NefMoto

I have a car I am working on with a sinter clutch and a flywheel, that is 2x lighter than stock.
This is a ME7.5...

What I have done so far, is increased KFMRES and KFMRESK. This helped in that the clutch can be let off carefully without the car stalling.

However, what I have not fixed yet, is rev oscillation when releasing the clutch and giving the car some throttle.
Because the clutch barely slips at all, it will catch and then release, the revs will overshoot then undershoot, and it goes into a resonance loop. Eventually the engine stalls unless you push the clutch in. I lowered the sensitivity of KFPED and that helped. Because the engine is almost rigidly bolted to the chassis, I increased the idle to 1000 rpm, as any lower there is very strong vibration through the chassis.

I wonder what I should be attacking next. The idle itself is good now, but taking off from a start and especially reverse gear could be significantly improved.
This is not technically idle, it is very low load and engaging-disengaging clutch and revs oscillating.
One thing I should probably look at is ARMD (I have just disabled this on WOT), but any pointers would be appreciated.

Is bit b_llrein set?

Good point. The car needs some work on the fuel system, this work will be finished tomorrow. After that I will log and check this.
I guess if it is set I need to work on the PID and if it is not set, then something else.

I have a Spec stage 3+ sintered clutch disc and pressure plate, along with a Spec 7.5lb flywheel installed on my '02 gti 1.8t The tune I have which is way underpowered for my bolt ons is from Nuespeed, stage 1.
The ONLY time I ever had clutch engagement problems was during the initial break in period which lasted about 1500 city miles. after that, engagment has been smooth as butter with very close to stock like drivability.
I have never messed with the tune, but again the tune is only a stage 1 Nuespeed. I'll send it to you for observation if you' like but Im pretty sure anything you are trying to change up, prolly is still stock on this tune. As far as I can tell with the def files I have Nuespeed only changed like 8 maps.
I've never had issue with stalling unless I let the clutch out too fast while at a low rpm. Good luck!

Oh ya! I also wanted to say during break in the clutch was very "on/off", no slippage and I also have 034 motorsport 85% durometer tranny and engine pendulum motor mounts and the dogbone mount is fully polyurathane 85% durometer even the front bushing in the dogbone, so I have as close to solid mounting as possible without actually having solid mounts.

This car is far beyond Stage 3, never mind Stage 1. The flywheel is I believe lighter than yours, and the clutch is a LOT more heavy duty than yours. It is broken in, and it is almost completely on/off.
I don't have problems with clutch engagement, I probably have problems with idle PID, but I will confirm after I log the flag silentbob mentioned.

Thanks for offering your file, but I'll pass.

WoW a Spec stage 3+ clutch and pressure plate is good for 525 ft/lbs torque. I'm still writing the tune I will replace the Nuespeed tune with...slowly learning between homework assignments for college.
Does your clutch disc have a solid hub or is it sprung? If its solid, I understand engagement will be nothing but on/off. Mine is double sprung. Please post your fingings when you get the chance after you've figured it out. would be greatly appreciated! Much thanks!

It is a multi plate clutch and it is good for over 750 ft lb. As on off as it gets. I consider myself quite used to heavy clutches, but I still manage to stall this thing almost all the time. There's just absolutely no inertia carried by the flywheel.
I don't know much more details about the clutch, since I did not build this car.

And sure, I will let you know the details.

Okay, I had a chance to log this now.
This is 100% some sort of PID controller going nuts, because the graph is perfect in shape.

I just pushed the throttle to a certain RPM and then slowly released the clutch.
I could not log b_llrein, because I don't have that defined in my .ecu file, and I would need to search the disassembly for it first.

What I could log was b_ll. The problem happens the moment it gets turned off.
Please see attached image.

In this image you can see that requested load does the same oscillation and throttle plate angle follows, even though I am holding the throttle in exactly the same position.

I also attached the log.

Both mifa_w and mrfa_w follow req load as well.

I must log wped_w, I think the problem might be coming from mrfgr_w.
Will verify tomorrow.

How does ignition angle look? Have you logged zwout?

Sure, download the log file.

Has exactly the same swings as everything else.

I solved the problem.

I significantly increased PVDNV and everything is OK now.

Can you summarize what PVDNV does?

I see that it is the P constant for the idle control PID, but how does that affect letting the clutch out?

