NefMoto

Hello all !

 i work on golf 6 R20 270hp and after good remap i have a only little gain.

i can feel hesitation du to my DSG Torque limiter.

i know it's possible to tell the DSG computer you are making less torque
I can scale/edit the torque values ( factor ) to appear lower in the med9 ecu.

do you know wath function or factor i need mod for that ?

sorry for my english !

Thank's in advance

If it was a good remap, you would have BIG gains, not only "little gain". you need to work harder and go back to the basics of the ECU.

Change torque limit in DSG. All other solutions is just not correct...

CDLF should have about 350NM from factory and in these cars DSG is limited to 380NM. You can also see torque limitation in measured blocks, should be 122.4

DSG software is 350Nm limiter, not 380.

Reducing MDNORM from 500 to 340 is enough for a stage 1. It's not true that this is an "incorrect" way to do it as a stage 1 won't require higher RPM or torque handling that a TCU flash will give.

Don´t claim something if you didnt check it...
There are few versions with more than 350NM allowed. In attach a DQ250 file from Golf R and other DQ250 file from Passat TDI. I think you will find the offsets yourself. Btw ori shift rpm should be 6700...

I still think it´s not a good idea to cheat torque and let the tcu calculate with 100NM less, maybe think about clutch slip regulation and wear. Change tcu torque limit is 5min job.

Of course it's incorrect, since all torque reported to the gearbox is wrong. 500 to 340 is especially stupid.
Torque is I feedforward in DSG, and you're reducing feedforward by over 30%, causing permanent clutch slip all the damn time.

If you want to cheat torque there is absolutely nothing required to do with mdnorm in the first place.

This thread is idiotic too - "I can feel" bla bla - no you can't feel shit. Log.

+1

...and if you do not have tool for DSG there are services now which offer you DSG tuning in sgo format. So you can flash with normal tools, e.g. VCP, AVDI, VVDI or some cracked VAS tool...

You should check yourself. That 380 is max torque is an upper protection tripwire mmr_MMaxGetriebeschutz_ko.

MmrTemperaMmax_kl caps TQ at 350 up until 138deg C then the TQ limit drops to 15Nm.

Incorrect is thus subjective. There is no clutch slip in a sensible stage 1 with TQ output called at 340. As for saying it has nothing to do with MDNORM, it absolutely does. The ECU will never report more than what MDNORM is capped to.

Attached Passat CFGB is 380NM from factory, can you explain this? ;-) In my opinion this map is active from 138° only, not before. But i dont have FR for it.

First of all, we're talking about Golf R DSG boxes here, you've thrown in a random Passat file. Secondly, mmr_MMaxGetriebeschutz_ko in that Passat file is 390Nm, and the vehicle output is officially 170PS/350Nm. The temp limiter MmrTemperaMmax_kl is still 350 though.

As for your opinion on how active that map works, that's wrong. See attachment. From 0-138deg, it will allow 350Nm. Between 138 and 145, it will interpolate between 350 and 15Nm, dropping TQ all the way to 145.

Only, if your stage1 has NOT more 340Nm indicated.

that´s correct:
ECU will limit all communicated torque to 340Nm.

Problem on this:
MO1 and M06 torque is indicated, so flywheeltroque is mutch less then 340Nm

All torque higher than 340Nm indicated are cutted and all feed forward torques are worng.

As well as all adapted torque-to-pressure values ...

Only mirco-slip-controler will learn, but has mutch trouble, if differend learn value for same communicated torque higher then 340Nm indicated.

So even on stage1 tuning this way is a VERY BAD idea !


to keep it short:
all ideas, NOT to tell actual torque correctly to DCTs is CRAP.

With this hysteresis around accurate torque reporting, you would think that 1Nm over "reported" would cause the clutches to slip. I understand how the DSG generates clamping pressure and how the microslip controller works, but a healthy DSG gearbox produces enough clamping pressure at reported 340Nm to hold much more. At the end of the day, there are thousands of DSG cars with MDNORM reduced running absolutely fine. I would never recommend running this with much more torque, but stock DSG with 340Nm reported is more than capable of holding stage 1 torque. The whole issue around the reported number is semantics at this point.

