41 replies to this topic

[Supercar]

I would like to convert my H-pattern shifter to sequential. That is at the shifter, not inside the gearbox. I want to design a mechanism that would transform the fore-aft strokes of a sequential shifter into the same cable motions that an H-shifter does.

Why would I want to do this? Because I think it will be faster, fewer mis-shifts and blown engines, and it will be cool too!

Has anyone heard about that being done before? I think there was one or two companies that actually used to make those but I cannot find them anywhere anymore.

Philip

[Canuck]

I'm certainly no fountain of knowledge here, but I pondered the same thing when looking at Ultima's GTR component car. I'm not very well versed in the shifting mechanisms in automotive manual transmissions, but I know my motorcycle stuff inside and out, and they're all sequential. I think it would be a fairly straight-forward operation to get a sequential action from your shifter. I don't know that it would be a nice, tight, short-throw setup however or even something you'd want to use all the time.

[Greg Locock]

If you are doing it mechanically I think you'll end up with a tracked drum that clicks between each gear. One track on the drum is the lateral position, one is the longitudinal position (and one more for the clutch).

Nightmare factor of about 7, I reckon.

I don't see how to regulate the shift speed. Perhaps that doesn't matter.

But as you say, cool.

[Supercar]

Originally posted by Greg Locock
If you are doing it mechanically I think you'll end up with a tracked drum that clicks between each gear. One track on the drum is the lateral position, one is the longitudinal position (and one more for the clutch).

I am thinking either a drum or a slotted plate, whatever is simpler and more compact, as long as it shifts well.

Probably no clutch, else I would have to move that shifter into the footwell. I do not think I have enought strength in my right hand to depress the clutch every time I shift.

[Greg Locock]

How about vac assist on the clutch? Howstuffworks has a good article, with a picture of the selector drum. I'm willing to scale the nightmare factor back to 6 having seen that article!

[Wolf]

I do seem to remember Lotus F1 cars at some time had some kind of shaft with grooves for engafing selector forks on top of gearbox. That seems to be simple enough (does anyone have a pic or cutaway, BTW?), and although not exactly what You're looking for, but it's something I pondered.

In that case fore-aft motion of shift lever needs to be converted only to rotation, and this is the mechanism I came up with (I modified an old mechanism for one-way conversion to bi-directional) :

[Supercar]

Originally posted by Wolf
... In that case fore-aft motion of shift lever needs to be converted only to rotation, and this is the mechanism I came up with

WOW! That must be reaching like 9.5 on the Greg's scale! After you've done all this I am now tempted to take my motorcycles apart and see how it is done there. I remember it was less complex, but I forgot how exactly, it was many years ago that I took a motorcycle gearbox apart.

Howstuffworks has some on sequential gearboxes, but they do not have the racheting mechanism detailed.
http://auto.howstuff...al-gearbox1.htm

[Supercar]

Originally posted by Greg Locock
How about vac assist on the clutch?

It's already there. In order to make is softer I would have to install a clutch vacuum booster as large as the brake booster. I have room on the other side of the engine bay, but that is not the plan. The plan is to have a dogbox and use no clutch at all.

[Engineguy]

A conversion external to the gearbox won't be like a motorcycle or racing gearbox internal treatment. To duplicate the cable travels of a twin cable H pattern system you'll need two cams or drums on a common shaft... one for the fore-aft cable and one for the lateral cable. Don't forget that the fore-aft (blue) cable must pause (dwell in the neutral position) while the lateral (red) cable moves during the 2-3 shift and the 4-5 shift. You'll need a lockout to prevent the camshaft from rotating from the neutral position to the reverse position.

[Canuck]

Forgive the rather crude elements presented here - modeling is not my strong suit.
This is strictly by memory - I do have the ratchet mechanism kicking around in a destroyed trans somewhere - when I find it I'll take some pictures but this should provide you with a reasonable idea.



