Yeah, I think that the power mode does actually affect shift behaviour. At least according to what’s known about its logic
M-Drive and MDM in non M cars
- Thread starter aus335iguy
- Start date
Maybe I'm expecting too much, but I had M3 DCT and PT shifts aren't remotely the same between power on/off regardless of throttle, just like they aren't on non-M with sport on/off. As far as MSD81 goes, sport doesn't just alter throttle maps. I assume power doesn't either on M3. I'll need to dig and see if anything except a physical sport button changes those other things in MSD81. Maybe just lack of sport button does it and it's all in DME. There's more than one mode in the trans, aside from logic level, can't possibly think it's just not in comfort 24/7, except maybe be level 5-6 that has too different of shift maps to tell.
The biggest problem is, can't really see any damn state of the M3 trans in anything on DME side and nothing communicated manually seems to effect it so far. Something somewhere must tell what mode it's in (comfort/sport).
If it is related with sending sport signal to dme, i have fixed similar behavior when switched “no sport button” option in TCU for kombi motorsport big letters coding. Then tcu used m3 routines and sends to dme invalid sport status, because m3 dme wants only Getriebeprogramm und Drivelogic. But Sport button did not light up also, maybe problem was in other place. So i patched program, to use m3 routine but send sport status. Now i can feel changes in engine.
@RSL, if you are using non m3 dme(even 1m), gts flash and m3 gws - DME didnt get sport signal from gearbox. dunno why it sends him via ANZEIGE_GETRIEBE 0x1D2 CTR_DISP_SPMOD_GRB. Maybe it is only for KOMBI display status "S". if no other thoughts, lets check.
@RSL, if you are using non m3 dme(even 1m), gts flash and m3 gws - DME didnt get sport signal from gearbox. dunno why it sends him via ANZEIGE_GETRIEBE 0x1D2 CTR_DISP_SPMOD_GRB. Maybe it is only for KOMBI display status "S". if no other thoughts, lets check.
Right, non-M, sport button tells the trans to enter sport mode, trans tells the Kombi to light sport and if CTR_DISP = 1h, LV_ETCU_SPT_SWI = 1, which sets everything.
On M3, there is no "sport" in Kombi and LV_ETCU_SPT_SWI is always 0 without a hack. M wheel button definitely triggers low torque request in 1M and B_sport flag shows active, but I don't know that's what toggles the sport factor torques. I'll have to look tonight.
What in the name of all that's Holy tells the M3 trans itself to enter sport mode without a sport button, just 0x399 or CIC CAN?
I know you said M3 supposedly enters sport on 5-6 automatically (and I will try to CAN log), but it's manually available on all levels in an M3, so something must turn it on.
Just for info, 0x198 doesn't do anything on mine. It's there, just never changes. Everything works though (shift map changes, DL stairs in Kombi, D/M, +/-, etc). Even if MSD81 doesn't use it, I'd expect M3 GWS/EGS would populate data.
I looked back as far as my logs had 0x198, I don't think it was even present before M3 GWS or SZL. The B0 counter runs, but that's it (high bits always 0s). The rest is FF C0 3F CF always/in every log. The net mgmt is active (0x42).
the 0x399 message is at least partly responsible for triggering the transmission to go to sports mode for m3 tcu program per @Olza 's discovery
second point to this, z4 tcu program is definitely listening for something else to switch to sports mode
exactly what i have been saying all along...
looking at this logically, it would be better to get at least get one variant to work first before mix and matching stuff
z4 would be easier in this respect as we now know the 0x399 message generated by msd81 is not exactly "complete" if we compare it with a true m3 message since its missing a bunch of other parameters (eg. shift lights, m_drv_act bit, etc)
the closest thing known to be working is the 1M version of MDM/M-Drive but that doesnt help since its 6MT not DCT.....
I agree that the focus needs to be on 1 setup to stand any chance of getting things working. Changing hardware and mixing and matching software versions just means any progress gets lost in the confusion.
From what I've seen the 1M software doesn't have anything additional to INA0S, just different CW settings in the calibration. Both need to have the M_ACT bit enabled.
I've changed the software to broadcast that M_ACT bit, RSL currently has a test program for that. It could be modified to retain the KOMBI light but I haven't looked at shift lights as they are a nice to have once everything else works.
So, The ideal test scenario would be what ? z4 INAOS DME with an M3 GWS and TCU and an M3 KOMBI and M3 DSC/SZL ?
Ideal test scenario is stock 335i everything to begin with.
Start by working out the DME as this seems to be the home of everything. Once you get to the point where the CAN outputs of the MSD81 match the M3 in terms of MDRIVE then start looking into other modules. It's only the TCU and DSC that are possible with the 335i as I it doesn't have VDC or variable steering weight.
