6HP Custom Tuning

doublespaces

Administrator
Oct 18, 2016
9,303
4,331
0
AZ
Ride
2009 E93 335i
Now that the XDF is out for the 6HP28 and 6HP21, what are all the possibilities we could see beyond what @RayBan has shown us in the OTS maps? TCU tuning has never been an option for this platform and I think there are a lot of intelligent people who have something to contribute. For example, how involved is it to simply raise the shift points up to say, 7500? How much extra stress does this put on the transmission, how is that calculated?
 
  • Like
Reactions: bradsm87

bradsm87

Corporal
Dec 15, 2016
162
53
0
Unfortunately, my MFactory LSD bolts have started to fall out, causing damage to my diff so no road testing for me, but I'm working on an initial flash and some alternative files to test certain functions. As soon as my diff is fixed, I'll be doing lots of testing.
 

doublespaces

Administrator
Oct 18, 2016
9,303
4,331
0
AZ
Ride
2009 E93 335i
Unfortunately, my MFactory LSD bolts have started to fall out, causing damage to my diff so no road testing for me, but I'm working on an initial flash and some alternative files to test certain functions. As soon as my diff is fixed, I'll be doing lots of testing.

Oh wow that sucks! When do you expect to be back up?

For me, shift point is the biggest interest. I hate seeing the car shift so soon, particularly when my single turbo is installed.
 

bradsm87

Corporal
Dec 15, 2016
162
53
0
I've added the torque value breakpoints to the XDF and will pass it on to xHP.

Next on the to-do list is to separate the upshift and downshift shift maps into separate XDF tables because with consecutive upshift and downshift columns on the same table, it's hard to visualise and if you put it into a graph, it's all zig-zaggy.

Breakpoints.PNG
 
Last edited:

bradsm87

Corporal
Dec 15, 2016
162
53
0
I've converted one of the shift map tables to separate upshift and downshift tables which makes it way, way easier to visualise the table contents rather than having upshift and downshift mixed between each other. It'd be pretty time consuming to do them all so not sure if I'll go further with it.

Before:
Combined up down.PNG



After:
Separate up down.PNG
 

bradsm87

Corporal
Dec 15, 2016
162
53
0
If anyone wants the shiftmap tables to display in vehicle speed rather than output shaft speed, Change the conversion in the tables from "X" to:

X/(38/11)*3.14159*((265*0.35*2)+(18*25.4))/1000000*60

Where 38 is crown wheel teeth, 11 is pinion teeth, 265 is tyre width, 0.35 is 35 profile, 18 is wheel size. Swap in your values.

That's for km/h. Add *0.6214 to the end in you want mph.

If you're unsure about crown and pinion gears, just put the diff ratio in in place of the "(38/11)".

Speed.PNG
 
  • Like
Reactions: Irafi and Jeffman

RayBan

Corporal
Oct 27, 2016
154
155
0
42
www.rbttuning.com
Hi guys,

there is a github repository available already:

https://github.com/jakemoroni/BMW-6HPxx-XDF

It's not operated or monitored by us, but I guess it's a good idea to coordinate community efforts in one place.

The base xdf file we published will work for HW# 7591971 and 7591972. That covers all post 03.2007 335i/335d/135i cars + most 325d and 330d. Other cars, with different hardware numbers will require different xdf's.
 

bradsm87

Corporal
Dec 15, 2016
162
53
0
Hi guys,

there is a github repository available already:

https://github.com/jakemoroni/BMW-6HPxx-XDF

It's not operated or monitored by us, but I guess it's a good idea to coordinate community efforts in one place.

The base xdf file we published will work for HW# 7591971 and 7591972. That covers all post 03.2007 335i/335d/135i cars + most 325d and 330d. Other cars, with different hardware numbers will require different xdf's.

So I found out the torque limiter map is withheld from the XDFs. What other maps that are needed to do a good TCU tune are withheld?
 

bradsm87

Corporal
Dec 15, 2016
162
53
0
What do we know about the tq limiter anyway?

