F1technical.net

I know it is available to buy 1 but it is so expensive, did anyone try to build 1 by yourself?

What would be the point?
Both H pattern and sequential are mechanicaly just different selector mechanisms.
It is possible to cobble up a very complex mechanism that could be added to the basic H pattern selector fork design to achieve a sequential selection.
It would be complex and subject to potential mis shifting and jamming.
The proper way would be to design a new casing to accept 'scrolls' acting on the forks etc.
This would basicaly be a new gearbox design.

Out of interest sake, what car are you wanting to do this for?
I know of a Japanese manufacuterer of aftermarket car parts who has built a kit to change H patern to sequential for a cable shift box. The kit is not gearbox invasive and is thus more or less plug and play for any cable shift gearbox. Similarly I have a South African engineer who has designed a similar system (independently) which works on a similar fasion although with fewer moving parts.

The SA developed kit is in my opinion a very pretty solution, PM me for more details on the SA built and developed kit.

Regards,

I agree 100% with mr gyro... what is the world coming to .

You can convert the shifter in car to a push pull, but at the end of the day it's still a manual H box transmission as you won't be changing the gate on the box or the way the forks are moved. Sequential shift forks are actuated by a barrel cam.


Infact I'd argue that it's a completely awful idea to convert an H box to sequential. If you are barrelling town the stright into a 2nd gear hairpin...? A block shift from 5th to 2nd at the end is the best way to do it. With this system you have to do 3 shifts in the time it takes to do 1. A true sequential box with have a calibration to shift either with or without the clutch, depending on application. With this you have to actuate the clutch yourself and you have no control over the shift speed or agression. Can you imagine forgetting to dip the clutch all the way and having this thing try to ram it into gear?

People tout the advatage of these kits as 'not missing a shift'. The answer to that is to learn to drive properly and accept that from time to time, you'll balls up a shift.


That's not to say it can't be done, but it's really Heath Robinson.

If you are after shift times you can get 95% of the way there by designing a nice H pattern shifter. If you are talking about something fitted into an OEM style casing designed for syncho shifts you shorten the fore/aft throw a lot and the side to side throw a little. Get a nice sensor setup for the shift cut, move the whole assembly close to the steering wheel and it will work pretty darn good and close to a sequential in terms of the shift time.

A french company called Selexy tried to market a sequential gear shift for "H" gate boxes. I don't know if any ever got sold.

scarbs wrote:A french company called Selexy tried to market a sequential gear shift for "H" gate boxes. I don't know if any ever got sold.

Not many I would have thought.
I built a hydraulicaly operated shift fork mechanism years ago for the mini box, that worked ok but was a bit heavy.
Needed a pump and all sorts.
It needed a more advanced electronic control system at the time (early 1970s).
The early developments of that gearbox, the early F1 auto boxes and slector mechanisms like the Selexy all came under a common description of gear shifting we used to use.
'Like trying to waggle a banana in a bowl of custard'.
It still describes many attempts at shift systems today.

Why sequential?
Sequential was only an FIA regulation issue not an essential transmission need for performance.
Or for that matter the best way to do the job.
Those regs destroyed transmission development almost completely in the early 90s.
Cost me a fortune, just like many others executed by the FIA regulations over the years.

You can change any ratio to any other ratio, no problem.
Now of course with cam barrels nobody knows any different, fixed in metal and the old boys network so to speak.
The BS has even taken it into perfromance road vehicles along with the continued use of the 19th century layshaft concept with all its torque sapping inefficiency.

Of course if you use a planetary gearset you can then change much more efficiently and far faster.
You can also use as many gears as you like, which is why the FIA limited the number of ratios to 7.
That was after I spoke to Garry Anderson at Jordan about the potential for an 8, 15 or 25 speed bevel epicyclic geartrain.
Garry asked me how many gears maximum would a driver be capable of using with his stop gap manual sequential idea.
Jordan didnt have a budget for the new auto/semi auto systems that I discussed with Ferrari at the time.
The Jordan car was built with 7 and the FIA latched onto that to establish the maximum ratios allowed.
Not sure why they mandated sequential, perhaps you could ask them.
I think it had something to do with coasting under braking when shifting from 7 to 3/2 but I may be wrong.

gixxer_drew wrote:If you are after shift times you can get 95% of the way there by designing a nice H pattern shifter. If you are talking about something fitted into an OEM style casing designed for syncho shifts you shorten the fore/aft throw a lot and the side to side throw a little. Get a nice sensor setup for the shift cut, move the whole assembly close to the steering wheel and it will work pretty darn good and close to a sequential in terms of the shift time.

Sequential shifting only works well if it is computor controlled and the shifting components are some form of pre-loaded quick shift mechanism, which the makers call seamless (wrongly).

