16bit register so when you enter the entire 6 digit calibration value there is some loss of fidelity so the last couple of values don't always match up with what you enter.

Thanks for the further insights. Yes, I suspect the trigger was the leaning out of the AFRs. However, I have no idea why the AFRs leaned out. Brand new injectors, coil packs and HPFP and one step colder plugs less than 2,000 kms "old". And the "lean condition" appears to have manifested on both banks, rather than only the bank housing cylinder 6. Logically therefore that would rule out fault with any of these parts. Not sure what to explore from there.

As for the coding of the injectors, I think you're correct. My mate used Pro Tool. Apparently when coding injectors using Pro Tool, one simply inputs the 3 digit numbers on the injectors without adding decimal points and the App does the rest

[RSL]

Originally Posted by ajm55

Thanks for the further insights. Yes, I suspect the trigger was the leaning out of the AFRs. However, I have no idea why the AFRs leaned out. Brand new injectors, coil packs and HPFP and one step colder plugs less than 2,000 kms "old". And the "lean condition" appears to have manifested on both banks, rather than only the bank housing cylinder 6. Logically therefore that would rule out fault with any of these parts. Not sure what to explore from there.

As for the coding of the injectors, I think you're correct. My mate used Pro Tool. Apparently when coding injectors using Pro Tool, one simply inputs the 3 digit numbers on the injectors without adding decimal points and the App does the rest

It was targeting pretty lean because of the load area the DME was working in, but the calculations are the problem. Using completely made up numbers, DME is calculating say 175g/s of air (and fuel to match) based on the low boost it's seeing, while there is actually, say, 350g/s of air coming in that needs fuel. The difference between calculated fuel and actual fuel need is done on the fly with trims from O2 feedback, and trims have a limit. If it reaches that limit and still needs more, you're out of gas, literally.