For the Alpina flash, it kicks in and tells the DME to reduce torque at around 630nm in the midrange. As a result, the DME reduces ignition timing. The current workaround is to adjust the DME tables to report a lower amount of torque to the TCU than what it's really running. This means the TCU can't produce the ideal shift because it doesn't know the real torque and the tables aren't mapped up to that much torque.

The table itself should be a torque limit per gear and 6 cells wide. I've already spent around 4 hours manually searching through differences between stock and Alpina files with both known possible torque conversion in both 8bit and 16bit HiLo decimal conversion with zero success.

Apparently the clues lie in the BMW/ZF 8HP OLS file which is floating around on various paid "mappacks" sold by dodgy looking members on dodgy looking forums.
 
  • Like
Reactions: Jeffman

doublespaces

Administrator
Oct 18, 2016
9,303
4,331
0
AZ
Ride
2009 E93 335i
For the Alpina flash, it kicks in and tells the DME to reduce torque at around 630nm in the midrange. As a result, the DME reduces ignition timing. The current workaround is to adjust the DME tables to report a lower amount of torque to the TCU than what it's really running. This means the TCU can't produce the ideal shift because it doesn't know the real torque and the tables aren't mapped up to that much torque.

The table itself should be a torque limit per gear and 6 cells wide. I've already spent around 4 hours manually searching through differences between stock and Alpina files with both known possible torque conversion in both 8bit and 16bit HiLo decimal conversion with zero success.

Apparently the clues lie in the BMW/ZF 8HP OLS file which is floating around on various paid "mappacks" sold by dodgy looking members on dodgy looking forums.

Perhaps a diff will not reveal the torque limiter differences. The actual change may simply be a modifier?
 

bradsm87

Corporal
Dec 15, 2016
162
53
0
Perhaps a diff will not reveal the torque limiter differences. The actual change may simply be a modifier?

The table that was posted by RBT Tuning earlier is a 6 cell table in NM units. I suspect that it might use a conversion that has not been used elsewhere in the XDF yet so the raw number in the ROM is very different to the actual NM figure. Myself and one other person have spent about 5 hours each searching so now we're trying to get the BMW/ZF 8HP OLS file which according to DWR at xHP, holds a lot of clues.
 

doublespaces

Administrator
Oct 18, 2016
9,303
4,331
0
AZ
Ride
2009 E93 335i
The table that was posted by RBT Tuning earlier is a 6 cell table in NM units. I suspect that it might use a conversion that has not been used elsewhere in the XDF yet so the raw number in the ROM is very different to the actual NM figure. Myself and one other person have spent about 5 hours each searching so now we're trying to get the BMW/ZF 8HP OLS file which according to DWR at xHP, holds a lot of clues.

Excuse my ignorance(I have no clue how this works) But why would you want an 8HP OLS file, just because its similar? There isn't a 6HP OLS file around?
 

bradsm87

Corporal
Dec 15, 2016
162
53
0
Excuse my ignorance(I have no clue how this works) But why would you want an 8HP OLS file, just because its similar? There isn't a 6HP OLS file around?

There is not a BMW/ZF 6HP OLS. According to DWR at xHP, The BMW/ZF 8HP is the closest. It may share some conversions etc. Disassembly and logic tracing can reveal the table locations and axises. One guy has put his hand up to do this. I'll manually search through differences to hopefully find a few too.
 

doublespaces

Administrator
Oct 18, 2016
9,303
4,331
0
AZ
Ride
2009 E93 335i
There is not a BMW/ZF 6HP OLS. According to DWR at xHP, The BMW/ZF 8HP is the closest. It may share some conversions etc. Disassembly and logic tracing can reveal the table locations and axises. One guy has put his hand up to do this. I'll manually search through differences to hopefully find a few too.