Ideally people can get to a point where they can choose MDRIVE settings for DSC and Drivelogic in the CIC and have them activate when the GWS sport button is pressed. This is my take on it anyway, I'm not sure what the aim is for others.
Ok so we need to be able to see what's going into and out of an M3 DME when the settings are saved and when they are activated. I believe that @amg6975 has those logs from his car so were covered. We also have a number of users here that can flash the dme and log.
@amg6975 would you mind sharing the CAN logs you have ?
Once we have those we need a test and log plan. What exactly do we think needs to be changed in the N54 ROMs to make this work?
Is it calibration data(switches in the ROM like what @RSL found in IKMOS ?) or are we talking program level changes ?
@amg6975 would you mind sharing the CAN logs you have ?
Once we have those we need a test and log plan. What exactly do we think needs to be changed in the N54 ROMs to make this work?
Is it calibration data(switches in the ROM like what @RSL found in IKMOS ?) or are we talking program level changes ?
We know M3 uses same IDs as others, just with different data and maybe not for the same things (i.e. 0x315). The other part of the problem is communication. Nothing "outside" seems to get in to trans.
IKM0S will change M settings to definitions, but doesn't change them in the trans (curious if a stock bin would). The same happens when M3 0x399 is broadcast manually...CIC updates, but still no response from trans.
At this point though, looks like it's going to definitely take CAN and I think we have the messages figured out, just some glitch in the delivery still.
If anyone thinks a DME generated 0x399 with the low bits on may work when a manual broadcast doesn't, I'll flash the custom change @carabuser did, but there seemed to be some question if the change made was correct. Any more comment on that?
This weekend, I was planning to flash/code M3 DSC and repeat everything to see if anything in there has a different link in the chain.
Definitely curious how Z4 handles lack of sport button to set trans sport mode.
B_sport sets torque factors and may set rng l down the line somewhere, but trans sport signal to Kombi definitely requests low. I haven't followed them down the line to see if they're mutually exclusive or if they both set everything at some point, but sport button changes multiples things and there are separate paths.
IKM0S will change M settings to definitions, but doesn't change them in the trans (curious if a stock bin would). The same happens when M3 0x399 is broadcast manually...CIC updates, but still no response from trans.
At this point though, looks like it's going to definitely take CAN and I think we have the messages figured out, just some glitch in the delivery still.
If anyone thinks a DME generated 0x399 with the low bits on may work when a manual broadcast doesn't, I'll flash the custom change @carabuser did, but there seemed to be some question if the change made was correct. Any more comment on that?
This weekend, I was planning to flash/code M3 DSC and repeat everything to see if anything in there has a different link in the chain.
Definitely curious how Z4 handles lack of sport button to set trans sport mode.
B_sport sets torque factors and may set rng l down the line somewhere, but trans sport signal to Kombi definitely requests low. I haven't followed them down the line to see if they're mutually exclusive or if they both set everything at some point, but sport button changes multiples things and there are separate paths.
The M3 DSC has a big part to play as its the one initiating the whole shebang as the SZL and DSC are "Married"
in x35 0x315 also under monitoring. routines checks both conditions.
- for Program level detection: always P1 for 335, P1-2-3 for Z4 with corresp. button level.
- for internal DCT "sportmode": sport button status (yes, 0x198) for for 335 or sport/sport+ for Z4
- for launch control checks, Z4 only, required vehicle status 3 or 11 (thats why LC didnt work when you reflash Z4is data to 335).
- for Program level detection: always P1 for 335, P1-2-3 for Z4 with corresp. button level.
- for internal DCT "sportmode": sport button status (yes, 0x198) for for 335 or sport/sport+ for Z4
- for launch control checks, Z4 only, required vehicle status 3 or 11 (thats why LC didnt work when you reflash Z4is data to 335).
it seems shiftlight not related to DCT ( its only DME thing. no? CAN SHLI_ST 0x206
1M is fine for SZL, thinking more specifically about comms to trans.
I just said last night that 0x198 never changes on mine with M3 GWS. M should send shifter positions and DL button presses, but it doesn't change ever, and I looked back months of CAN logs. Always the same except counter bits.
Olza, could you trace through the TCU to find where the drivelogic is set?
I know we can set it using the fixed values in the TCU calibration but I think it should use the signal from ST_MDRV as shown in this picture you posted here: https://www.spoolstreet.com/threads/m-drive-and-mdm-in-non-m-cars.7155/post-107160
That would explain why @RSL can't get his drivelogic values passed from the DME to the trans.
that mdrv values parsed only if that correct MDRV_ST comes.
335 DME looks for sport signal. m3 gws cant give it. instead of it it looks for drivelogic switch button (this is switchable in config and potentially can be changet back to sport button analysis).
335 DME looks for sport signal. m3 gws cant give it. instead of it it looks for drivelogic switch button (this is switchable in config and potentially can be changet back to sport button analysis).
Last edited:
