Has anyone turned a manual gearbox into a semi-auto one?

[xxChrisxx]

I thought the OP was inquiring more about a home brew approach to AMTs.

[woohoo]

Sorry, should have been more clear.
My point was that there are quite a few of these transmissions our there, and so there must be quite a few resources for them.
Its also the same system used in the E36 of the BMW M3.

Broken down, it is an electrical clutch pull, with three solenoids controlling the movement of the gear lever.
Each controlled by a potentiometer.

Of course it is easy for an armchair fan to talk

[marcush.]

I think the mono clutch approach is not really suited for performance use .
A conversion to a two clutch system is a new gearbox esentially as you need to split the box in two (clutch one for gears 1,3,5 and clutch 2 for 2,4,r for example) .
Considering the engineering effort needed for application/mapping /integration only that´s a exceeding the possibilities of a very talented amateur by a long ,long shot.
You could potentially get a ETK or CCP-TCU to be even able to manipulate the settings but still no access to the engine side...which is necessary..
Not feasible unless you are stuffed with money the gearbox manufacturers will be glad to help.

[autogyro]

F1 uses mono clutch and is at the peak of layshaft 'conventional' gearbox systems.
Dual clutch layshaft systems may or may not shift faster than single clutch layshaft systems depending on the shift/dog mechanism used and the electronic control program but there is a disadvantage with extra weight and torque loss.

All modern control systems working on layshaft boxes limit the shifts to sequential or 'semi' sequential use and prevent full driver control over gear shifting.
If you want to win improve your gear shifting skills dont add a 'lego' kit to do it badly for you.

[marcush.]

F1 uses mono clutch and is at the peak of layshaft 'conventional' gearbox systems.
Dual clutch layshaft systems may or may not shift faster than single clutch layshaft systems depending on the shift/dog mechanism used and the electronic control program but there is a disadvantage with extra weight and torque loss.

All modern control systems working on layshaft boxes limit the shifts to sequential or 'semi' sequential use and prevent full driver control over gear shifting.
If you want to win improve your gear shifting skills dont add a 'lego' kit to do it badly for you.

yes on all accounts.
I had the idea the guy wanted to convert a synchro box ....i should have read if that´s the case before makking that claim.

[HowlerMonkey]

uhm... Ferrari, Alfa Romeo and FIAT have been doing this for years..

The "F1" shifting mechanism in the Ferraris, the "Selespeed" in Alfa Romeos, the "Dualogic" of the FIAT are all the same family of electro hydraulic shifting attached to a normal manual gearbox

Yes they did but your answer is not related to the topic of turning a manual gearbox into a semi-auto one.

[riff_raff]

Sorry for coming late to the discussion. The question in the OP was about modifying a M/C engine/gearbox for automated shifting. In theory, it is entirely possible to accomplish, and has been done with numerous manual gearboxes. But as many others have pointed out, it would require substantial amounts of modification to the engine controls, and serious amounts of software and control system development for the gearbox itself.

Having said all that, your choice to modify a M/C gearbox makes some things less difficult, since most M/C gearboxes use large numbers of gear sets with close ratio spacing. And they also use dog rings for shifting, which are very durable and make the gearbox less sensitive to imprecise speed synchronization during a shift. However, M/C engines also have much lower polar inertias, so it would likely be more difficult to accurately control its engine speed during an automated shift event.

[marcush.]

isn´t it exactly the other ways round -a dog box does need the active input of the driver performing a synchronising action -the throttle blib do allow engagement of the next gear (on downshifts) and quick and modulated shift action on upshifts to not hurt the dogs..
At least that´s the case with car type gear steps and the mass involved with a car.

I confess a synchro type gearshift will always take some time to allow the gears to be changed so not really the choice for ultimate speed shift .

I ´d think the big thing here is less the gearbox actuatio as is the integration into your drivetrain software..It´s not like you just need an ígnition cutout and everythings fine ?
An automated box will bang in the gears no matter if all is fine and well or something is not quite fitting in terms of load revs or whatever.
No wonder in the early days of automatic shifting in motorsports there were lots of issues .As we see todays this has been inverted entirely and transmission collapses are a thing of the past now as are blown engines due to overrevs ...as the powertrain brains do not allow for anything shift when parameters are not plausible or within threshole.