I thought there was a different anti-stall system unrelated to idle control in ME7?

PVDNV is for idle control when vehicle is moving as I understand it.

I did not delve into it too much - I just saw the map, and thought I'd try it before anything else. Upped it some and the problem was gone...

The oscillations in my log looked like a PID controller with too low P and it seems I was right.

whats the address for that map?

That's a good question. Unfortunately I am not a long distance mind reader, and don't have any clue which binary you are using ;)

oh right!..forgot about that.

[Identification]
HWNumber          = {0261207446}
SWNumber          = {1037362937}
PartNumber        = {06A906032HS }
SWVersion         = {0007}
EngineId          = {1.8L R4/5VT     }

Thanks prj!

I'm going to need to do this in my S4.  I have a very light twin plate clutch and I've had a few stalls while coming to a stop and declutched.

Increase KFMRES, KFMRESK as well as the map I mentioned...

PVDNV?  That's my next step as it seems logical.  I've already bumped up the torque reserve as you mentioned as well.

Hey prj, do you remember where the address is for PVDNV?

On what SW?

Identification]
HWNumber          = {0261207446}
SWNumber          = {1037362937}
PartNumber        = {06A906032HS }
SWVersion         = {0007}
EngineId          = {1.8L R4/5VT     }

Is 0x11221 and factor -0.000000+X*0.001526 correct for the M-box for PVDNV?  I'm interested in playing with this map as well to correct a stumble at idle when the car is rolling at 1-2 mph and the clutch isn't pressed in.

Prj, would it be likely that these things would involved when there are problems with stalling in cold weather with thick gearbox oil in addition to idle torque reserve?

You mean with the clutch in ?
I just changed P because I had P oscillation.

Does not mean this will solve all the problems in the world.
If you have stalling when cold try tweaking KFMRESTA a little bit and see if it makes it better.

KFMRESKH is a nice map for it, but unfortunately only used when cat heating is on.
However, it is possible to hack the ECU to run always from KFMRESKH.

ya forgoted me agains..  :(

I found this in an audu tt 225 ols with a pretty extensive definition list:
Characteristic P-gain at start (10x1)

() / -

U / min
-100.000 -50.000 -10.000 10.000 50.000 100.000 150.000 200.000 250.000 300.000
- 0 0.0443 0.0351 0.0351 0.0366 0.0458 0.0595 0.0595 0.0595 0.0595 0.0595

but it's PVDNST

up...

PRJ do you mind posting what you had in the PDVNV map stock?  Trying to match it up (and see if at least one definition file is sane to copy from).

Thanks!

From a quick look in mbox code 0x11221 should be PVDNV  (6x1). I have no idea if this matches what def files say. (or if you are working with 551M)

I know its a little dusty thread, but since I am already working on the problem for some time now that the engine stalls when stopping and declutching I have some oscillations when open throttle a little and engage the clutch. I am going to check tomorrow if tweaking PVDNV is going to help out my problem.

I have a Spec stage3 sinterclutch with 4,5kg flywheel which makes it terrible to drive around when dieing at stops all the time.
(Already tried FKKVS to compensate a little fuel short on low RPM which helped already by 60% but dint solve the problem, also somebody else said solving this sort of issue with adjusting ignition time and tweaking acceleration enrichment, only ignition did something)

You need logs.

1. Log
2. Deduce what the problem is
3. Modify parameters to solve the problem

Changing things randomly is never going to solve such issues.

hello
in this binary anyone found PVDNV ???
or only these ?

It is not necessarily present in every binary. Can be called something else.

Basically IF you have the exact same issue you need to adjust the RPM pid controller when the car is not idle.
For example this controller controls the engine RPM when the car is moving in gear with no throttle applied.

Don't just search for value and add + to it without logging and finding out what the problem is.
If your fueling is bad at some lower pulsewidths, it can also cause such yoyo effects...

Another one that will cause random stalling with a light flywheel is when your fuel resume is set to stock values (too low).

LOG LOG LOG LOG :)

I tweaked the PVDN a little rough so sometimes its going to oscilate at idle, when coming from overrun.
The biggest part of the problem is solved, if I declutch now when i just release the gas pedal it picks it right up at 1000 rpm. The only problem still remaining is when declutch at full overrun coming from above 3k rpm and let is slow down between 1250-2000 rpm, when declutched it still stalls. If I declutch above 3k rpm it will pick up as it should.
 
