NefMoto

I'm having some issues with idle hanging sometimes.  This is not that big a problem by itself, but when I get off the throttle without disengaging the clutch revs will go to around 1250 rpm and if I'm on a flat or slight down slope revs will just start to jump up and down between 1250 and 1500, this is with my foot completely off throttle.  It only stops when I push the clutch idle will then sit around 1250 until the car really slows down and then it goes back down to normal 780 rpm.  Interesting thing is that it looks like throttle is actually opening and closing causing the jerkyiness?  I have been looking at the logs for ages and don't seem to be able to figure out what is causing this?  Intervention?   Idle PID?

Note the car is an RS6 converted to manual, twin plate clutch, GTX30 turbos

I originally thought I had a hardware issue, have now ruled that out after testing alternate Throttle body, Accel pedal, putting the N249 and stock DVs back in place, etc

Attached is a log with as many variables that I thought might be able to help.

Checked clutch switch is working?

Yeah seems to be working variable reads 0 during jerky behaviour then 1 when I push the clutch in and the oscillations stop but revs still hang until car gets really slow.

Just checked and the clutch switch it definitely working  B_kuppl is switching when clutch is pushed and is reading 0 doing WOT and during the time the revs are hanging and it's jerky

Ok so I have the hanging revs stopped.  Thanks PRJ!

But as the revs drop in the same place it's still very jerky.  From the logs there seems to be a lot of things oscillating, but I'm not sure which one is causing which....

 wdkba is oscillating even though wped is 0
 rl and rlsol are oscillating around each other? Could this be causing the ECU to open/close throttle?  Or is this just the result of the throttle oscillating?

Attached is new log

i don't think this is the right log.

oops that was completely wrong, sorry.  Here is the correct log, accelerating, foot off accel pedal and then letting car coast back to idle speed.

What was the solution?

He had it posted, must have edited it?

I think I can glean from the title of his log what he did, but I had a similar issue on a previous setup, so I was hoping for a resolution/confirmation to the original post.

Something go lowered from 400 to 100 IIRC.

Hey guys sorry a little quick to say it was fixed, removed from post as it was not fixed and did not want to mislead

Tried again yesterday and still a problem on slight downhill (light load)???  Trying to figure out exactly which variables to log but most of them don't exist via ME7Logger and I don't know ASM c166 to figure out RAM locations.

Looks like mifa is causing milsol to oscillate is causing rlsol to oscillate, but really hard for me to log enough of MDFAW to figure out where is it coming from.

Thought it might be either intervention, torque reserves or idle PID, but as this is auto ecu converted to manual I'm sure there are a lot of differences due to Torque converter etc.  I really wish there was another 4.2 40v ECU that I could use that was originally manual with 2 MAFs.

If someone would be so kind as to assist with the XDF location etc for KFMIZUOF for my RS6 bin I could eliminate 1 of the possibilities.....

Oh if someone is kind enough to assist my bin is:

http://nefariousmotorsports.com/forum/index.php?topic=2645.0title= (http://nefariousmotorsports.com/forum/index.php?topic=2645.0title=)

Most likely slow path idle pid still not quite what it should be.

Is that your stock bin or your actual current bin?

BTW, I'm pretty sure my issue (with similar symptoms) was either torque reserve or idle PID, and not intervention.  Doesn't necessarily say much about your case though, just thought I'd throw that in there.

That's my original bin

My next step was to test reducing the torque reserves and see how that goes.

Wish I had the original KFMRES tables and idle pid from a manual 4.2 40v, but don't have a properly defined xdf or ols for one.

jibberjive does your car have lightened flywheel and/or twin plate clutch?

It's nothing to do with torque reserve, the slow path idle pid is most likely interfering.

One car had a stock dual mass flywheel, and the other had a steel single piece lightweight flywheel.

Is all modules recoded?

Sounds to me as they are not, I know from other TIP/manual swaps the idle was raised to like 1200-1300rpm until vehicle speed was almost zero, then it went down to normal.

Yes, they are. Everything is on manual and ABS/ESP operates normally as well.
TCU is disconnected.

Without having looked at your logs it's no surprise that idle control is acting weird if you change from an automatic to a manual transmission. You will most probably have to make big changes to the hole function to get a good result because the system behaves quite different now without the torque converter. Maybe start to compare the parameters to the ones of a S6 manual.

Totally agree, but funnily enough this is the only thing that is really a pain, the rest is pretty good and that's with 850cc injectors and A8 MAFs.  I wish I had a defined XDF or ols for a 4.2 40v manual (A6/S6/A8/S8) but unfortunately I have not been able to get my hands on one.  I guess the only other option I have is to learn to use IDA pro and find maps myself...

I was considering comparing 2.7TT files between manual and auto (say M and L) and see which IDLE PID/Torque Reserve etc tables where different, but have not had a chance yet.

Bische, as PRJ says everything is correctly coded, no errors in the ECU or the ABS.  Only code I have is the cluster throwing a single code about not being able to talk to the TCU.  She idles great the only issue is this one while coasting/slowing down while in gear, as Silentbob says though I think it is probably something in relation to the TC auto settings.

What, exactly, do you need to find in your bin?

If the cluster tries to speak with the TCU it is not coded correctly.

I think you are wasting time with idle PID and reserve, the ECU thinks it is still an Tiptronic and sets a raised idle as you coast down in speed.

The jerking is likely related to this as the raised idle PID is calibrated to operate against a slipping TC and not a locked up clutch. I think you should look in the Idle speed algorithm and see if you can find wheres a Tiptronic controls a raised idle.

I would try set CWCFS = 0, CWCFS is at 10189.

For the final time, it does not. I know what I am doing with these. The idle PID is completely recalibrated on this car. The ECU does NOT think it's a Tip, clutch switch works, cruise control works, clutch pedal disengages cruise, kfmresk works etc.
The problem is 90% solved already as well, most of it had nothing to do with idle pid. It's actually perfectly visible from the logs what the problem was.

S_fs is completely irrelevant, it does absolutely nothing when b_autget is not set.
Making this thread was a mistake in the first place.

Hey guys,

Just wanted to say thanks all you have all got me to the point of getting this sorted, the problem with the oscillations was actually in FUEDK/FUEREG.

ILMRMX
ILMRMN
CWFUEDK
CWFUEREG

I just used RS4 values as I don't have a def file for a V8 manual... If anyone gets hold of an XDF for S6/S8 manual it would be great to see if there are any major differences between the v6/v8.

Calm down, I did not intend to say you or anyone else is wrong, thats why I wrote "I think". I also did not know you were in the process of diagnosing the condition.

I spoke from my own experience, not from a RS6 in particular but 1.8t swaps. And according to my understanding, B_mxgecod is set over CAN from the cluster. And since Milka wrote he still has a cluster code looking for the TCU, I did a 1 + 1 = 2.5 and wrote what I did.  ;)

Thanks for documenting the solution, this thread got me really interested about the RS6 V8  :)

This was only 10% of solution, 90% is something else :)

I think the jerkiness is only due to CWFUEREG.