The pins on the end of the drum are engaged by the slotted lever which is free to rotate on it's pinned axis. A spring between the blue and red elements keeps pressure on the pins so it doesn't flop about. The lower, green lever and the blue lever rotate on the same axis however the blue lever is not fixed to the green lever's shaft - it is free to rotate within the limits imposed by the (not shown) spring that ties the blue and green elements together such that a rotational input on the green shaft will cause the green element to rotate until it takes up the slack on the spring between it's leading edge and the blue elements trailing edge. This turn pushes or pulls the red lever which engages the pins and rotates the drum. The spring pressure on the red lever is such that when the primary (blue/green) spring pressure returns the lever to it's resting position, the red lever can lift and travel over the next pin to position itself for pulling the next gear. I know somewhere in that mess should be an adjustment pin that centers the drum-pins within the lever, and provides for the centering action.

Or - I could have it totally cocked up...

[Canuck]

This image is taken from one of Indian's parts manuals. They are notoriously inaccurate but again the idea is there.

[12.9:1]

I know I've seen kits based on electro/hydraulic-or pneumatics, thought I can't find one now, in my experience such hybrid systems may seem to technical, they offer broad (and relatively simple) adaptability - adjustability. We are talking about a one off here ?

A quick Google search and a hint at the - it's either exactly right or it's just not working! mechanical system.

http://www.cartronic.../seqshift_e.pdf

[Engineguy]

Originally posted by 12.9:1
...and a hint at the - it's either exactly right or it's just not working! mechanical system.

In this type of conversion, some slop is a good thing... there are internal detents in the transmission that precisely hold it in gear... you've got to let them do their job and not fight them... your mechanism needs to simulate removing your hand from the (H pattern) shifter.

[AS110]

Back in the '60's there was a conversion to make a three on the tree into a straight line floor shift.It was really a vertical H pattern,but spring loaded and you could shift just pulling back or pushing forward.How would your car work as a 3 speed...?

[Supercar]

Originally posted by 12.9:1
I know I've seen kits based on electro/hydraulic-or pneumatics, thought I can't find one now, in my experience such hybrid systems may seem to technical, they offer broad (and relatively simple) adaptability - adjustability. We are talking about a one off here ?

A quick Google search and a hint at the - it's either exactly right or it's just not working! mechanical system.

http://www.cartronic.../seqshift_e.pdf

This is exactly what I want! Ror € 2.400,00 I should be able to make my own. One off or a few, depending on if it works and if it can be manufactured easily.

http://www.cartronic...e/por_sequ.html

Philip

[Supercar]

Originally posted by AS110
Back in the '60's there was a conversion to make a three on the tree into a straight line floor shift.It was really a vertical H pattern,but spring loaded and you could shift just pulling back or pushing forward.How would your car work as a 3 speed...?

That's what I heard too - thee speed! Yes, I usually use only 3,4 and 5th gear on my 6-speed when I am at at track.

[Supercar]

I think I now remember the shifter mechanism on my '83 CZ-125 516. It used two oppositely sprung blocks with slanted tips inside a drum that had internal teeth. Somehow these blocks were able to rotate the drum and jump over those teeth. There were only like 5 parts in that whole mechanism. The drum totated the grooved shift shaft that moved the forks.

Oh, I think I remember. There also was a stationary segment there that pushed one the blocks in so that it could jump the drum's teeth. Sorry, CAD crashes my laptop, otherwise I would have sent a drawing.

My even older CZ-125 511 and CZ-250/500 514/514 had a grooved plate that moved the shift forks. I am thinking a plate is easier to make and customize than a grooved shaft. But both should work.

Thanks for all the comments and suggestions!

[soubriquet]

Originally posted by AS110
Back in the '60's there was a conversion to make a three on the tree into a straight line floor shift.It was really a vertical H pattern,but spring loaded and you could shift just pulling back or pushing forward.How would your car work as a 3 speed...?

I fitted one of these to a Holden 3-speed. My very hazy recollection is it was able to work in a straight line because there were two external links to the shift forks, so no across the gate motion was required. Simply the correct sequence of engagement and disengagement.

[AS110]

And they were crash 1st,so the 2nd - 1st shift was never attempted in motion....sober at least.Here in NZ they were mainly used on Holdens too....B/C,or Berry & Chung I think.

