Discussion about issues encountered whilst converting to M3 Flash and GWS

[Bimaxtremeretro]

I think I may have found it but it could depend on who's version of INPA you're using. This is a pic from my INPA that I downloaded from Bimmergeeks. Note shift + F6View attachment 10368

Yep thats exactly it.

Anything to do with throttle changes, that needs to be deleted, cruise etc.

Again guys, use the bimmergeeks standard tools and import ur own daten 60.1 - do not use the mikes package installer, it's rubbish and the way it leaves registry entries is a joke. It will conflice with other software like gini or anything else your using ediabas for.

 

[Bimaxtremeretro]

Guys,

Is it possible to get 1 summary final summary of what the 3 or 4 of you have done exactly and the symptoms we have at the moment.

Stoke is saying dd seems fine right, dyz on the other hand wot shifts are fine. So between the 3 or 4 of u with everything fitted, we need to put together a summary of who has what and exactly symptoms. Everything on the car especially cruise MUST be functional without errors then we can look at the throttle and work it oit.

The reason i say cruise must be functional is because if it's not, it's a telltail sigh the dme egs/dkg and dsc are not happy. Then we can make corrections. There is no point me throwing out 10 ideas which might or might not work when I don't have an easy way to see what everyone has done.

For example: the guys with m3 dsc, I'm almost 100% certain will need to get rid of $544 or $541 and change to $540. Making cruise with braking or active cruise work on m3 dsc is beyond this. Let's just get everything else working and shifting as it should error free and go from that.

We need to report all the errors and work through each one at a time.

To have the kombi properly functional it really needs to be coded to your car not m3. For testing sake m3 is default coding is fine I guess. Just remember what is see in the trace file is barely half the story.

We really need this master summary to compare everything.

 

[Bimaxtremeretro]

Guys, there is also so much wrong misinformation out there re dme and tcu relationship shifting torque etc, it's not funny.

We need to take a step back and think logically.

For reverse engineering purposes, if we compare any bmw dme from the last 20 years between AT SMG DCT vs the MT variation we quickly see they are almost identical with 1 key difference.

That Is, on the non MT versions, what happens to throttle once the AT is reporting shift.

real world, to compare, what happens in a manual if we don't release the throttle while engaging the clutch to shift, at very low part throttle scenario ? And what happens when we do the same thing at full throttle hugh rpm and shift like a bastard.... same shit...

Those who have tried and have enough experience in MT , is this in any way similar or can be described to what your feeling ?

Derek, since you have brass balls, can you try something for me, I've never done it but Would be cool to know.

Let's stick a MT flash on your msd81 and see what happens. If it doesn't start (the msd80 100% does) it will just be the clutch switch but since the CAS still thinks it's a DKG it should should still send the start signal to the dme.

Then drive the car and see if that behaviour is the same.

After that, fuck it, lets grab all the throttle tables from the m3 dme and put them on the 81 and see what happens.

We just need some valid consistent data if you guys wanna get to the bottom of this.

 

[dyezak]

I’m flying back home today and have this weekend to work in the garage. I’m pretty sure I can finish this. I’ll get everyone some updates soon.

 

[Stokes]

Guys,

Is it possible to get 1 summary final summary of what the 3 or 4 of you have done exactly and the symptoms we have at the moment.

Stoke is saying dd seems fine right, dyz on the other hand wot shifts are fine. So between the 3 or 4 of u with everything fitted, we need to put together a summary of who has what and exactly symptoms. Everything on the car especially cruise MUST be functional without errors then we can look at the throttle and work it oit.

We need to report all the errors and work through each one at a time.

To have the kombi properly functional it really needs to be coded to your car not m3. For testing sake m3 is default coding is fine I guess. Just remember what is see in the trace file is barely half the story.

We really need this master summary to compare everything.

I don't want to add to the confusion... I've had limited testing... more time can be spent this weekend on it.

In general, I am experiencing the same (good WOT shifts - extremely fast... too fast I think; sluggish or slower feeling shifts non-WOT). Overall, the shifts are different from before. Might be a good idea to drive a M3 back to back with the non-M3 (335is/135i) to see what the difference is.

For testing purposes, the shifting doesn't seem serious, but also I am not sure long term it's good for the clutch.

I can default code/configure the KOMBI back with the VO set for E82 and see what I get and use that as a basis for further experimenting.

 

[RSL]

Please post logs of the shifts with torque output, torque limit, pedal, throttle, etc.

Guys, there is also so much wrong misinformation out there re dme and tcu relationship shifting torque etc, it's not funny.

Please elaborate on what misinformation is going around, it sounds like you're disagreeing that torque output has any bearing on shifts or how modules respond for them.

We don't know how torque request (pedal) or output is tuned, where limiters/offsets/factors are adjusted, who has what interrupts/paths enabled or disabled, how throttle tables have been altered, etc. It would remove a lot of other variables if everyone were running the same DME bin, but so far, only 1 person has posted a log up to this point with recurring torque pulls while the RPM wandered up and down post-shift.

M3 throttle tables might be a start, but corresponding load to torque and load target tables would be the more complete picture.

 

[aus335iguy]

I think Bill who is BQ tuned is our best candidate. BQ has the most experience with DCTs and my cars shifts felt much snappier with one of his tunes. Have you had any time to do anything on this @LMB335is ?

 

[TireDeathRacing]

Ton of logs from jb4 just pulled them all off the indash Android, I haven't looked at any of them.

