MDCT... a mistake?

[enigma]

Its not a "some cars have it, some don't" its a driving style that causes it. I can make my car do it, I just choose not to.

[TLud]

Originally Posted by enigma

Its not a "some cars have it, some don't" its a driving style that causes it. I can make my car do it, I just choose not to.

I 100% agree with your belief that driving style accounts for most (albeit not all) of the reported issues. The question becomes which of these issues should simply be adapted to as the nature of the beast and which are severe or problematic enough that they should be changed/fixed?

[swamp2]

Enigma: I will have to mostly disagree with you on this recent exchange.

This new issue with the low speed lag when between braking and accelerating is a NEW issue with the NEW software. This issue happened with the last verison of the software but in my personal experience it was very rare. Others have concurred with this finding. With the new version of the software it is that it happens a majority of the time, but not quite all of the time. It is a bug and was either 1. An unintended consequence of other fixes or 2. An oversight based on opinions about how people actually drive, perhaps comparing US vs German driving styles. That being said I do agree that the DCT is not an automatic and requires some adjusting to and perhaps even a slight change to your driving style. And of course you still remain in control and can force the box to pretty much always do what you want, when you want. Bottom line is that the D modes, before the software update, were far superior in this particular regard which relates to the software correctly predicting what the driver actually wants at low speeds. Nonetheless this well reported issue is still a bug, and I'd pretty much guarantee it will be fixed. Just because you are clever enough to prevent the bug from happenening, perhaps with 100% reliability, does not mean it is not a bug.

[enigma]

Originally Posted by swamp2

Enigma: I will have to mostly disagree with you on this recent exchange.

This new issue with the low speed lag when between braking and accelerating is a NEW issue with the NEW software.

Yep, guess why..

As we both know you can only preselect up or down. The old software ALWAYS preselected the lower gear as long as it could. People complained that this caused upshift lag. So BMW changed it to preselect up in many cases. So now people found a case where they could trick it and get downshift lag.

Some people just like to complain rather than understand the machine. The gearbox cannot preselect both up and down at the same time. If you know you are going to need a burst of acceleration you have all the controls you need to get the car to do the right thing. It just cannot always guess right without the driver giving it a hint.

autos with torque converters get away with this because the converter unlocks at low speeds.

[swamp2]

Originally Posted by enigma

Yep, guess why..

As we both know you can only preselect up or down. The old software ALWAYS preselected the lower gear as long as it could. People complained that this caused upshift lag. So BMW changed it to preselect up in many cases. So now people found a case where they could trick it and get downshift lag.

Some people just like to complain rather than understand the machine. The gearbox cannot preselect both up and down at the same time. If you know you are going to need a burst of acceleration you have all the controls you need to get the car to do the right thing. It just cannot always guess right without the driver giving it a hint.

autos with torque converters get away with this because the converter unlocks at low speeds.

An interesting possibility. However, I don't think it is the difficulty in next gear prediction that is causing this particular lag. I have two reasons to think in this way.

1. If they have added a bias toward preselection of higher gears (or in fact simply made it always do so), then the situation would be completely reversed and now all downshifts would suffer from a similar delay after paddle actuation. However, this did not happen. I would say there has been no change in the delay between downshift paddle operation and the actual beginning of the first clutching operation. This kind of shoots holes in both of our theory that preselection was the root of the upshift lag between paddle actuation and the beginning of clutching. It points to the previous upshift lag as again just a bug, glaring and simple, but just a bug.

2. When this issues shows itself shifting does not need to be involved and typically is not. Coasting into a slow neighborhood corner in D and say currently in 2nd gear, "trail brake" lightly through the corner and then quickly get on the gas. You will most often get the lag but the car does not shift to 1st nor to 3rd. It simply stays in 2nd, the throttle system literally ignores the gas pedal movement for a bit, and then revs when the software begins "catching up" with the pedal. No matter what the preselection algorithm was doing a gear change was not involved. As well it is not a delay that seems to be caused by the clutch on 2nd getting stuck open. Is it really reasonable to think that the throttle command was ignored, all the time while in 2nd gear, simply because the software could not appropriately choose/decide if 1st or 3rd as the preselected next gear?

Lots of speculation going on here for sure, but I think my case is still fairly strong.

Again, despite the seemingly glaring/obvious/intrusive/serious nature of this issue. I still think it is a bug that will be fixed.

[Bdez]

Without wanting to add to the conjecture through rumours etc I got this comment from my dealer after discussion how I was experiencing the problem (which is no different to anyone else’s comments here). Hope I’m repeating it faithfully but anyway …

The comment through BMWA was that some of the fuel economy characteristics of the software were interfering with DCT performance which was causing or contributing to the issue at the low end.

Before anyone tries to prove me wrong or points out the obvious that fuel economy is not a primary design feature of the M3, I’m not a engineer and am not saying that this is the golden answer to the problem - I think what the comment indicates is that the issue is not a simple fix as software issues often aren’t.

My biggest gripe with the problem is changing lanes from a standing stop and getting the lag - not a problem in S mode at all though !

Anyway , just one more bit of info to consider with all the rest of the comments.

[BigAl2]

I picked up my E93 DCT in early September and IMO 3500km is not even enough to completely understand all of the different setups you can try to get the most performance and or comfort out of your M3. There are some folks here dissing the DCT that have no idea. If you are commuting to work in traffic (no probs on the way in at 6:30, but the shits at 4:30 PM) a 6MT sucks. If your M3 is a weekend car for cruising the windy switchback and you want to feel the road then the 6MT is for you. Personally I think it's a helluva tranny. Too bad I've parked it and driving an Acura TL with snows for the winter. Next April though I'm gonna be luvinit again.

[Johndoe]

Originally Posted by bdh2009

My biggest gripe with the problem is changing lanes from a standing stop and getting the lag - not a problem in D mode at all though !.

Did you mean not a problem in "S" mode..?

[Bdez]

Originally Posted by Johndoe

Did you mean not a problem in "S" mode..?

Thanks for the pickup - I did mean "S" !