XHP DCT - Experience Thread

[carabuser]

I thought it is only me.

When the engine is warm it feels also like a tad more than before. Solution working from me giving initially more load request i.e. push the paddle harder for a second and clutches close immediately.

Yeh, same for me. Have to give it a good push to get it moving.

I think there's some tables in the trans related to clutch engagement from a stop that xHP haven't fully understood. The Z4 had the opposite problem initially where it would close the clutches immediately even without and pedal input so you'd go way to fast.

 

[carabuser]

I have creep disabled.

Are you on stage 2 or 3?

 

[RSL]

Not using their maps, but never liked stock 335is standing start engagment. TCC or something I logged is active until it's closed and the car is moving. M3 bin/disabled creep is very different.

You can do some DME tuning to mitigate it, but if it could be dialed in the TCU, that would be best.

 

[JohnDaviz]

Yeh, same for me. Have to give it a good push to get it moving.

I think there's some tables in the trans related to clutch engagement from a stop that xHP haven't fully understood. The Z4 had the opposite problem initially where it would close the clutches immediately even without and pedal input so you'd go way to fast.

You can do some DME tuning to mitigate it

I was thinking to overcome this issue with increased torque request in the lower rpm/load range up to 1000rpm.

 

[RSL]

I actually found the opposite, but it's worth playing around to find something that suits you. I liked the 1M tables or close by a large margin for part-throttle on non-M DCT bin, but there are other things in play on the 1M bin with them.

What it appears to come down to is the more torque/request, the more it does it's offset thing at standing start. You can probably add enough request to overcome max offset, but any normal stop will probably be a full-on launch then.

 

[houtan]

Are you on stage 2 or 3?

I am on stage 3. not sure if it matters, but i set the DSC off setting to performance launch vice burnout.

I also reached out to XHP regarding the rev limiter setting. Probably old news to everyone else, but they said the rev limiter needs to be raised in the DME before the raising the max shift point revs in D and S mode will work. Are you guys able to raise the rev limit on the N54? I have not heard of anyone doing it on the early N55s.

 

[JohnDaviz]

I am on stage 3. not sure if it matters, but i set the DSC off setting to performance launch vice burnout.

I also reached out to XHP regarding the rev limiter setting. Probably old news to everyone else, but they said the rev limiter needs to be raised in the DME before the raising the max shift point revs in D and S mode will work. Are you guys able to raise the rev limit on the N54? I have not heard of anyone doing it on the early N55s.

There are two different things.
a) Limiter hitting before DME Stock Rev Limit
b) Increasing Rev Limit. I increased to 7200 DME and TCU and still not working.

 

[houtan]

There are two different things.
a) Limiter hitting before DME Stock Rev Limit
b) Increasing Rev Limit. I increased to 7200 DME and TCU and still not working.

Thanks. so i thought i was experience letter "a", but it turned out to be the torque reduction setting for me. Now shifting at 6500 feels normal.

For "b", do you know if the settings to raise the rev limit work on a different car? Also, are those tables available in the n54 xdf's?

 

[carabuser]

Thanks. so i thought i was experience letter "a", but it turned out to be the torque reduction setting for me. Now shifting at 6500 feels normal.

For "b", do you know if the settings to raise the rev limit work on a different car? Also, are those tables available in the n54 xdf's?

All of the rev limit tables should be there. With the N54 there's stacks of documentation and even a public MSD81 DAMOS so even if the table isn't in the XDF it's easy to add in.

 

[dzid_]

Preface.
While I don't care about drag times, I do enjoy to slide around a bit, mostly low speed. I am Rally kind of guy.
I asked XHP whether their software would help with initiating a slide while driving in a any way, N-D clutch drop, paddles hold to clutch-kick, or the burnout feature. The answer was basically No. They said burnout mode is only for take-off. But I decided to buy it anyway.
Current setup:

1. 2011 135i N55 DCT
2. 135000 miles. It probably has still the original trans oil.
3. MHD ~E30 fuel and catlless downpipe + IC. So probably just under 400hp
4. XHP DCT (no OTS map)

Current flash:

1. MHD reflashed:
   1. E30 map - manual gearbox ,
   2. Linear Throttle
   3. No burp.
2. XHP flashed with:
   • Launch RPM @ 3000rpm
   • Torque reduction @500Nm
   • Torque punch @5
   • Creep off
   • Kickdown Launch - performance mode 3000rpm