[Stian1979]

I have taken apart some gearboxes from both motorbikes and car's and to be honest I can't see anny way to make a H-phatern into a sequensial withour making a complex mecanism.
I think it would be easier to control it by hydraulic sylinders and electric selenoyd valves so you could have f1style gearing on the stearing weel.

[scooperman]

I was searching the web to see if anyone had done a sequential shift for the T5. Still have not found anyone making kits, but I did see this picture




It is on the History page at this site...http://www.ofiengineering.se

[scarbs]

A french company 'Selexy' used to make a remote sequential selector mechanism, this was aimed as a retro fit to H gate gearboxes. IIRC it was RaceCar Engineering magazine that ran the story. effectively it was the motorcycle style indexed drum that operated forks to rotate the conventional shift linkage.
they don;t appear to be around any more but finding the RCE article may lead you to your solution.

[slucas]

Have a look here; www/ikeya-f.co.jp/en/index.html

(I can't remember how to highlight it so you can just click on it, sorry)

[SCO]

whilst we are on clever selector mechanisms does anybody know how the wilson pre-selector boxes work?

[cheapracer]

Ok, Ok I know this one!!

I was in a workshop last year and noticed a very slick, simple and compact mechanism on the side of a Yankee V8 gearbox (BB 396 Chev box from memory) that looked very familiar to me as a 1970's Maico motocross selector unit - a local Aussie makes the unit (and has a patent I was told) to convert external GM/Ford "H" patterns into a true sequential shifter.

If someone out there wants to know more I can give the contact details of the friend who introduced me to the Guy who owns the workshop who knows all about it.

[Fat Boy]

Just out of curiousity, why do you want to convert away from the "H"? There are good reasons to keep it an "H", and there's really no speed to be gained in making it an inline shifter.

It's done at the highest levels because there are enough gears to make finding each one difficult (i.e. with a 7-speed). In the lower formula, it's used because it's in vogue. For 5 or 6 gears, an H pattern works quite well, and the ability to skip gears can be very nice.

[cheapracer]

Originally posted by Fat Boy
Just out of curiousity, why do you want to convert away from the "H"? There are good reasons to keep it an "H", and there's really no speed to be gained in making it an inline shifter.

It's done at the highest levels because there are enough gears to make finding each one difficult (i.e. with a 7-speed). In the lower formula, it's used because it's in vogue. For 5 or 6 gears, an H pattern works quite well, and the ability to skip gears can be very nice.

1 for - is for as you say

1 against - is time - sequential can/should be quicker than the (typical) 2nd to 3rd and 4th to 5th cross gate shift.

About 90% of race drivers I've ridden with couldn't shift with a clutch properly anyway and are suprised when I show them how I do it and how much time they save. Oh, another "I'm a teacher and I didn't know it", wow, I could have made Senna into a Superguy

[Fat Boy]

Originally posted by cheapracer


1 for - is for as you say

1 against - is time - sequential can/should be quicker than the (typical) 2nd to 3rd and 4th to 5th cross gate shift.

About 90% of race drivers I've ridden with couldn't shift with a clutch properly anyway and are suprised when I show them how I do it and how much time they save. Oh, another "I'm a teacher and I didn't know it", wow, I could have made Senna into a Superguy

The time spent crossing gates on a really tight H pattern is pretty much the same as it takes on a sequential, as long as they're both manual (no hydraulics or air shift). You can run an ignition cut on an H-pattern and get down to 150ms or so shift time. You'll maybe get it 20-30ms faster on a sequential, if you're really on your game. That's about as good as it gets. There's definitely time in shifting properly, no question, but between the 2 types of boxes the slicing is pretty thin as long as they're both set up properly, though. NASCAR gearboxes on a road course are a good example of getting it _done_ with an H pattern.

If there is not a shift cut, then the difference between an H and sequential is nill. The slow bit now (or at least should be) is your foot. Keep in mind, we aren't talking about a ground up sequential box, we're talking about an inline mechanical system to convert an H to a sequential. More parts to carry around, more parts to maintain, more parts to break.

