carlhook
Newbie
Karma: +0/-0
 Offline
Posts: 23
|
 |
« on: December 21, 2023, 09:04:27 AM »
|
|
|
Hi all,
I have been chasing my tail with this one for a while now.
2011 VW Scirocco CCZB DSG6.
K04-064
Unitronic Intake (yes, I need to fix my maf scaling, but I know this is not the cause, as it was identical on stock intake)
Redtop coils
BKR8EIX gapped at 0.6mm
S3 stock new downpipe with stock cats.
All else stock.
Very mild (Golf6R Stage 1 type) tune, requesting 1.55BAR at Peak then to 1.1BAR by redline. (due to cats, I want to be all MOT legal on daily)
Main issue - slight misfire when cold WOT at 5000RPM that turns into slugish/non smooth acceleration when hot at high RPM + closing TPS.
Maybe you can spot the problem.
It's not knocking.
It's not overboosting - thou maybe it uses Throttle to regulate boost?
It's not fueling issue - rail pressure is ok, injection duty is ok. MAX 7.7ms
rl_w is never reaching rlsol.
rlmxko is way over rlsol.
mifa follows mibas and never reached mimax.
mimax follows mrfa at 100
basically no timing pull, just 2-3degrees from 1-2cyl.
EGTs are ok, and lambda enrichment is disabled for testing.
Lambda Control is -10% because I MAF is not fully perfectly scaled (Unitronic Intake) - but TPS was closing identical with stock intake, so not the issue.
So no knock, not fuel issue, but still slight stutter/misfire when cold at 5k rpm and closing TPS all the time.
What am I missing?
I did notice that it closes less TPS when cold, which then I guess causes the misfire.
When hot it closes more TPS but no misfire, just you can feel it not "Pull" as good.
Logs and all OLS ORIGINAL AND MODIFIED files attached.
I have VEHICAL Logger, and I know there must be a parameter to log for all related TPS closing problems, but I dont know which ones to log.
|
|
|
|
|
Logged
|
|
|
|
|
prj
|
|
« Reply #1 on: December 21, 2023, 01:01:57 PM »
|
|
|
Tune is a hot mess lol.
1. These engines don't like 0.77 lambda one bit. 0.81-0.82 is where they make power, if you have too much knock then reduce boost...
2. You need to linearize boost pid from scratch, it's written here how to do it, but to get a manageable map these need a patch iirc.
3. Throttle cut - most likely your KFMIRL and KFMIOP are not the mathematical inverse of each other or you're exceeding injection timing window. Look at the FR for torque coordination and log inputs/outputs.
|
|
|
|
|
Logged
|
|
|
|
carlhook
Newbie
Karma: +0/-0
Offline
Posts: 23
|
|
« Reply #2 on: December 21, 2023, 03:34:17 PM »
|
|
|
Thanks prj.
The tune is a mess I know, been chasing my tail with this TPS closing forever.
1. I increased Lambda for now as I am also battling the misfire issue which actually melted electrode of a plug yesterday.
I had it at 0.85 forever, so will do 0.82 now. Thanks for the knowledge.
2. The ECU file is patched to do full Openloop, but thats dumb. I discovered CWAWGTV and KFWGV. If I set CWAWGTV to 4 then it uses KFWGV as pre-control for PID and I can get it to be actually quite ok. Not in the logs you see now, as its requesting more boost than the pre-control was tuned for. But I will look into redoing PID, but after I have found the cause of misfire but more importantly the throttle closing.
Does not make much sense tuning PID if TPS keeps closing for now.
3. They are not, thats true. Its some stage 3 base I used, where it does this. (screenshot attached)
I am slowly working the whole TPS functions backwards.
I will make sure to add the torque coordination inputs/outputs for the log.
What is the injection timing window on these? Is it as simple as monitoring the injection time (ms)? If so, whats the "sweet spot"?
Also, then I assume there is a parameter that I can log for "exceeding injection window"?
|
|
|
|
|
Logged
|
|
|
|
|
cherry
|
|
« Reply #3 on: December 21, 2023, 06:47:42 PM »
|
|
|
Maybe just a mechanical problem with soft exhaust valve springs?
|
|
|
|
|
Logged
|
|
|
|
|
prj
|
|
« Reply #4 on: December 22, 2023, 12:46:49 AM »
|
|
|
What is the injection timing window on these? Is it as simple as monitoring the injection time (ms)? If so, whats the "sweet spot"?
