Wednesday 19th September 2007

Paddle Shift M3 Gets A US Debut

BMW prioritises launch of double-clutch gearbox

---

Go paddling in your M3

Despite the SMG paddle shift ’box being standard fit in the existing M5, buyers of the new M3 will have to wait a little longer before they can change gear without taking their hands off the wheel.

This is because BMW have taken a clean sheet approach for the E92 (kind of) by nicking the much-respected DSG twin-clutch idea from VW and refining it with the help of specialists Getrag. Apparently the new set-up will be called MDCT, and features seven ratios which can be swapped either by an X5-style curvy gear stick or wheel-mounted paddles.

Demand for the system has been sufficient enough to bring forward the launch, although the actual dates aren't yet clear. The word on the street is that the LA autoshow this November or Detroit motorshow next January will get the nod for the launch, possibly alongside the 4-door variant and maybe a Touring concept to gauge public opinion.

[DoctorD]

This is not news. BMW had already communicated several months ago that M-DCT would be fitted in production cars from March '08.

Also it is not BMW that have nicked the idea from BMW, but in fact Getrag who have collaborated with Borg Warner (who originally developed DSG for VW) to bring DCT type boxes to the wider market. Previous BW DCT systems were limited by the torque they could handle, but Getrag have been leading the way in developing higher torque units which have been then further developed with BMW and Ferrari. I understand that BMW are still evaluating M-DCT to ensure it provides the level of sportiness to replace SMG, and they are also considering its use to replace the current ZF supplied Auto boxes in future BMW models.

Edited by DoctorD on Wednesday 19th September 12:20

[MK4 Slowride]

It is a very good gearbox. Also it'll be nice to see the M varient across the range, saloon, touring, coupe & convertible like you get with most other brands top end motors.

[Wolvreen]

DSG seems to be the way to go for clutchless.
Are Porsche doing their own or thru Getrag?

[Mr Whippy]

Will it ever be as good as an auto in pure auto mode though, without wearing the clutch plates excessively?

Dave

[edb49]

DoctorD said:

This is not news. BMW had already communicated several months ago that M-DCT would be fitted in production cars from March '08.

I think the news element of this is that BMW are bringing forward the launch, or at least that's how I read it.

Also, when BMW say that demand is so high for the DSG, do they really mean demand is very low for the M3 and they're having trouble shifting as many units as they thought they would?

[alex_123_fra]

edb49 said:

Also, when BMW say that demand is so high for the DSG, do they really mean demand is very low for the M3 and they're having trouble shifting as many units as they thought they would?

[DoctorD]

edb49 said:

Also, when BMW say that demand is so high for the DSG, do they really mean demand is very low for the M3 and they're having trouble shifting as many units as they thought they would?

I think it reflects the fact that a large percentage of the M3 order book have been transferred (i.e. delayed) to wait for the M-DCT model, hence they need to focus on this M-DCT version and produce it as early as possible.

[Dabe]

A funny fact is the M5 is available in the US with a manual 6, and it's told to represent 50% percent of sales. Maybe the fact that the M3 will be mated to a DCT is creating a new interest for that kinda tranny.

[edb49]

DCT gives the hottest numbers, 0-60 time etc. Everyone likes to pretent their Raikkonnen by flicking paddles behind the steering wheel too.

[Mr Whippy]

edb49 said:

DCT gives the hottest numbers, 0-60 time etc. Everyone likes to pretent their Raikkonnen by flicking paddles behind the steering wheel too.

I can imagine with two or three 'instant' shifts (basically 0 time) vs two or maybe three shifts on a manual adding up to a second or so, the 0-100 time may well drop, in real terms, by ~ 5-10%, which is quite stunning but perfectly reasonable.

Not sure I personally would like the DCT for general road use, but on a track and for the CSL that is going to add up to alot I would imagine, and for those who like paddle shifts and outright performance...!

Dave

[Ashok]

Anything is better than SMGIII

[wushuhsu]

Man, Porsche has already lost the engine wars (no DFI, ridiculously low horsepower in its production Cayman, Cayman S, 911S lines, no twin turbo less than $100,000). Now it's falling more and more behind in other areas like DSG, which it's been struggling to get out for years. I'm so glad I bought a BMW 335i twin turbo rather than a Porsche. I'm not even going to mention headlights that turn off automatically, standarad swiveling Xenon headlights, remote keyless entry and push button start, better navigation system, and iPod and Bluetooth.