Experiences after one good ride:

1. Cold start, very little creep, shift to R - engine bogged down - relatively normal for me; happened before.
2. No creep - still creeps tiny bit. At the moment I don't like it as it still requires brake pedal. WIll see.
3. Shifts initially a bit shaky, but were quickly adapting or warming up. One time when accelerating in 2nd gear I think clutches started slipping as the vehicle started to oscilate longitudinally and my head was banging. It's hard to even guess a cause with both MHD switched to Manual gearbox and XHP flashed. We are working blind essentialy
4. When stepping on the pedal, I heard clutch locking. Ok, that indicates to me either the adaptations were running higher microslip and adapting torque2pressure curves or kisspoint or basically it was normal microslip, but control wasn't as good yet. I believe stock software also runs microslip most of the time for good NVH, but normally I cannot feel the locking as much. Anyways...
5. Some shifts were buttery smooth comfy shifts. Kinda long. That's very nice for D ! Though, who knows what adaptations or driver profiling will do them later.
6. Now one thing that I don't like in D is holding the gear. In stock, that would only happend in Sport or S. Now it is always and it doesn't match my driving style. I often accelerate hard and coast. Now when I am costing it holds 2nd gear forever. I really dislike it in D and I will have to hold accelerator for it to upshift. Did XHP enabled it somehow? Not appreciated.
7. Launching with mid pedal is pretty good. I think thanks to lack of creep. But I will be testing with creep on. Especially that there is a second way (emergency way to launch )
8. Kick-down launch - this thing is good now. I already forgot how it was original, but I was never using it as it was slow. Not sure if it is thanks to turned off creep, but the engine speed rises quite quick and the clutch is applied agressively. It spins my crappy cold tires with no problem, even without burnout mode. Now, the big problem is the DTC traction control. In my car DSC is fucked. I was writing about it here. This time with the wheel slip, I got so massive traction intervention bang (from exhaust and brakes) that my whole body felt a joilt forward. It's crazy. But I am glad maybe there is a hope to at least partially fix it with M1 DSC software. For now I have to disable DSC completly, which is not ideal, because cruise-control doesn't work without it.
9. To my great, great, great surprise, contrary to what XHP suggested to me, the kick-down switch launching (allows RPM to rise and then apply clutch) works with non-zero speed. I think it does that as long as the RPMs are below the specified threshold, or basically there is no room for rpm to rise anyway. It works only in 1st gear, but that is good enough for me. It works in any driving mode, but you really need to have accelerator kick-down enabled. What it does effectively is a clutch-kick when pressing kick-down switch - Imagine driving around a small round-about and trying to slide the rear end. On old sofware, it would be really hard at low speed - the car would just accelerate gradually and eventually understeer. Here, you press accelerator kick-down, the RPMs rise (clutch slips) even though you were already rolling 10mph and then the clutch engages and you get torque surge and you know what happens next... At least my car would never be able to do that before. I understand why BMW didn't ship the car like that. It's a big liability, that with simply pressing the kick-down the car will slide around. But I love, it's better than some double-paddles hold in porcshe PDK. No fuss, just pedal to the floor, (as long as the DSC is off)

I will play a bit more with maybe turning creep back on and managing shiftpoints a bit.
Then rant:
I am not planning on getting OTS maps right now. Main advantage I see for me is that OTS disables kickdown in manual.
I don't really know what OTS maps gives, as the map description is just verbal. I assume they adjusted clutch engagement rates and shift lines a bit, but really there is only so much to be done with shift lines.
I don't think I actually care about ultra-short upshifts. I mean, if the shift gives me punch, I prefer for that punch to last longer! Adding more toruqe reduction and more punch reduces shift time anyway, so I again I am not sure what OTS map gives.
This is whole ultra short upshifts is just marketing from OEMs anyway and it's a patch to cover for bad control. That said, I do care about shift delay (clutch fill + torque swap phases time); not so much about upshifts speed change duration (which is what these parameters modify).
The only good thing about short upshifts (short speed change) is that since the control is a bit bad (e.g. input torque increase is inhibited after shift started), then in "aggressive" city driving spending less time in a shift can be better.