A sequential shift mechanism fitted for manual shifting is too remote for the driver to 'feel' the shift and is far less efficient or as potentialy fast as an H pattern mechanism used by a 'skilled' driver.
Sequential is based on the 'hairy' biker using his big booted foot to jam a gearshift.
Not much 'feel' doing that is there.
Any gear shift with a layshaft gearbox is only as fast as the time it takes to throw the mechanism from one gear engaged to another gear engaged, against the rotating inertia and speed difference of the geartrains rotating components.
With manual shifting 'feel is most important'.
With electro/hydro/pneumatic, the shift mechanism is practicaly irrelevent.

autogyro wrote:
snip
Of course if you use a planetary gearset you can then change much more efficiently and far faster.
You can also use as many gears as you like, which is why the FIA limited the number of ratios to 7.
That was after I spoke to Garry Anderson at Jordan about the potential for an 8, 15 or 25 speed bevel epicyclic geartrain.
Garry asked me how many gears maximum would a driver be capable of using with his stop gap manual sequential idea.
Jordan didnt have a budget for the new auto/semi auto systems that I discussed with Ferrari at the time.
The Jordan car was built with 7 and the FIA latched onto that to establish the maximum ratios allowed.
Not sure why they mandated sequential, perhaps you could ask them.
I think it had something to do with coasting under braking when shifting from 7 to 3/2 but I may be wrong.

William did a test with a CVT some time in the 90's I think but it was quickly banned, I'm sure partly based on it
wouldn't sound right just going around at a constant ~16000rpm

William did a test with a CVT some time in the 90's I think but it was quickly banned, I'm sure partly based on it
wouldn't sound right just going around at a constant ~16000rpm

David Coulthart tested the car.
I had a meeting with Patrick Head at the time.
The Williams CVT was based on the VanDorne twin cone system.
The problem with all CVT transmissions, Van Dorne, Perbury or others, is that if they are used in a high performance application they all have to use a large input of energy to move and control the variable diameter cones (Van Doorne), the friction discs (Perbury), or other moving components. The toloroidal disc based CVTs like the Perbury also use a friction fluid to maintain traction within its variable mechanism.
Tony Rudd of Lotus and I looked at an electromagnetic fluid for a similar purpose and for remote activation of fluid clutch packs way back in the 80s.
The modern flybrid system is based on the toloroidal system with its high energy control system and friction fluid.
It is ok for low power application but would be unreliable for a high power hybrid system like the 2014 F1 ERH/ERK powertrains.

The CVT concept was nothing to do with my bevel epicyclic geartrain designs of the time.
These were stepped ratio geartrains and would have varied engine speed through fixed ratios exactly the same as the ancient layshaft geartrain.
The difference was in the shift system, which was hydraulic the same as in a conventional automatic gearbox.
However, there was no torque converter and a conventional clutch was used for starts.
My current system uses electro magnetics for shifts and needs no clutch at all.
It is also a complete gearbox and energy recovery/apply unit all in one, with only one bearing in use in 8th gear.

The FIA banned all 'automatic' gearboxes and specified stepped ratio gearboxes with a maximum of 7 gears, it will be 8 gears in 2014.
Downshifts were allowed to be automatic, which is another reason why sequential gained an unfair lead.
It being very difficult to program a control system to activate anything other than an approximate 500 rpm engine speed increase between gears on downshifts using an ancient layshaft gearbox other than in sequence.
Trying to do anything else with the current masking by the high downforce levels, gives insufficient driver feel and control over the shifts and with the regulation controls over traction limiting and diff design, there is no way to prevent rear from breaking away during rear wheel brakeing and energy harvesting, or of effectively controlling tyre wear.
This has totaly lost the old driver skill, manualy changing down under braking and replaced it with a juggling act on push buttons like an arcade game.

I would love to wake up to a new formula that encouraged proper powertrain development instead of model aeroplane technology.

Back to the original question of this old topic

Has anybody seen any animation of how the barrel cam moves the H-pattern rod?

I've seen the kit below in use on a couple of rally stages in a VAG car. It relaces the shifter box and worked on the 02A type gearboxes.

I've seen in being used a couple of offroad stages without problems... however can't say I'm a fan. Too much reliability riscs without any huge gains if you ask me.

http://www.sqsracing.com <- they make all sorts of these conversion shifters.

Hey.

I've been building the H pattern gearbox paddle shifter for the last 2 years. I got paddle shifting without the (expensive) gearbox. It's basically a pneumatic multiposition cylinders controlled by sophisticated hardware and software.

You can see it here:

Nicely executed project.
Unfortunately without the 'seamless' 'quickshift' mechanism in the actual gear train there will be no chance of a perfectly matched engagement.
Also the system forces the gear changes to be sequential which takes away the capability of selecting gears 'across' the gate i.e 5th to 2nd.
Driver feel is also taken away.

In 1976 I built a mini with a similar system with hydro mechanical shifts.
This later surfaced developed in the 1989 Ferrari of Mansel.
I later changed this to a bevel epicyclic stepped ratio automatic clutch flite unit with manual shift electronic actuation.
The hydraulic clutch shift engagements were fool proof.
Leyland wanted to use my system in the Metro 6R4 after 1978 but the company was then on the way out.
Motor sport might have gone in a different direction otherwise.
It is ONLY regulation that prevents epicyclic gear trains from bettering layshaft in racing and the status quo in road car manufacture.

Wow. 1976.. that's before I was born. I cannot imagine doing this 10 years ago, let alone 40 years. Hats down for that project. Honestly.

I think that finely tuned paddle shifting system puts less stress on the gearbox than the manual torturing.
I built a data logging features to my ecu so I can easily tune every stroke and movement to see if the stroke is struggling with the gears or is the force actually preventing the shift. And with pneumatics, I can control the speed of the movement, so I don't move too fast and damage the gearbox forks or something else. Also, the ecu is capable of pressing clutch (pneumatic/hydraulic) with two speeds. One speed for up (quick release) and second for down (slow release). It's easily tunable with air flow controller.