BMW M52 (Turbo) rough idle after ECU firmware update

[Xaphan]

I recently acquired this car from a friend who built it. He had a friend from the track tune it, and the car ran well. The car never idled perfectly, but it idled where they set it and it was all good. Fast forward to when i got the car: I drove it for a week or so, and I was eager to learn and improve. When i first plugged in the ECU to my pc, I backed up the tune. Then I checked if there were any updates to download for the G4X. Turns out the ECU was a few versions behind. I read through the changelog, and nothing in red pertained to my tune so I proceeded to update the ECU. The update went through just fine. After that I calibrated the TPS and MAP. Next, I loaded up the original tune and fired it up...

When the car first started, it idled up to where it was set (1200 rpm i believe, for that ect) for a few seconds and then began to idle down, then up again, then down, and so on. I changed nothing in the original tune. I've tried changing a few values around in the idle control settings, none of which stop this erratic idle. 

I'm not a tuner, and I'm still learning. But I can't figure out why the exact tune with no mechanical changes is causing this idle issue after a firmware update. Any ideas or help would be greatly appreciated!

• BMW M52 Turbo (GT35)
• Link G4X ECU
• Link LAMBDA
• ISC still there
• car seems to feel fine when you rev into the higher RPMs
• I suspect conflict between limits or values in the idle parameters (but I'm no tuner so...)

 

SHORT YOUTUBE VIDEO OF CARS IDLE STATE, AND LIVE INFO FROM PCLINK

 

m52 tune 1 (JAK).pclx m52 idle after update.llgx

[Laminar]

Your idle valve is opening, trying to get the RPMs from where they start around 780rpm up to your target of 1356rpm. But as the idle valve opens, MAP increases and once you hit your MAP lockout of 50kPa, the system kicks out of closed loop idle control, and initiates a new cycle.

Check out the idle configuration of the E36 base map (C:\Link G5\PCLink G5\Sample Maps\). It looks like your tune still has the stock Idle Base Position Table values, but the Idle Target RPM Table has been increased far above stock. The point of the base position table is that it's an idle valve position that's approximately what you need to achieve your target RPM. But because the base position table was set up for an idle RPM lower that what you're running, the base position is too low and the PID loop has a bunch of work to do to achieve the right idle.

Increase your MAP lockout to like 90kPa to get it out of the way for now. Then do some testing to find the idle valve position that gives you stable idle at each ECT value in your table, and populate the Idle Base Position table with those numbers. You can bring the MAP lockout back in if you want it, it stops the ECU from entering idle mode when you're not idling, but your TPS should take care of that so the MAP lockout isn't a crucial function.

[Xaphan]

@Laminar, 

first, thank you so much for taking the time to help me. 

I will adjust the tune according to your suggestion and see how far I get. I'm still lost on essentially what happened during/after the update. Either way, Ill try this and report back as soon as I can. 

THANK YOU VERY MUCH.

[Xaphan]

So, I adjusted the MAP value and that stopped the surging. Sucess! Thank you for that.

Ive been watching some webinars and combing through forums and I believe that I have a better grasp on the process and I will continue to test and post results. Unfortunately, I work for the next 2 evenings but I'm going to try to spend at least an hour after work with the car until my off days. 

I will try: re-do idle control settings in open loop, lower my RPM lockout value, experiment with dashpot, experiment with ISC values, and turn on closed loop after I populate the idle base position table with corrected ISC values. I will report back with progress and results.

 

 

[Xaphan]

Progress. The car no longer stalls, but is far from smooth and won't maintain a target rpm.  I even used the compare feature to make sure nothing changed in the tune or perimeters. Using @Laminar advice, the MAP lockout valve change did help. Repopulating the idle base table helped, but the car still runs rough, not at all like it was .Clearly there is something else going on. 

I still can't understand what changed.

Latest log attached 

PC Datalog - 2024-08-30 8;11;56 pm.llgx

[essb00]

Still reaching the MAP lockout rpm on the 2nd log. You need to increase MAP lockout RPM further (maybe try 75kPa). You also need to fix the idle base position values to be closer where it sits with stable idle. Idle actuator integral gain 0.5, idle ignition control proportional gain 1.0.

...or perhaps copy all the idle parameters off the E36 Xtreme Plugin sample map as a good starting point.

Another thing affecting the idle - lambda/afr seems too lean. You need also to adjust the master fuel.

[Xaphan]

after work i will try your advice. if i dont have luck, ill attempt to try the e36 base idle settings. if none of that work, i'm reverting versions until someone that knows what theyre doing can look at it.

i really appreciate the help. 

[Adamw]

You can use the file compare function to see any changes or new features after an update, but by the looks of it your problem is more that idle control has never been tuned rather than anything related to the update.  Everything looks miles away from what it looks like it needs.

I would suggest you check base timing though as depending how old your firmware was there was a change in 6.20 that could change the trigger offset. 

But assuming your timing is correct then I have made some adjustments to your idle settings to get it closer in the map attached.  

m52 tune 2.pclx

[Xaphan]

@Adamw thank you for doing that. The compare feature was very helpful. I learned about it in webinar and have been using it since. I'll try your corrections later today when I get off work.  

I reverted back to the 6.22.8 firmware and loaded up the original tune that I received the car with. It will now hold idle. Not perfect, better than before. Very well could have been the trigger offset, ill look at it. 

I attached a couple of logs from last night after I got the old firmware on the ecu. 

 

 

PC Datalog - 2024-08-31 7;47;30 pm.llgx PC Datalog - 2024-08-31 7;50;27 pm.llgx