I would love to see XHP add options regarding downshifts though, as these are sluggish on this car.
Ok, enough of rant

 

[houtan]

The main feature I wanted was disable of the auto shift, which requires buying the map pack because it’s only available in stage 3.

It’s far from perfect for me as well. My biggest issue is some sort of torque limiter kicks in about 500 rpm before redline. It can supposedly be fixed by increasing the torque reduction setting, which worked sort of. But the second I tweaked my tune it started doing it again. It is going to be really frustrating if I have to constantly change this setting, especially when it was never necessary on the stock flash.

If I had a choice, I would keep the stock flash, creep disable, and auto shift disable.

 

[houtan]

checking back in here to see if anyone is self tuning and running this on an N55. I have been exchanging messages with XHP for the last couple of weeks. If i flash their stage 3 map, my car looses power around 6400. If i flash the stock trans flash, the car drives perfect and it shifts at 69XX. XHP says it's the engine tune.

So i searched my bin and saw a couple of torque tables related to the DCT that i had raised. Put one table back to stock and it seemed to help. The Stage 3 map reved to 6700 in manual mode, but power is still being cut early. My last test is to put the second table to stock and see if that helps.

It would be nice if there was some sort of documentation that mentions what tables should be stock for their flash to work properly or if someone would share what tables need to be stock. I sure will once i figure this out.

 

[Olza]

I guess we can change that torque-cut intervention 500 rpm before cutoff (in gts flash that hysteresis is different. Does it so important?
Also, if xhp used 335 shift maps then there mostly 6400-6500 shift rpm. You need to use 135is or other maps with 6900 cutoff.
Can you rewrite own bin with xhp? I can say you addresses and values to change

 

[RSL]

Yes, you can flash custom calibrations with XHP. I was neck deep in it before the M3 bin. Post or send the addresses and GTS values

 

[houtan]

Not sure if this helps, but when using my stock transmission flash, in manual mode, dsc off, sport button pressed, my transmission was shifting at 69XX RPM.

Per the various test I did for XHP, one of the custom settings is to change the shift points in D and S mode. I changed all shift points for all gears to 6k for D and S mode and it worked. Changed them to 7k and it was at 6400.

There was one table in the Torque manager of my XDF called Maximum Torque Characteristic, I have always run that at 999 from 1,000 to 7000 RPM. I changed that table to stock, and it helped a little bit, i could get to 6,700 ish RPM. So better, but not fully there. This is with Torque reduction disabled. Maybe the car is kicking the rev limiter in early in anticipation of the shift. I am going to try setting torque reduction to 1,000 to see if that helps.

 

[Olza]

took a quick look.
about shift limits. there are many conditions like shift lock, forced upshift, slippage and so on. engine not in limp or warmup?
maybe you have shift lock control signal from DME (do you use Stage3?).

change 6500 at 0x702E offset to 7000 (0x1964 to 0x1B58).

about cutoff offsets. you can try this:
target gear rpm + offset compared with shift lock control limit for overrun rpm threshold. GTS values for example.

also i think i saw some torque demand corrections near rpm limit, didnt find it now, maybe later in process...
at your own risk! its only mine subjective data, based on code analysis ))

 

[RSL]

I assume slippage table is what causes early upshifts on launch, or part of it. There's a table right after the slippage limits that looks like torques, are those torque reduction/corrections/offsets? What are the X break errors from (input vs. output internal, engine vs. wheel torque, some wheel slip factor, etc.)? Any conversion factor on break values or is it straight nm torque?

M3 shown, second table starts at 11BD0.

 

[Olza]

no this is not a torque, this is offset to reported idle speed target from dme (you can force them to 780 rpm in TCU). not sure it is downshift )

about rpm lim torque intervention. seems it works via that rpm thresholds (see picture) shift_lock_can_rpmlim-engine_speed and if throttle >20%

so TCU uses TORQ_AVL_MIN (TQ_LOSS?) targeting, decreasing torque by 3 nm each cycle.

 

[RSL]

Thanks, so hard to tell what some things are just looking at them. After I posted, I thought breaks might be rev gradients or something along those lines.

I'd have to check what TORQ_AVL_MIN crosses to, but there is a bare min torque based on lots of conditions (reserve/anti-stall).

 

[JorgeL]

My car.