As far as shifting properly with a clutch, aren't those two things mutually exclusive on a racecar box?

[J. Edlund]

Couldn't it be possible to control the three forks with pneumatic, hydraulic or electrical actuators instead, perhaps using one actuator per fork? I seem to recall that something like that was used on the first halfautmatic gearboxes used in F1.

[phantom II]

http://www.mastershift.com/ Try this site. I spoke to the owner up in Carlisle last week at the biggest Corvette show in the world, who says a auto clutch mechanism is in the works. A friend of mine has the system on his C6 Z06. The speed of the shifts is adjustable to brutal. Using the clutch is optional but advisable. The shifts are sequential.

Originally posted by scooperman
I was searching the web to see if anyone had done a sequential shift for the T5. Still have not found anyone making kits, but I did see this picture

[AdamLarnachJr]

You will spend less money (and time) on a new sequential gearbox than you will converting an H pattern and having it function properly.

If you are missing that many shifts with an H pattern, either you're synchros are shot or you aren't shifting properly. Secondly, shift speed is also determined by your rotating mass. If your looking for something close to competition speeds, you might as well invest in a proper sequential gearbox... and even then those require constant attention and calibration.

If you want just a cool toy to mess with, build a pneumatic system and write some software to control your solenoids. You can use rams to actuate the clutch, shifter, and throttle.

[cheapracer]

Originally posted by AdamLarnachJr
You will spend less money (and time) on a new sequential gearbox than you will converting an H pattern and having it function properly.

.

Just not true, I used (box on the floor) and watched the action of (mentioned in my post) the shifter myself - for those who know what a Maico rotating shift/selector plate looks like it's just pure and foolproof simplicity, I will try to find a pic.

...
and by the way, a crappy response.

Some of us raised 3 kids, put them through good schools, they had new shoes (if you don't know what that means then we aren't on the same level, some other Fathers here will understand that clearly) had Mum stay at home and we worked one job - and we went motor racing as well by engineering our own low cost solutions to things - we don't just run out and "spend less money ( ) on a new whatever".

Anyway the racing boys "go and buy a new toy mentality" makes me laugh because in a Race car shop it's $100 bucks, in a Rally shop it's $45 and put a dollar on the fridge and get your own beer out of the fridge, in a Speedway or Offroad shop it's $28 and you'll get a free beer on a Friday night - all for the same part

[cheapracer]

http://www.mxbikes.c...last=1211263737

Not a great picture but you can see in the 6th pic down a Maico rotating shift plate. It rotates by means of a forked ratchet. 2 grooves in the plate at different points either go in towards the middle or outwards forcing the following pins of the selector forks (2 on the single shaft under the shift plate) to follow moving the gears across on their respective shafts (left and center shaft).

The one I saw simply operates in the same principle on the 2 typical external gearbox shift levers and is faster in theory because it can control the 2 levers independently in the one action on the previous 'H' cross gate shift.

[Canuck]

Harley went away from the plate-type sequential mechanism to the rotating drum in the early 80s I believe. Harley being the paragon cutting edge technology that they are, don't often abandon great technology to adopt a lesser one (they usually just stick with the lesser, foregoing change at any cost).

In this parts catalogue, you get a decent idea of how it worked. You can see the fork blocks that carry the forks have little bushings that ride in the shift plate. Directly above the shift plate is the ratchet mechanism, and above and to the left is the actual shifter arm.

[minigeff]

am i right in thinking aston's had a normal gearbox which was operated through a sequential system in the first model vanquish?

i know the gearbox in that was shite, and thought that was the problem.

i'd just stick with the normal gearbox, if it aint broke etc etc.

[AdamLarnachJr]

Originally posted by cheapracer


Just not true, I used (box on the floor) and watched the action of (mentioned in my post) the shifter myself - for those who know what a Maico rotating shift/selector plate looks like it's just pure and foolproof simplicity, I will try to find a pic.

...
and by the way, a crappy response.

