An Old DCT Lag Rant

[CanAutM3]

Originally Posted by SenorFunkyPants

Actually F1 use a manually operated single clutch sequential gearbox.

No they don't.

A manually operated sequential gearbox is what you find on most motorcycles, where it is human power that moves the clutch actuators and the gear change cylinder. On an F1, it is servos that actuate the clutch and cylinder, hence automated.

I guess I see your point to the fact the servos act at the request of a human

Originally Posted by SenorFunkyPants

Actually F1 use a manually operated single clutch sequential gearbox.

The DCT essentially shifts in zero ms as there is no moment when torque is not being applied to the output shaft. During the "shift" there will be some loss to heat as one clutch pack is disengaging while the other is engaging but there is no break in torque output...which is why your passengers head doesn't move during the shift (best observed in S2).

We agree on this one

[US///M3]

Originally Posted by M3takesNYC

F1 actually uses a dual clutch don't they?

No

So quit telling your friends your M3 has an F1 gearbox.
the F1's is similar to a motorcycle gearbox, it has no synchronizers, the engine ECU matches the gearbox speed, for the specific gear and corresponding road speed.
It's a manual geabox, with automated selectors changing gears...auto-manual.

[SenorFunkyPants]

Originally Posted by CanAutM3

No they don't.

A manually operated sequential gearbox is what you find on most motorcycles, where it is human power that moves the clutch actuators and the gear change cylinder. On an F1, it is servos that actuate the clutch and cylinder, hence automated.

I guess I see your point to the fact the servos act at the request of a human

F1 "seamless" gearbox technology is quite interesting - they get round the single clutch plate restriction by using a system that allows two gears to be engaged at once...its hard to find a definitive description of exactly how they work but this seems to give an idea.
http://www.zeroshift.com/pdf/RcarN6V15_Zeroshift.pdf

[CanAutM3]

Originally Posted by SenorFunkyPants

F1 "seamless" gearbox technology is quite interesting - they get round the single clutch plate restriction by using a system that allows two gears to be engaged at once...its hard to find a definitive description of exactly how they work but this seems to give an idea.
http://www.zeroshift.com/pdf/RcarN6V15_Zeroshift.pdf

That was an interresting read. Thanks for sharing

[MikeBlaszczak]

Originally Posted by Radiation Joe

I doubt you were around when this problem was plaguing us, so I'm not sure your perspective is quite right for chiming in.

Nothing's wrong with my perspective.

[Radiation Joe]

Originally Posted by M3takesNYC

Your still complaining or crying over a stupid transmission on a car that had some lag?

Must really suck to be consumed by such minutia in life.

Sure it bugs me if my car is not working exactly right but the thought lasts a day if that. 2 or 3 years later and this still brings out this much emotion in you-scary

Originally Posted by MikeBlaszczak

Nothing's wrong with my perspective.

Typical internet experts. No clue what the background of the story is, but feel they have to share their misinformed opinions.

I spent two year fighting with BMWNA over a seriously unsafe condition that they knew about but would not admit. This was not some 50msec argument about shift times; it was a car that would sit in an intersection with a semi bearing down on me that was stalled due to a transmission error.

I put between 25k and 30k miles per year on my car. When it tries to kill me, I get upset.

[CanAutM3]

Originally Posted by Radiation Joe

Typical internet experts. No clue what the background of the story is, but feel they have to share their misinformed opinions.

I spent two year fighting with BMWNA over a seriously unsafe condition that they knew about but would not admit. This was not some 50msec argument about shift times; it was a car that would sit in an intersection with a semi bearing down on me that was stalled due to a transmission error.

I put between 25k and 30k miles per year on my car. When it tries to kill me, I get upset.

I know where you are coming from, I got one of the very first DCT here in Canada back in 2008. The lag, when you experienced it, was pretty dreadful. I learnt to drive around it the best I could, but the lag would pop up once in a while in the most inopportune moment.

With the various software updates it has become much less of an issue. There is sometimes a slight hesitation, but no more of the dreaded lag. The counterpoint is that the transmission has become "clunkier" with the software fixes, as it is rushing into the downshifts.

I guess some had/have it worse than others.

Regardless of this issue, I have been truly enjoying the DCT

[MikeBlaszczak]

Originally Posted by Radiation Joe

No clue what the background of the story is, but feel they have to share their misinformed opinions.

Please fill me in: is this the drunk Radiation Joe, or the sober Radiation Joe?

You've repeatedly shared all the background information. You didn't like the car and thought you should sue. You didn't sue and instead just bought another one. The only person you have to blame is yourself so you got drunk and re-posted your rant to get more attention.

[Radiation Joe]

Originally Posted by MikeBlaszczak

Please fill me in: is this the drunk Radiation Joe, or the sober Radiation Joe?

You've repeatedly shared all the background information. You didn't like the car and thought you should sue. You didn't sue and instead just bought another one. The only person you have to blame is yourself so you got drunk and re-posted your rant to get more attention.

Yup. You pretty much nailed it on the head. Congratulations.

[swamp2]

Originally Posted by CanAutM3

IMO, if the most performance oriented shift on a DCT would interrupt power delivery for 50-100ms, it would entirely negate the whole point of DCT transmissions. If the best single clutch can do it in 40-50ms, where is the performance benefit of DCT ?

I've never heard of them being that fast. There were claims that the E46 SMG (or perhaps SMG II) could shift in less than 100 ms. Actual testing revealed shifts around 250 ms. I posted a link to the article with the test results here long ago. I'm sure single clutch automated manuals have improved since that time. However, separating manufacturers claims of shifts times vs. actual shift times can be difficult.

