[Garageaholic]

Hey Guys - I'm at the tail end of an N54 Swap into an E36 and after a number of test drives, I realize that I do not have ANY boost being produced going past the throttle body into the intake manifold.

I can hear the turbos both when I rev and when I put the car under load (wot). And I can confirm I have + boost at the TMAP sensor. So I know the turbos themselves are creating boost, but it's like the boost isn't getting into the intake manifold, past the throttle body, like the throttle body isn't opening up enough to allow that positive pressure to build up in the intake.

So I can rev the engine all I want and can hear the turbos whine all day long, and when I give it load, the turbos spool up but no acceleration or boost.

Now I have the ANTunes Harness hooked up for now (CAN Emulator), and the system is on CAS3 with key/dme/button. These are the only codes I get (In order of priority as shown in INPA)

CDBE - Idle Speed Control Valve CAN Bank 2
2A18 - DTML Pump Heater
2F71 - (should not cause limp mode)
2F81/2F83 - VANOS solenoids (I will check/clean/inspect as necessary)
Updated 8/15/19 - Actually, they are:
2F83 - Engine Switch off time, no matching fault system
2F81 - Engine Switch off time, damaging exhaust gas
2FA4 - (should not cause limp mode)
2E8E - (should not cause limp mode)
2AAF - (should not cause limp mode)
2F4F - Vehicle Speed Sensor

(those above that said "should not be in limp mode" are confirmed by ANTunes that they are normal due to non-communication from those other DME's, so I'm accepting that it is not causing or contributing to the issue)

Here's what I checked:

Throttle body and gears all look good. No corrosion, pushed pins, connection issues.
Vacuum Lines for leaks
Obvious charge pipe leaks
All electrical connections look good
Turbo Wastegate Actuation looks good
Reset all adaptations that I could, reset codes. the above codes are all that continually come back.


This is my post of the build if you were interested!
https://www.e90post.com/forums/showthread.php?t=1605973

Thank you guys so much for your help. I really appreciate it. Hope that my contribution in my build thread inspires you all to try something like this yourselves.

Thank you!!!
Frank

[feuer]

2FA4 - I think you will need to reflash the ECU (possibly ECU isn't set for 335i)
2AAF - possible lack of fuel (if so ECU will keep throttle closed to protect engine damage)

[mad1stgen]

The only codes from that list that will cause a limp mode, are the vanos codes.

The 2FA4 is due to your CAN emulator most likely. 2AAF is meaningless by itself.

Wire in the original Kombi, and skip the emulator for now.

[Garageaholic]

Originally Posted by feuer

2FA4 - I think you will need to reflash the ECU (possibly ECU isn't set for 335i)
2AAF - possible lack of fuel (if so ECU will keep throttle closed to protect engine damage)

The supplier of the harness said those codes (not bolded) are OK to run and shouldn't put the car in limp mode.

I have my low pressure fuel pump (in tank walbro 255lph) running full speed on switched power (not using EKP with this harness)

Originally Posted by mad1stgen

The only codes from that list that will cause a limp mode, are the vanos codes.

The 2FA4 is due to your CAN emulator most likely. 2AAF is meaningless by itself.

Wire in the original Kombi, and skip the emulator for now.

The kombi is the the body controller that's integrated into the fuse panel under the dash, right?

[Garageaholic]

Actually, i stand corrected. Those 2 VANOS codes are not VANOS codes at all. they are actually:

2F83 - Engine Switch off time, no matching fault system
2F81 - Engine Switch off time, damaging exhaust gas

[vtl]

2FA4 - Incorrect data record. What have you got flashed on the DME? This has happened to me when I ran the PPK flash on my car without modifying the BIN to bypass the security check. Basically limits the power so the car is barely driveable. Is this what you're experiencing?

Have you got MHD? Can you do a datalog?

[Garageaholic]

Good points! i don't have MHD I don't have an android device. I can't do it on iphone I see...

[feuer]

Originally Posted by Garageaholic

The supplier of the harness said those codes (not bolded) are OK to run and shouldn't put the car in limp mode.

I have my low pressure fuel pump (in tank walbro 255lph) running full speed on switched power (not using EKP with this harness)

I don't know what ECU would think if there isn't EKP to control the fuel flow.


The kombi is the the body controller that's integrated into the fuse panel under the dash, right?

Kombi is instrument cluster. Light module is by fuse panel.

.

[dpaul]

I think vtl is correct - you've likely got the wrong zusb flashed on your DME

[Garageaholic]

This was from a stock 335i with a cobb tune only. It was in a crash and I took the drivetrain out and put it into the E36. The engine was working perfectly before I took it out.

I ONLY have CAS hooked up (with key and start button), So I understand there will be communication codes.

I bought a harness that was to trick the DME with all CAN codes necessary to get the car out of limp mode. But I feel like it still is.


So I checked the wastegate diaphragms. They hold strong no leaks, and full close at 7psi.

I checked the boost solenoids and they held vacuum and idled at 12psi, so they were good too.