Depending on actual situation and amound of troque you are running.
On launch and high speed gradients, you will get in trouble, if the actual torque differs less the 10Nm to communicated torqure, even if the inertia of the engine is low.


Yes, if you are in status "fahren" and the micro-slip-control adapt to mirco-slip-speed-limit and press "correct" to actual troque.
So it will work in this special condition. But you will see, micro-slipp-controller needs some time to reduce slipping.
All controled situations will get in trouble, if the communcated torque is incorrect. So off-going clutch during torque-hand-over e.g. or most importanted: launches poweron-downshifts.

Cannot match to this !
Take a look at clutch slip of the shafts and calculate the energie you loose by slip, you will see how mutch stress you will put into the system, when you running at torques higher MDNORM indicated.

Running alltime short-hard shifts and fast launches, you will not sense the shocks, you will get by incorrect troque, because the kind of application shocks by physiks.

But if you want to have a smoose and findly clutch control even on high torque, you need to tell correct torque to DCT.

And this will not happend, if you cut the high torque communicated on you "stage1" tuning.

The point is:
as long as you are running lower the MDNORM indicated, the DCT will work with proper communicted torque and work well.
This will be the cause of most times running a stage1 tuning.

But racing or mutch time running above MDNORM, will have bad impact to livetime of clutches and DCTs

Sorry in fact i was talking from wrong car, CFGB is only 350NM. But CFGC is 380NM. I only want to say that there are also cars with more then 350NM limit, and i still doubt the temperature map is limiting to 350NM. I never touched this map and never had any limitation, so i expect this map is active "from" 138° or there is another single temperature value to make this map active first.

I never tried to cheat engine torque output because it simply makes no sense, these DSG are easy to read/write since many many years. I understand you say it´s working, but you can say nothing about clutch wear over lifetime or stress from oil etc. Sure the controller adapts microslip, but then better bend torque output instead of making hard cut at 340NM...

No! Deffendly not.

Alltime, when clutch is controlled, you will have a lot slip, when you bend torque.
So better (most certainly not good) let the DCT run like it is used to and only bend or cut at very high torque, witch are mostly used in status "fahren" und mirco-slip-contoller will do his best to catch the engine speed.

There is nothing subjective.
If you reduce MDNORM from 500 to 340 on any DSG box, the feed-forward will be insufficent and the clutches will be CONSTANTLY in the microslip adjustment.
They will be slipping much more than the microslip target non-stop all the time on any increasing torque gradient. Adaptation will also do jack shit, because while on increasing gradient the control will be constantly undershooting, on decreasing gradient it will be as well, as they will be grabbing too hard and the adaptation cancels itself out.
To even suggest such a thing is "okay" means having ZERO knowledge of how the DSG regulates clutch pressure.

If you want to cheat torque a little on the top-end (and you will still have slip issues, but at least not in all daily driving), there are much different methods. You should never ever change mdnorm on any DSG box to underscale reported torque, if you care at all for the longevity of the transmission.

Just because you don't see it slipping 1000 rpm, doesn't mean it's not there.
I really suggest educating yourself a little more on how the microslip controller and the feed-forward works before making stupid statements that someone is going to take for gospel and damage their gearbox.

Changing mdnorm was used in the beginning when noone could r/w DSG.

Imho the only right way is to change TQ-limit in DSG-SW.

Regards

It was never neccessary to change MDNORM, even then it was always enough to smoothly underscale the torque at the top end a little.
MDNORM screws everything up, also in normal driving.

I agree with you, but this is what many people done in the beginning as a "workaround".

Bumping this because i stumbled upon it after already going through this issue for a couple months. Eventually i figured it out... But this would have been very helpful as a beginner in DSG tuning.

I increased MDNORM to 450, and modified the IOP axis coordinates, and also modified the MAP related to torque loss. Because the original MDNORM is only 400NM, minus the torque loss, the maximum is only 360NM. The actual torque of my car is about 400NM.

Interesting topic and i also could not wrap my mind around this but have much better understanding know.
Increase MDNORM to 450.