Some of us raised 3 kids, put them through good schools, they had new shoes (if you don't know what that means then we aren't on the same level, some other Fathers here will understand that clearly) had Mum stay at home and we worked one job - and we went motor racing as well by engineering our own low cost solutions to things - we don't just run out and "spend less money ( ) on a new whatever".

Anyway the racing boys "go and buy a new toy mentality" makes me laugh because in a Race car shop it's $100 bucks, in a Rally shop it's $45 and put a dollar on the fridge and get your own beer out of the fridge, in a Speedway or Offroad shop it's $28 and you'll get a free beer on a Friday night - all for the same part

Cheapracer: Yes, you can get them to work, I've seen a few graduate projects do just that through hydraulic and pneumatic means, but I have yet to see one shift quicker than any person could, and all of them integrated in some way with the engine.

Let's define proper operation here. The idea behind a sequential gearbox is for quicker shifts and packaging... basically performance. For the whole system to work properly you will need a way of actuating the clutch, cutting throttle, fuel and/or ignition, reading rpm and matching rpm with any given gear, and obviously actuating the gear.

Now if your in it just for sh!ts and giggles and don't care about performance, yes you'd be correct that any garage operation jury-rigged setup will work fine, but just to shift the gears and nothing more.

Now you say my initial response is "crappy", yet it appears you do not understand how complicated a proper "sequential" gearbox system actually works. The mechanical bits are a very small part of the equation. If you'd like I could elaborate as I've spent a fair amount of time this year calibrating Ricardo's along with the Megaline shift systems (used on the Audi R10) and their related systems. It's all pretty straightforward, but to put it into perspective one bad pots sensor cost us our finish at Le Mans this year. As I said, a proper sequential system is just that "a system". The original poster wanted a "faster, fewer mis-shifts". A bag of bolts and sheet metal will not do this.

[Tony Matthews]

I don't know if this is of any interest - it's the F2000 F1 gearbox, the barrel shifter based on a Yamaha design. There is a vane-type actuator within the barrel and the 'can' on the front is the rotary sensor.

[robroy]

Originally posted by Tony Matthews
I don't know if this is of any interest - it's the F2000 F1 gearbox, the barrel shifter based on a Yamaha design. There is a vane-type actuator within the barrel and the 'can' on the front is the rotary sensor. [/B]

Wow! Love the welds and carbon effect finish.

[Christiaan]

If I am correct ur trying to emulate the selespeed from Alfa Romeo. I'll tell you what

THAT ALFA SELESPEED SYTEM WAS DESIGNED BY SATAN!

It has an actuator driven by a pump shifting what is essentially a manual gearbox. I have been told that if one wants to go sequential, then full automatic is the way to go

[britishtrident]

As many have already pointed out a Heath-Robinson system is a pointless exercise, it has to be a blank paper design to realise any advantage.

It never ceases to amaze me how big a part fashion plays in racing car design ----- if you have an engine that goes all to 15,000 rpm plus yes a properly designed sequential shift transmision is a good idea but for mere mortals the H pattern has been working pretty well since before WW1. For a time it face real competition as a luxury sports car and racing gearbox from the Wilson pre-selector gearbox.

The H pattern has real advantages -- eg being able to go straight from 5th to 2nd gear, the only things that have given the H pattern a bad name is the poor linkages of most FWD cars and the spring bias towards the 3rd-4th speed plane that became common on tin tops at the end of the 1970s.

If OP ever gets a chance drive one of the better H pattern 4 speed RWD manual gearboxes from the 1960's/1970's eg most of the Ford boxes (particularly the "2000e") or the Rootes/Chrysler UK boxes he will realise the adavantages of the H pattern.

[bobqzzi]

http://www.sqsracing.com/ makes what you are looking for for several cable shift boxes

[MaxRPM]

Hi to everyone, first post .
I noticed that Lusomotors http://www.lusomotors.com/parts.html do a sequential conversion for the Ford Mk9 box for 599 euro.
Looking at the picture of the bolt-on device, I'd guess it contains a stubby change barrel operated by the lever that drives followers in the conventional 'H' pattern. A fairly compact solution if this is what you are after.