[horton]

wushuhsu said:

Man, Porsche has already lost the engine wars (no DFI, ridiculously low horsepower in its production Cayman, Cayman S, 911S lines, no twin turbo less than $100,000). Now it's falling more and more behind in other areas like DSG, which it's been struggling to get out for years. I'm so glad I bought a BMW 335i twin turbo rather than a Porsche. I'm not even going to mention headlights that turn off automatically, standarad swiveling Xenon headlights, remote keyless entry and push button start, better navigation system, and iPod and Bluetooth.

If you don't want a manual gearchange, and things like better headlights and navigation are important to you, then I'm sure a BMW would be a better choice - but then again I doubt things like that are priorities for the core of Porsche owners.

[Mr Whippy]

Just thinking about DSG.

So, what happens to the engine speed during a shift? Does it change instantly, or do the clutches slip over the period the revs need to change?

I assume under a full-bore acceleration then the clutch shifts over instantly and the engine energy gives almost an acceleration boost.

But what happens in day to day driving? To rise from 4000>6500rpm say, on a downshift, surely the electronic engine blip takes around 0.5s - 1.0s, so is the shift complete, but the clutches ramp in and out as the rpm's move from the pre.gear rpm and the new.gear rpm?


Ultimately it still sounds alot down to the software being smooth and knowing you want to be... I can be aggressive on upshifts and then even half way through a shift go smooth... I'm guessing DSG has the same I/O limitations as SMGIII in this respect. It will see your intent and shift appropriately, but it will never be as smooth as the proper auto or manual until it's sensors can poll more variables that indicate driver intention...!?

Dave

[Craigy_97elise]

wushuhsu said:

Man, Porsche has already lost the engine wars (no DFI, ridiculously low horsepower in its production Cayman, Cayman S, 911S lines, no twin turbo less than $100,000). Now it's falling more and more behind in other areas like DSG, which it's been struggling to get out for years. I'm so glad I bought a BMW 335i twin turbo rather than a Porsche. I'm not even going to mention headlights that turn off automatically, standarad swiveling Xenon headlights, remote keyless entry and push button start, better navigation system, and iPod and Bluetooth.

I think you ought to read up a little on Porsche and dual-clutch gearboxes. Porsche used them in versions of the 956 and 962 racecars in the 1980's. Porsche also hold a few patents in the area and they are under development for production models at the moment.

As to the comments about "losing the engine wars" due to not having a twin turbo under $100K... well if porsche can profit from selling over $100K because people will buy it, why not? It might help explain why Porsche is so profitable.

The Cayman engine is slightly hemmed by marketing since Porsche obviously don't want the cayman to damage sales of the 911.

I doubt Porsche are as technologically inferior as you suggest...

Craigy

[Craigy_97elise]

Mr Whippy said:

Just thinking about DSG.

So, what happens to the engine speed during a shift? Does it change instantly, or do the clutches slip over the period the revs need to change?

I assume under a full-bore acceleration then the clutch shifts over instantly and the engine energy gives almost an acceleration boost.

But what happens in day to day driving? To rise from 4000>6500rpm say, on a downshift, surely the electronic engine blip takes around 0.5s - 1.0s, so is the shift complete, but the clutches ramp in and out as the rpm's move from the pre.gear rpm and the new.gear rpm?


Ultimately it still sounds alot down to the software being smooth and knowing you want to be... I can be aggressive on upshifts and then even half way through a shift go smooth... I'm guessing DSG has the same I/O limitations as SMGIII in this respect. It will see your intent and shift appropriately, but it will never be as smooth as the proper auto or manual until it's sensors can poll more variables that indicate driver intention...!?

Dave

According to what I have read, Borg+Warner DSG (ie. the VW system) shifts take about 8 milliseconds (that's 0.008s for people who don't know what a millisecond is).

Apart from anything else, this allows the engine little time to drop revs to match the new gear (assuming an upshift is taking place), or to blip up to match the new speed if there's a downshift.

In a given environment, the engine may have to drop 2000rpm in this time (eg. shifting from 6500-4500). Bear in mind that at 6500rpm, the engine itself is turning over about 108 times a second, otherwise known as about 0.009s per rev.

Therefore, roughly speaking, the engine has to match the new gear inside about 1 revolution. I'd expect the clutch to have to slip a bit in order to accommodate this.

If any of this reasoning is flawed, can someone correct me? (I am thinking the flywheel may play a part too).

Craig

[Mr Whippy]

That maths sounds right.

Basically, it changes gear in 8ms, but it's not transmitting maximum torque, probably, any faster than an SMGIII gearbox... ie, the key is the time the engine takes to change rpm when the shift time is this fast!?

All quite interesting, it'd be nice to actually see a proper gear > gear operational cycle of what happens. I'm guessing all out the issue is still conrod strength etc, decellerating the rotating masses in one revolution by maybe 50% is alot of stress I'd imagine!?

Dave