[Tommy Cookers]

@ r r and marcush
mostly the car dog-box runs at engine speed and the bike dog-box at about 40% of engine speed
is this is a big factor eg relative to providing the right conditions for the dogs to work well ?

even shaft-drive bikes have not had truly engine-speed boxes since 1969 (such had notoriously poor shifts)
the eg 1200cc bike box and clutch will need much greater torque capacity than a 1200cc car equivalent (if this exists)
the H16 BRM was notorious for having its clutch 'in the wrong place', the input side of the gearbox not the output side of the engine
(this seems to have been for 4wd capability)
but the F1 Cosworth (and other F1 units ?) now has the clutch on the input side of the gearbox ?
to contribute towards a suitably high crankshaft resonant frequency

presumably the paddle-shift technology can be very helpful and liberating to the designer

[autogyro]

The BRM H16 had two crankshafts geared together.
The position of the clutch was as a result of dynamic balance requirements rather than 4x4.
Ferguson four wheeled drive was looked at though as Pilbeam had developed an F1 BRM car previously with it.

In the late 70s I designed an epicyclic electricaly shifted MC gearbox for racing.
Honda looked at it but decided an auto or semi auto gearbox would detract from the hairy motorcyclist image they wanted to promote through their racing.
Ducati used a pulse ignition shifting device based on the concept which improved the ghastly sequential MC stepped layshaft shifting of the gearboxes which are forced to accept increased torque from the input so as to reduce the input shaft rpm to workable levels.
They still use the same heath robinson designs today.
I talked to Norton about using a proper epicyclic semi auto power train using their engines when they gave my late friend Ken Wallis a couple of their engines for his autogyro development.
The Norton would have had a completely re-thought control layout to do away with the ancient MC controls still used today.
The brakes would have both been on foot pedals, push button gear shifts up and down (why on earth use paddles!?) and just one lever, the clutch. Of course the Norton rotary was far too fast for the Japanese so the world economic system banned it and the demise of Norton is now history.

[machin]

I have no doubt that Auto will see this as an abomination, however, the OP did ask "Has anyone turned a manual gearbox into a semi-auto one?"

I'm not going to get into the positives and negatives of such a system, but the answer is Yes: according to this website:-

http://web.archive.org/web/200803160512 ... ml?id=1459

"The [Subaru WRC2007's] gearbox retains the road car’s H pattern but uses a hydraulically and electronically controlled semi-automatic gear change system to shift each gear in fewer than 0.1seconds. "

I'm guessing that the H-pattern was retained because the rules required the team to use a "production" gearbox, but the team clearly thought it beneficial to provide the driver with sequential control -presumably via paddles mounted behind the steering wheel.

If the OP could find details of this system, maybe he could copy this?

[andylaurence]

There's several H-pattern to sequential kits and they can be pneumatically actuated with off-the-shelf kits should you so wish. Here's an example of one installed this week and you could automate it with a XoomBox. You'd want to get a dog-engagement gearset to replace the synchros. As far as the OP's desire to do a conversion on a motorbike engine, it's not rocket science - you need something to pull/push the lever, something to blip the throttle and an ECU like the XoomBox to cut power to the engine and fire the gear lever in the right direction.

[autogyro]

I have no doubt that Auto will see this as an abomination, however, the OP did ask "Has anyone turned a manual gearbox into a semi-auto one?"

I'm not going to get into the positives and negatives of such a system, but the answer is Yes: according to this website:-

http://web.archive.org/web/200803160512 ... ml?id=1459

"The [Subaru WRC2007's] gearbox retains the road car’s H pattern but uses a hydraulically and electronically controlled semi-automatic gear change system to shift each gear in fewer than 0.1seconds. "

I'm guessing that the H-pattern was retained because the rules required the team to use a "production" gearbox, but the team clearly thought it beneficial to provide the driver with sequential control -presumably via paddles mounted behind the steering wheel.

If the OP could find details of this system, maybe he could copy this?

I visited Prodrive in 2002 and consulted with David Dupont.
Subjects covered under confidentiality, were development of their H pattern semi and auto shift gearbox and modifications to a Group N rally spec engine for use in a high performance autogyro for record attempts.

We discovered that the two horizontally opposed cylinder heads were symmetrical and could be reversed to give an updraft induction with the exhaust feeding the turbocharger on top for aviation use.
A different camshaft and ambient atmospheric controlled boost pressure gave 200 bhp at 4000 rpm at 15,000 feet altitude perfect for altitude record attempts with the limited prop diameter.

The Subaru flat four is IMO an ideal engine for light aircraft so long as you don't compromise it with reduction gearing.
Subaru of course avoided the application because of the modern attitude to marketing, regulation and health and safety.
We live in an emaciated engineering world.

The gearbox worked OK as a compromise.

David is now at Hewland Engineering.