 

[Stokes]

I reset variants and then took a quick drive. Took a couple logs. Here are two... part throttle and mostly WOT. The part throttle had a funny feathering happening between 2nd to 3rd when still mostly cold. Didn't happen again after that. The torque act clutch reading is higher than I recall seeing before. With MHD, I can select the Manual Trans option and quick flash to try.

https://datazap.me/u/cstokes54/two-samples-mdct?log=1&data=2-3-12

 

[RSL]

The warm one definitely looks better, but not sure if it's the temp or the WOT. Have a log of the bad/good part-throttles after the variant reset for comparison? I *think* with the DCT flash option, they were just disabling torque interrupt by ignition or a path, not sure if there is anything critically different on their MT option, but certainly easy to try.

Thanks for posting TireDeathRacing. I've only looked at a couple quickly, but have never been a big fan of JB4+DCT explicitly for shifts. I've tried it several times and even with well-tweaked BEFs, it's just noticeably (to me) not the same as flash only. No such thing as too much data though. Any idea what the conversion factor is for torque on the new dedicated param/firmware?

Might be best to include what driving mode(s) are being logged too when they are posted. If we see a change in logged parameters between drive modes and see that some modes work better than others, would give some indication what potential changes may help and where. One mode might be better at part-throttle and another at WOT, etc.

This is just 1 potential facet in the scheme of things, but could show results with a few quick changes. Is anyone running (or able/willing to run) a stock or unlocked map that can be altered and tested?

+1 for BQ on DCT, wouldn't hurt at all to have him in on this.

 

[Stokes]

Ok, take this with a grain of salt until I get more logs. I flashed MHD with the manual transmission option this morning and the part throttle shifting was a lot more like I expected. I had a short drive, so not enough time to catch a good set of logs. Hoping someone else with more drive time and mhd can try the same.

https://datazap.me/u/cstokes54/mdct-mhd-set-mt-dtc

DTC was on for the WOT log.. Tires broke loose and TC intervened... Hence not the best logs.

 

[LMB335is]

I think Bill who is BQ tuned is our best candidate. BQ has the most experience with DCTs and my cars shifts felt much snappier with one of his tunes. Have you had any time to do anything on this @LMB335is ?

Pinion bearings finally arrive today so I can drop off my diff Monday to get the ring gear changed. Once that's done I'll have everything in place. It will take my shop a day to swap out the drivetrain then I can get busy on rest. Heading to Barbados for a few days next week and the EU in June so I'm hoping May will be the month.

 

[dyezak]

Ok, take this with a grain of salt until I get more logs. I flashed MHD with the manual transmission option this morning and the part throttle shifting was a lot more like I expected. I had a short drive, so not enough time to catch a good set of logs. Hoping someone else with more drive time and mhd can try the same.

https://datazap.me/u/cstokes54/mdct-mhd-set-mt-dtc

DTC was on for the WOT log.. Tires broke loose and TC intervened... Hence not the best logs.

VERY interesting!

 

[dyezak]

Ok, take this with a grain of salt until I get more logs. I flashed MHD with the manual transmission option this morning and the part throttle shifting was a lot more like I expected. I had a short drive, so not enough time to catch a good set of logs. Hoping someone else with more drive time and mhd can try the same.

https://datazap.me/u/cstokes54/mdct-mhd-set-mt-dtc

DTC was on for the WOT log.. Tires broke loose and TC intervened... Hence not the best logs.

Well, I just thought through this. @Bimaxtremeretro has been saying all along he used the 1M flash for the conversion car he did. ALL 1M's came with an MT so he was using a MT flash. All of us are factory DCT cars and have our DCT flash.

I wonder if @V8bait or @jyamona would be able to explain to us what difference the MHD "MT" option has and why that might be impacting our situation.

 

[Jake@MHD]

The only thing the MHD option does is set the Load to Torque tables and a few other tables a little differently (higher values because there is no TCU to bitch about too high torque). It does not set any transmission flags in the DME. Those cannot be set in the flash, and actually are either coded to the car or auto-learned by certain CAN packets.

Also, saying all 1M bins are MT flashes is false. A 1M bin (IKM0S file) has the tables and logic to work on any trans, same as the 335is (INA0S file) is the same thing used on both. When he used the 1M flash, at power latch phase the trans is recognized due to packets from the DCT TCU. I assume the 1M bin works better because it has logic support for the M-button. I'm sure BMW tested a DCT in a 1M at some point too..

 

[Stokes]

Hold off deciding anything per my experience until others can verify.

I hope to get more and varied drive time today/tomorrow.

 

[Stokes]

I just had a few minutes to look at my own logs for comparison of before (non-M DCT) and after (M DCT) programming. Same MHD Stage 2+ (v1.0 for N55) and fueling. Torque Act Request is considerably higher now. Before WOT was peaking @ 514. Now it peaks @ 598. Part throttle now is showing 540NM at the beginning of the range and dropping as expected.

Not sure if it is related.

 

[RSL]

I haven't had time to look, but good deal on the test. Torque output will change significantly with AFR and less significantly on other things, but if there are noticeable shift differences with it at all, means it can be tuned around to some extent.

600nm isn't high at all on a tuned N54 DCT, but there may be things on the M setup that expect less since they're not exactly torque monsters out of the box and run higher line pressures.

 

[aus335iguy]

If the DME flash fixes it perhaps we don’t need all the other modules ?

 

[aus335iguy]

Great work @Strokes ! Can anyone verify ?