Nice, tell him to come over here and hash it out.
 

mfish123

Lurker
Mar 9, 2017
15
5
0
Ride
2009 BMW 335i xdrive E90 Sedan
EDIT: I'll leave this post up to inspire / inform others but I think I figured it out. I multiplied multiple cells by 0.9%. I'm seeing that the bottom 110 row has a higher value than the rows above (since they were all lowered by the same percentage). As @bradsm87 has alluded to he thinks the 110 row corresponds with the kickdown switch. So that would make sense that I rev higher when pressing the kickdown switch. Full details below.....

So I've been doing some testing to lower the "redline" (max rpm shiftpoint when WOT). What I'm finding is that if you lower the OSS in the relevant fields in the relevant shiftmap the car will obey as long as you don't press down the kickdown switch. I'm on stock turbos so no sense revving to 7,000 rpms except maybe in 1st gear.

However, if I use the kick down switch in D mode (and more than likely in S mode too but didn't test) to downshift and then pull to the top of the gear it doesn't obey the lower redline. Is there maybe a separate shiftmap table for when you use the kickdown switch? There are a few without a description in the XDF that I didn't touch. Or maybe some hidden toggle that overrides the max OSS when you use the kickdown switch

6th to 3rd downshift with kickdown switch - http://www.datazap.me/u/mfish123/6th-3rd-downshift-0?log=0&data=5-7-23-26-30

2nd to 3rd upshift in D mode with lower shiftpoints - I forgot to totally disable DSC hence the negative timing in 2nd with spinning wheels - http://www.datazap.me/u/mfish123/3rd-4th-lowered-redline

1st to 2nd to part of 3rd - I purposely made the redline a little higher in 1st - http://www.datazap.me/u/mfish123/1st-2nd-part-3rd?log=0&data=23
 
Last edited:

bradsm87

Corporal
Dec 15, 2016
162
53
0
My understanding is the bottom row (110) in the shiftmap tables are for the kickdown switch. You'll notice the values in them are very different to the rest. If you pull those values closer to the values above them, the car should behave better with the kickdown switch. Let us know if this works.

When I get my car back, I'll remove my BMS kickdown blocker and start playing with it.
 

bradsm87

Corporal
Dec 15, 2016
162
53
0
So what I've read seems to indicate that the pressure regulators in the mechatronics units in the 6HP21 max out at 4.6 Bar which is kind of confusing/contradictory because stock HS 1>2 in common maps upshift targets 5.6 bar and 3>4 is 5.15 bar :confused:

So Alpina M mode 3>4 already maxes out this 4.6 bar and the 1>2 isn't far behind at 4.1. Probably only a small reduction in the timespan possible for the 3>4 shift :( If we're already well and truly maxing the shift pressures by 630nm, there is really no advantage that I can see by increasing the torque limiter on an Alpina flash, other than just for convenience as not to have to mess with the DME flash to get below the limiter.

It would be so damn cool to put in higher max pressure solenoids then find and update the solenoid scaling to suit in the TCU.

I can't help but to dream about what we could acheive together if xHP worked with the community with tuning discussions :(
 
  • Like
Reactions: Rob09msport

doublespaces

Administrator
Oct 18, 2016
9,303
4,331
0
AZ
Ride
2009 E93 335i
So what I've read seems to indicate that the pressure regulators in the mechatronics units in the 6HP21 max out at 4.6 Bar which is kind of confusing/contradictory because stock HS 1>2 in common maps upshift targets 5.6 bar and 3>4 is 5.15 bar :confused:

So Alpina M mode 3>4 already maxes out this 4.6 bar and the 1>2 isn't far behind at 4.1. Probably only a small reduction in the timespan possible for the 3>4 shift :( If we're already well and truly maxing the shift pressures by 630nm, there is really no advantage that I can see by increasing the torque limiter on an Alpina flash, other than just for convenience as not to have to mess with the DME flash to get below the limiter.

It would be so damn cool to put in higher max pressure solenoids then find and update the solenoid scaling to suit in the TCU.

I can't help but to dream about what we could acheive together if xHP worked with the community with tuning discussions :(

Where did you read about the pressure limits? And what is happening when you set it above that value, nothing? Or is there a tuning issue at that point?