Again, this engine worked perfectly before it was removed, so I don't think it's a piece part issue. Think there is something in one of those fault codes. Thought that maybe I hooked something up wrong, or forgot to tighten a clamp, but no.

So what I did is hook up a boost gauge to the actual charge pipe right before the throttle body. Then I hooked up the vacuum gauge on the intake manifold. Just a simple racing of the engine and I saw that boost was easily generated on the charge pipe but not getting into the intake manifold.

I will do a quick test drive and see how much boost is actually NOT getting into the intake manifold. I think this is a control issue.

[Garageaholic]

Oh yes, and this CAN Emulator also fools the DME into thinking the EKP relay is there, so LP fuel pump is running on switched power constant on.

[Garageaholic]

Anyone have any ideas? I'm going for another test drive today, plan to install boost gauge on both intake manifold and charge pipe monitoring what's going on. Then getting more fault codes as needed. Could definitely use the help.

I'm the kind of guy that once the problem is found, i document it here in as much detail as I can, to help the community learn from my swapping mistakes and setbacks.

Thanks again for your help,

[georgelb]

Maybe I'm misunderstanding something, but your video appears to be working correctly. You have a vacuum gauge that is showing vacuum at idle, reducing near zero when the throttle plate opens, at which point boost starts to build.

When you let off the accelerator the throttle plate snaps shut, vacuum increases, and boost tapers off. If you're looking for a boost reading from the other side of the throttle plate you'll need to hook the boost gauge to that side.

[Garageaholic]

Originally Posted by georgelb

Maybe I'm misunderstanding something, but your video appears to be working correctly. You have a vacuum gauge that is showing vacuum at idle, reducing near zero when the throttle plate opens, at which point boost starts to build.

When you let off the accelerator the throttle plate snaps shut, vacuum increases, and boost tapers off. If you're looking for a boost reading from the other side of the throttle plate you'll need to hook the boost gauge to that side.

Ok, so you are right, but more news today!

It works! (but only for the first 60 seconds). I didn't do anything different, but I know what changed. I used to let the car idle before taking it out. But now I just started it up and took it out. It worked amazingly for the first 60 seconds.
...But after the first 60 seconds, the car went into Limp Mode, or "reduced power" mode and it wasn't producing boost, presumably the wastegates were held open by the boost solenoids and the result was reduced boost to the intake manifold as a protective manner. So I pull over to the side of the road, key off, wait 5 sec, key back on, and the car was back up to full power! Then 60 seconds, limp mode. later rinse repeat. I timed this repeatedly and got it down to exactly 60 seconds, which is a pretty whole number, so sounds like it was purposely programmed by BMW to reduced power mode.

The codes before and after are shown below (nothing different than described earlier). Within the first 60 seconds:

0x2AB4 DME INTERNAL ERROR
0X2F58
0XCBDE THROTTLE
0X2A18 DTML
0X2F71 E-BOX
0X2F83

After the 0:60 second mark, all the same as above with additional:
0x2FA4 (THIS MIGHT BE MY ISSUE - Usually this means the zusb number is not compatible with the Dme, this is now my focus)
0x2F81
0x2E8E INTELLIGENT BATTERY SENSOR

Youtube video of the behavior coming Wednesday. Thank you again for all your help!

From Boost Addicts, I read this to address the 0x2FA4 code. Can you verify if this would fix the issue?

"I'd recommend to first find the latest PPK ZB number for your car using the kmm_ATSH.txt file in SP Daten (There are different ZB numbers to suit different combos of 135i, 335i, AT, MT, Euro, US etc.), flash that ZB number to your DME using WinKFP. From here, your car should run but at low power and have 2FA4 and probably 2FA3 error codes (You can ignore 2FA3). From there, read the DME with BBFlash, Change 4C9C8 from 00 to 01, then flash the BIN to the DME. That should do the trick nicely."

[dpaul]

Originally Posted by Garageaholic

The codes before and after are shown below (nothing different than described earlier). Within the first 60 seconds:

0x2AB4 DME INTERNAL ERROR
0X2F58
0XCBDE THROTTLE
0X2A18 DTML
0X2F71 E-BOX
0X2F83

After the 0:60 second mark, all the same as above with additional:
0x2FA4 (THIS MIGHT BE MY ISSUE - Usually this means the zusb number is not compatible with the Dme, this is now my focus)
0x2F81
0x2E8E INTELLIGENT BATTERY SENSOR

From Boost Addicts, I read this to address the 0x2FA4 code. Can you verify if this would fix the issue?

"I'd recommend to first find the latest PPK ZB number for your car using the kmm_ATSH.txt file in SP Daten (There are different ZB numbers to suit different combos of 135i, 335i, AT, MT, Euro, US etc.), flash that ZB number to your DME using WinKFP. From here, your car should run but at low power and have 2FA4 and probably 2FA3 error codes (You can ignore 2FA3). From there, read the DME with BBFlash, Change 4C9C8 from 00 to 01, then flash the BIN to the DME. That should do the trick nicely."