Keep in mind that the performance of the M-DCT varies widely depending on both the Drivelogic mode, how aggressively one is driving as well as which shift is being considered (a 6-7 shift can be faster than a 1-2 due to the closer ratios). There is nothing hurting performance vs. any manual for shifts

Originally Posted by CanAutM3

It is physically impossible to have the two clutches simultaneously engaged. Either one, or both are slipping. Static friction is stronger than dynamic friction. This implies that the first clutch remains fully engaged even if the second clutch is progressively applied (and slips). As the second clutch is applied with more pressure, it transmits more and more torque to the transmission. As soon as the first clutch is sufficiently released, it starts to slip and the gearbox gets driven only by the second clutch. The second clutch then continues slipping until the gearbox spins at the same speed. With good timing of clutch engagement and disengagement, the inertia of the engine can be recuperated in the process. With the power of electronic control, this can be (relatively) easily achieved.

I do not see any physical reason for both clutches to be disengaged simultaneously and interrupt power delivery during an upshift. Even if both clutches are simultaneously partially engaged, we are talking milliseconds here. IMO, the slippage on a standing start is far more severe. To minimize wear in less aggressive shift modes, I can see a combination of engine management adjustments (a very slight cut in engine power output to accelerate the RPM drop) and clutch application to make the shift smoother and reduce the severity of the clutch slippage.

Sounds completely reasonable and in line with descriptions I and other forum members have stated in the past.

Originally Posted by CanAutM3

I believe that the quoted DCT shift times is the time it takes from one clutch being fully engaged to the second one being fully engaged. However, contrary to a single clutch transmission, power continues to be transmitted in the process. Hence the "zero shift time" for the purpose of the discussions in this thread.

IMO, it is possible to make the transmission of power truly seamless with a DCT .

It may be possible. The testing I did prior (and posted) was not conclusive. It may even be possible that there is a positive surge (derivative of acceleration) during an upshift from very clever engine and or clutch phasing that dumps a bit of extra flywheel and drive train inertia into the wheels causing a small acceleration GAIN during the short shift process.

Originally Posted by CanAutM3

From the Getrag website:

The GETRAG Powershift dual clutch turns the dream of fully seamless gearshifting into reality. While one gear is engaged, the system has already preselected the next. Once the relevant rpm has been reached, one clutch is opened while the second is closed simultaneously, precluding any interruption in tractive force. The intelligent software control means that sportiness and convenience are guaranteed alongside high efficiency.

I could not find any significant technical data though...

Marketing, marketing...

Unfortunately, without some additional testing the "nitty gritty" details remain a bit unknown.

[V8FunNaturally]

Originally Posted by swamp2

I'm not sure I can answer that, probably only Getrag and/or BMW engineers can... It's probably the exact same group of engineers who can answer these questions about lag as well!

Again my hunch is that some shifts do have an effective zero shift time, however, clearly not all shifts in all modes have this. There is a fine balance in being able to phase 2 clutches without dangerously having both partially engaged at the same time. Also, there is the question of dumping excess crank/flywheel angular momentum into the drive train. This will help performance albeit probably minimally. This momentum dump can probably be controlled as to it being fast or sharp or slower and steady again depending on the Drivelogic setting. It also may be the case that the mode that feels the fastest might not actually be the best performer. I suspect it is the former but certainly cannot prove that.

I certainly feel a wide range of shift times myself, again depending on level of aggressiveness in driving along with Drivelogic mode. My estimate of 50-100 ms for WOT upper Drivelogic modes is based on a variety of quotes from OEMs, along with CarTest simulation, observations about vehicle performance as well as my somewhat crude testing.

should have gotten a MT.
+100

I told you so in 2007/2008. My concern with the to be released DCT then was that the algorithm used to decide the shifts was a much bigger risk for the buyer than the whole mechanical side of it. Already then I argued that the DCT control algorithm woud be (there were leaks) artificial intelligence/Adaptive software based and that was the red flag for me to choose MT over DCT. No matter how fast DCT can mechanically switch gears from the time it "intelligently" decides to do so the driver has already given up two degree of freedom : choosing the gear he want whenever he wants and engaging it softly/aggressively if he chooses or anywhere in between. The DCT manual mode is fine for the track, it isn't as practical as MT for daily driving even though you can select different level of aggressiveness you are sampling a smaller space of possibility than the MT can never being able to dose the clutch release yourself as you desire it dynamically at every instance.
6MT

[SenorFunkyPants]

And there speaks a man with little idea of what he is talking about.

[CanAutM3]

Originally Posted by swamp2

Marketing, marketing...

Unfortunately, without some additional testing the "nitty gritty" details remain a bit unknown.

I was recently watching the Evo interview with Andreas Preuninger on the new 991 GT3. He mentions a "reaction" time of less than 100ms on the PDK for the GT3, making it the fastest transmission available on the street...

Now, "reaction" time, what is that supposed to mean? Paddle actuation to completed shift? Paddle actuation to shift initiation?

You are right, it is mostly about marketing. Without precise testing or in depth technical details from the manufacturers, we can only speculate. I personaly am a believer that it is possible to get uninterupted power delivery with DCT .

[swamp2]

Originally Posted by sunsweet

should have gotten a MT.
+100

I told you so in 2007/2008. ...

These comments are really so beyond irrelevant.

What part about the DCT lag being both rare (now) and being somewhat like a "pimple on the face of supermodel" do you keep missing and completely dismissing? Even with the lag I am 100% happy with my decision and would never go back. It is better or the track, at the strip and around town as well and I have used it at all of those places.

Last but not least the 6MT in the M3 is hardly an excellent, crisp and short throw transmission...

[The Automotive Enthusiast]

No lag is being experienced on my end. I do experience bucking when the car is cold though. I was never a fan of BMW's manual transmission. Too rubber feeling.