I am sure to bring the PVDN way more back because I bounched it up to much for my thoughts. However I am going to log and post it.

Here is my bin and xdf already:

What RPM is your fuel resume at? I am sure it is set too low.

What do you mean by fuel resume?

Searched for it but you are the only user using this :)
Translated it to my language but its not clear to me.

Check BBSAWE in FR

I believe prj's talking about the map KFNWEGM

Yup.

I tried to be a bit prj'ish and avoided spoonfeeding.  ;)

why? we already have one prj.

I was finally able to log some.

I surely need to decrease PVDN a little because with a little throttle and clutching the rpm start to oscillate, but that's for later.
I cannot find the map KFNWEGM which Prj is telling and also the map BBSAWE. But I see that BBSAWE is for raising fuel cut by tester, so what is that supposed to say?

I already raised idle to 1000. Also experimented with NRLMN and NRLMNLLR which gave the result (in combination with PVDN) that it will hold steady on 1000 rpm when declutch right after closing throttle, but it sucks to drive like this and shouldn't be the way of driving.

Logs:

Altough I cannot get the right resolution in ECUxPlot, must be the filtering?

EDIT:
Got something already.

(http://s27.postimg.org/4u1w2e3zz/Overrun1.jpg) (http://postimg.org/image/4u1w2e3zz/)

You are flooding it. 7ms injection time at 1000 rpm???

I believe this comes from FKKVS, which I adjusted in the first place and resulted into a better response of the engine.

See the pics below

Some additional info about my setup:
Wiseco pistons
Pauter conrods
Dahlbeck intake manifold
ATP exhaust manifold
Garrett GT2871R 0.64 A/R
Cat cams 268 degrees intake cam / 258 degrees ex
3" exhaust front to back w/o cat
Siemens DEKA 630cc injectors
4,5kg flywheel
Spec sinter raceclutch  stage3 i thought (single disc)
Stock MAF

Finally its getting better but I am not still there yet.
I've 0-ed between 60 and 100 degrees C in KFBAKL which resulted in a likely 3ms inj on 1000 rpm. I also put back stock FKKVS and adjusted on idle.

The engine doesnt not stalls anymore buuuut Its still not how I want to have it, because its still lacking at 500rpm and raises up again to 1000.

Are there any other Fueling maps besides KFBAKL and FKKVS which I can adjust for low rpm settings? Or any other suggestions?

I just read in functionsrahmen KFBAKL is for acceleration and KFVAKL is for deceleration, so I will adjust KFVAKL also and log what it does.

Cheers!

Engine not really stalling anymore, but I am still with the issue that the rpm drops to low before picking up again.
Somehow its still injecting even when the throttle blade is fully closed.

Is there any way to get this rpm drop fixed?

Someone can help find PVDN in 4B0906018DF ?

Odd seeing prj asking for help :o

HI. Do you resolved your problem? what was the problem? fkkvs , kfvakl or kfmres/k? Thanks!

This was 9 years ago, did you miss that fact?
And 9 years ago I was able to log the ECU, find the problem and solve it. Then I posted the solution for everyone. 9 years later this is still used.

You weren't even on the site in 2012.

Some relevant info for thread:
Over the past 9 years I have learned that idle issues are very often the result of incorrectly tuned injectors.
Another thing that can help with light flywheels is setting SAWE to deactivate a little earlier.
And finally, in my specific case the P term needed work because the oscillation were caused by incorrect P term for flywheel weight.

Nothing wrong with asking a question, please dont take it as an insult.

I wasn't trying to poke fun or anything about it. If anything I admire you posting your problem and working threw it to find a solution publicly. It helps the community and shows proper forum etiquette. Also for what it's worth I have been here sense 2012 too. Forgot my login credentials and an old email I couldn't get into threw iCloud.

Back that time I was not having enough knownledge to known what the exact issue was.
In another thread Prj throw me to issue and that was the idle controller, which was not calibrated for my setup, to poor and caused stalling of the engine.

I managed to fix this.

Off side topic:
We are here to learn, but you need to put effort in it. That is the way to get there.

Ps. Keep the timeline in mind ;)