First things first: tell us if the DME, CAS and key all come from the same car that the engine was removed from.

[Garageaholic]

Originally Posted by dpaul

First things first: tell us if the DME, CAS and key all come from the same car that the engine was removed from.

Fair question - And yes it was. All from the same car. I had the car at my disposal so I took it all out. It had a Cobb Tune but no JB4 or piggyback added on.

This harness I purchased connected all the necessary wires to interface the CAS to the DME including the Z_USB wire, START, and the BU_PA wire.

Thanks

[dpaul]

Thus the ZUSB must be correct. Of course, the Cobb flash not only changes the parameter files but makes changes in the DME programming to prevent you from reading the parameter files. So, just guessing here but the simplest explanation is some sort of incompatibility between the Cobb flash and the CAN filter/emulator built into your aftermarket harness. Why not just use WinKFP and flash appropriate stock files back to the DME? You've got nothing to lose.

[Garageaholic]

I plan to do JUST THAT tonight. Appreciate all your help with this standby!

[Garageaholic]

So I removed the Cobb Flash, removed the CAN Emulator. I now have many more fault codes, but ran the same tests with the same results (after 60 minutes, the 2FA4 code comes back and car goes into reduced power mode).

So admittedly I am struggling with WINKfp. I have the complete toolset, including INPA and have used INPA for much of the fault codes, but I don't have much experience with WINKfp.

I know it's beyond the purview of this thread subject, but I'm looking for a bit of a tutorial on how I can take the BIN, Change 4C9C8 from 00 to 01, then reflash the BIN to the DME.

Finally, I am able to keep the engine running while actively clearing the fault codes and the car operates normally until another minute goes by and it goes into reduced power again.

FYI: the Car is a WC73
The hardware P/N on the DME itself is 7604567
The hardware P/N from INPA readout is 7616431

I am not sure but feel like a mismatch of these P/N's is not good, regardless if it was this was before the drivetrain was removed from the E92.

Anyway, If I can get just a bit of a leg up for WinKFP Flashing, I think that would be a huge help. Thank you,
Frank

[dpaul]

Don't worry about the numbers on the DME itself.

Just use WinKFP in comfort mode, press "update ZUSB" and let it select the appropriate ZUSB automatically.

HW, SW and ZUSB numbers don't relate necessarily to anything actually printed on the module.


Maybe more info that you want or need but here are brief descriptions of BMW nomenclature relating to WinKFP and the daten files relevant to its function:

ZB-NR (ZUSB) specifies an allowed combination of TYP-NR, HR-NR and SW-HR files for a given module.

HR-NR (hardware number) specifies a program file (file extension "0pa" in the daten). This is the actual program that the DME runs.
It's not a number stamped on the module. I would normally think of this as software, and think of the module itself as hardware but that is not the way BMW uses the terms.

SW-NR (software number) specifies the filename of the data or calibration or map file (file extension "0da"). It is not a "program" that the DME runs. It is a set of data tables used by the program running in the DME with parameters related to engine operation (i.e. ignition timing, fueling, boost, VANOS etc). This is file that is typically altered by tuners.

DAT file gives the combination of 0pa and 0da used for a particular ZUSB

HWD file gives complete list of 0pa filenames, starting with first and ending with most current for the particular module

HIS files gives ZB old, ZB new pairs i.e. update sequence of factory revisions

Finally, my guess is that what you found on Boost Addicts is not relevant or at least not the first thing to get into. If it is not clear to you, be aware that you cannot make arbitrary alterations in either program or data files because the DME will not accept them unless their "signature" is altered or the code in the DME which checks the "signature" is altered first. Defeating this protection was a major accomplishment of some very accomplished software engineers/tuners, respect to those folks. So you have to use either MHD, KESS, BB flash (although specified in the ***********s thread, not that stable and not supported) or some other aftermarket system to flash altered files.

[Garageaholic]

Thank you! Great education. And yes, I plan to issue a detailed youtube video about all of this. It seems to be scattered amongst youtube videos with no sound and the many many forums that have different info in different formats.

So a couple of straightforward questions about WinKFP regarding and your info. I have version 5.3.1.

1. Did you mean HWH instead of HWD? I see that in my DATEN FILES.
2. What is the folder that WinKFP looks for when it starts up? When I click Comfort mode, all I get is F1 Identify ECU Family, F2 Choose Custom File and F3 Update ECU. I don't see anything related to ZUSB
3. What exactly is "assembly line data" in WinKFP as it relates to *.0pa, *.0da, and DAT files?
4. So how does produce a BIN file? is there a program that takes the *.0pa and *.0da files and put them together into a single BIN file?

Generally, I'm having trouble with it autodetecting. I have the Daten files for E89.

Thanks for your continued support. I am still trying to figure out how to use this to either remove or fix the 2FA4 code.

And I don't think MHD is the answer to that level of flashing. I really want to better understand the technical navigation of the daten files through NCSexpert, and WinKFP

[digitalashley]

How did you go in the end? What solved your 60second into limp mode issue?