Also, then I assume there is a parameter that I can log for "exceeding injection window"? Start reading. Your entire problem is that you used some random file instead of learning why every change is made. KFMIRL/KFMIOP needs to be mathematical inverse and gearbox needs to be flashed to get rid of tq limit. |
|
|
|
|
Logged
|
|
|
|
carlhook
Newbie
Karma: +0/-0
Offline
Posts: 23
|
|
« Reply #5 on: December 22, 2023, 10:00:30 AM »
|
|
|
Start reading. Your entire problem is that you used some random file instead of learning why every change is made.
KFMIRL/KFMIOP needs to be mathematical inverse and gearbox needs to be flashed to get rid of tq limit.
Not exactly true. I do understand most of changes made, and I see no load or torque intervention in logs. Gearbox reflashed. I will re-do the MIRL/MIOP and report back. Also slowly working my way back via wdksb_w. wdkoufs_w is the furtherst I have found the limit originate from. |
|
|
|
|
Logged
|
|
|
|
|
prj
|
|
« Reply #6 on: December 22, 2023, 11:05:51 AM »
|
|
|
I see no load or torque intervention in logs. Why is your throttle shutting then? Because there isn't load or torque intervention? Of course there is. But yes, you can't see it :p Look at MED9 torque coordination in FR and log it. |
|
|
|
|
Logged
|
|
|
|
carlhook
Newbie
Karma: +0/-0
Offline
Posts: 23
|
|
« Reply #7 on: January 03, 2024, 04:00:27 AM »
|
|
|
A update.
Flashed this file (lot lower load req) to try and keep it safe while I find the limiter.
So I see B_enimbgr (Condition: release for regulation of the torque limitation via injection angle) is set TRUE right before TPS starts to close to limit load or torque.
And then Conv_bLdMaxLimActv_mp (Flag for torque load limitation active) follows it.
Limiters for torque are set to 420Nm but the Com_trqLim is limited to 360ish Nm.
Also I see ACCI_trqNLim_mp wants to be at 3276.8 but is limited by AccPed_trqDes to just 2200ish.
Also B_nozwe (condition no intervention of torque-model via ignition angle) is TRUE.
So basically it's limiting torque via injection angle limitation because of load limit or torque limit? thus its all limiting rlmax.
It's not temp, fuel, air, knock, etc. Its just the lovely torque model.
Something is missing...
|
|
|
|
« Last Edit: January 03, 2024, 04:03:04 AM by carlhook »
|
Logged
|
|
|
|
|
prj
|
|
« Reply #8 on: January 03, 2024, 04:08:24 AM »
|
|
|
So I see B_enimbgr (Condition: release for regulation of the torque limitation via injection angle) is set TRUE right before TPS starts to close to limit load or torque. Yeah you're out of injection window and on top of that DSG is untuned and limiting tq. |
|
|
|
|
Logged
|
|
|
|
carlhook
Newbie
Karma: +0/-0
Offline
Posts: 23
|
|
« Reply #9 on: January 03, 2024, 04:17:27 AM »
|
|
|
Yeah you're out of injection window and on top of that DSG is untuned and limiting tq.
Ok, I will reflash the DSG as I dont recall what I have flashed in there now. But I just can't wrap my head around the injection window problem, is there a single flag/condition that = out of injection window? As how B_enimbgr is translated now, it reads as its using injection timing to limit load, not that it limits load due to injection timing. FR - B_enimbgr AWEA LOK Condition: Enable to control the torque limitation via injection angle Also you can see in the LOG that B_enimbgr is TRUE from 2500RPM to redline...I doubt out of injection window in all rev-range. + in my other stupid files, I managed to push 30% more airflow through it but with similar TPS intervention. A bit puzzled. |
|
|
|
« Last Edit: January 03, 2024, 04:22:50 AM by carlhook »
|
Logged
|
|
|
|
|
prj
|
|
« Reply #10 on: January 03, 2024, 07:55:34 AM »
|
|
|
I didn't look at the logs, yes can be that you're not out of inj window.
Make KFMIRL/KFMIOP mathematical inverse and flash DSG to have higher tq limit for starters e.g. 500nm.
|
|
|
|
|
Logged
|
|
|
|
carlhook
Newbie
Karma: +0/-0
Offline
Posts: 23
|
|
« Reply #11 on: January 04, 2024, 02:04:06 PM »
|
|
|
I have read so much conflicting stuff on MIRL/IOP. Making the whole map new to scale to higher loads screws with idle, even if the map is perfect inverse. It's not perfect inverse stock, so who gives? Then I read about modifying just the last 2 columns and load axis to fit required load, but in other DSG posts you talk about not allowing "bent" graphs and maintaining perfectly linear map. Then I have the tuning file servers doing weird stuff like seen in the attachements. I kept all else same, just reduced MDNORM to stock 400NM as I am not even close to that with my load requests being like 190 tops. (I had it at 500, so as far as I understand it could make TCU ask for intervention in torque. I know that I must fix tcu fully, I have file ready, will flash in the weekend) Then I read about MDMAXNMOT but its referenced once in FR and is not a thing in my DAMOS. Still not sure where do I tell ecu what NM is what torq% Life is not made simpler as there is like a ton of info in this forum, but I find best post about IOP/MIRL in DSG tuning section....  My goal is to create the 6R Stage 1 type of tune on this CCZB MED17.5, to be as smooth as can be, retain all protections and then create a post talking about all changes, and reasons why so the next guy can learn, but without wasting a ton of time on bs. Just found out about HP Tuners, they have my ECU supported, just no idea how good their map definitions and interfaces are. |
|
|
|
|
Logged
|
|
|
|
|
prj
|
|
« Reply #12 on: January 04, 2024, 09:16:52 PM »
|
|
|
You don't understand torque in ECU at all.
1. Do not fuck with MDNORM, set it to stock value.
2. If you don't need more load than is in stock KFMIRL map, set KFMIRL/KFMIOP stock.
The other shit you have is a great example how NOT to tune if you want your DSG clutches to have any life to them.
If you need more torque than stock MDNORM you will have to scale the whole map by the same amount as you scale MDNORM.
And also rescale the idle controller.
Or really, just do whatever the fuck you want lol. Garbage in garbage out.
|
|
|
|
« Last Edit: January 04, 2024, 09:19:08 PM by prj »
|
Logged
|
|
|
|
James1
Jr. Member
Karma: +0/-0
Offline
Posts: 30
|
|
« Reply #13 on: January 05, 2024, 03:59:05 AM »
|
|
|
7.7 ms way too much. This is a DI engine. You are out of fuel.
|
|
|
|
|
Logged
|
|
|
|
carlhook
Newbie
Karma: +0/-0
Offline
Posts: 23
|
|
« Reply #14 on: January 10, 2024, 08:15:39 AM »
|
|
|
OK, update.
So turns out if you use KFLDHBN as boost map, it uses TPS to limit the load.
I have attached 2 files, one is with KFLDHBN as load/boost limit other is KFLDHBN raised away and LDRXN as main limit.
Then also the following changes:
KLSWBHTKR raised from stock 850-600 (at 5500rpm) to 900 for all rev range from 3k rpm.
There are multiple maps, but this I think is the main EGT enrichment map per RPM/EGT. - in logs I still hit it, thus rich lambda gets dumped, but until I actually match my calculated EGT to the real worlds ones using real sensors, I want to stay safe. (full stock downpipe, for MOT)
KFFLAF - LPFP duty as mine is openloop and dumb, but I doubt the TPS fix is due to fuel as between the runs and logs, even at higher boost levels, my injection time is very similar.
KFKHFM - still thinks more air is going in that it is, so it lambda controlls -10% now at WOT, this is due to shit Unitronic Intake design, I will work on fixing this next.
KFLDHBN - all max to raise out of way for now.
I will test with stock KFFLAF but I doubt its the culprit for TPS, I think it was the KFLDHBN.
|
|
|
|
|
Logged
|
|
|
|
|
