Invoke: Dave the Resurrector.


Something has just occurred to me. It's based upon a sequential conversion mechanism that I once saw for the VW Type 1 transaxle.

Take the drum and ratchet out of a motorcycle gearbox, and fit it into a small enclosure that can be mounted in a remote part of the car, such as behind the diff.

Instead of three shift rods, fit the drum with three shift cables.

Open the Miata transmission. Remove the main shift rod and shifter turret. Attach one of the shift cables directly to the end of each selector shaft.

Just read through this for the first time. I don't have a lot of technical knowledge with trannys, but didn't the 3rd gen MR2 come with an optional 6 speed sequential? Granted it is sitting sideways in the rear of the car, might that be a plausible solution?

FWD trannys have the diff built into them. Basically rendering it useless for a rwd application.

That's a fairly clever bit of thinking Joe... Push-Pull Cables aren't particularly hard nor expensive to have made, we do them for throttles fairly regularly.

See here... PUSH-PULL.COM This is who does all our cables.

One thing I'd point out that I hadn't seen mentioned earlier... the sound "dog" boxes make doesn't have anything to do with the dog vs synchro gear engagement. The sound comes from the gears being straight cut rather than helical.

Thanks. They're pretty common in marine applications, in many OEM FWD shifters, in rear-to-mid transaxle conversions (think VW / Porsche) etc., as well.

I came into the idea while thinking about how to convert a VW transaxle to computer control, using solenoids or linear motors rather than cables, but the principle would be the same.

I've never opened a M/C gearbox (they scare me) but I am assuming that you basically have three selector forks (or rods) riding on the drum, such that each increment of the drum causes one fork to move to the neutral position, and another to the engaged position.

Now, whether or not any M/C box uses a gear pairing that is compatible with an H-pattern is another question, and one I don't know the answer to. IOW, 1-2 on one fork, 3-4 on another, and 5-6 on another.

Eh?

I don't recall seeing anybody refer to the sound of the gearbox, unless you're talking about mazda/nissan's (incorrect) reference to FM's Quaife gearset. While not absolutely straight, that gearset does have a much shallower helix angle, and is thus likely louder. But you're right, in that it obviously retains the stock synchros.

Still, I wouldn't want a dog box on the street even if I could get one. My focus here is entirely about how to convert from an H-pattern shifter to an inline shifter, so as to alleviate the possibility of incorrect shifts.

The reason I brought up the straight (dog) gears is that it would aid in the sequential process. Now I just have to look up wtf Joe is talking about.

Short version:

Straight-cut gears have nothing to do with dogs. A dog box can use helical gears, and a synchro gearbox can use straight gears.

Ah, I thought a dog box utilized straight cut gears, thank yee. But the straight gears is what gives it strength vs. the helical?

And I meant wtf about your VW transaxle. I am afraid I know nothing about VW's.

This was the quote that made me think that people could be confusing the sound of racing boxes as coming from the dog engagement vs the straight cut gears...

Most dog boxes DO use straight gears, the reason they do that though is because they're primarily racing pieces and yes straight cut gears are stronger and don't load the shafts axially like a helical gear does. One other reason not to use dog engagement on a street car... the dog rings get replaced fairly regularly.

Joe, I don't really know anything about bike trannies unfortunately so your guess is as good as mine about how exactly they transfer drum to fork motion inside the box. I wonder if anyone has a broken sportbike lying around...

Tex seems to know his way around a bike...

Oh.

The transaxles used in the air-cooled models were sort of semi-modular, in that the nosecone (where the shifter terminated at) could be removed from the main body of the transmission, exposing the ends of the selector rods themselves. You could then bolt a different system on, to convert to a side-shift mechanism, a hydraulic shifter, etc.

A while back, as a sort of thought experiment, I was toying with the idea of cutting down a VW transaxle (removing the whole bellhousing / clutch section) and using it as a differential for a bike engine. Figured that as a bonus, a person could retain the stock 3'rd gear as well as 4'th, thus obtaining a high / low range selector, as well as retaining the stock reverse gear. To avoid having to run an additional shifter, I thought it'd be interesting to go electric, with linear motors directly operating the forks.


That part is actually pretty straightforward:

http://img40.imagefra.me/img/img40/2...9m_c0347c0.gif

The three shift forks slide side-to-side on the shaft. The tabs (in green) ride in grooves on the drum. As the drum rotates, it causes the forks to move.

Here's another example, in which the three forks are split across two different shafts.

http://www.gadgetjq.com/transmission%20r1150gs.gif

Normally the forks engage the dogs directly. Instead, attach one cable to each fork.

The trick is to find a transmission which uses a gear pairing which is compatible with the H-pattern. IOW, 1-2 must share a fork, 3-4 another, and 5-6 another. The more drum pictures I look at, the less I'm convinced we'll find that combo. It seems that the transition from any gear to any other gear always seems to involve two different forks for some reason.

ok I get it now ;)

Your electronic controlling now has me amused though...

Looking at the above examples it looks like you'd need a few basic things to roll your own...

1) Steal the drum ratchet mechanism from a bike trans.

2) Measure the needed shift fork movement

3) Fab up extensions from shift rod to drum

4) Have a drum CNC cut

5) Fab a housing for the above

Am I missing anything?

6) Profit!

Seriously, though. I fear that by the time you start having new drums cut on a CNC machine and such, the "low budget" nature of this particular approach is lost. At that point, it would likely be just as well to adopt an Ikeya-style design.

Thinking about this further since my last post I'm inclined to agree that either copying the Ikeya style or modifying an existing one to work is probably the easiest and most likely cheapest solution.

Any number of Japanese gearboxes have similar gearshift mechanisms, for instance the Toyota T50 in my Locost is nearly identical in shifter operation to a Miata 5 spd (the location of the miata in a turret is different, the actuation method almost identical). That said it seems like the challenges in adapting an Ikeya would be:
1) Find a box with similar actuation method and length/width of shift gates at the shift rod.
2) Fab an adapter to attach the mechanism to the box.

Without having seen an blown apart Ikeya in person I'm not sure how easily modded for varying shift gate throws they might be.

Well, they offer one for the FD.

Any of you triangle-heads know how similar the FD tranny's gear selector is to the Miata's?

I've considered this a bit myself and come to the following conclusions:

1.) A Bus (Hayabusa) makes more torque than a 1.6L Miata engine almost everywhere. With the 200+lb weight savings of going from the Miata iron block and diesel transmission to a Bus motor, I think you'd have a fantastic street car. The Quaife inline reverser box is $1600, but Champion Motorcars (a defunct Locost kit builder) used a much cheaper, much simpler box. I'm sure the manufacturer could be found. The transmission on the Bus motor would have a hard time with the abuse, and you'd burn up clutches pretty fast, but they're very easy to swap out. Adapter flanges already exist to attach universal/ CV joints onto the Bus output shaft. With an aftermarket gearset, tool steel shift rods, billet shift forks, and a super bad ass clutch setup (solid hub, APE plates, super stiff springs) you could probably get by for a LONG time. With a really stripped track day/ street/ F Mod (Street Mod?) autocross toy you could possibly keep it together forever.

Now that I write all that down... I'll be needing a Miata roller.


2.) The aftermarket Harley transmissions are <$3k, and almost all of them can be had in right side drive (correct input/ output shaft arrangement). And if they can make a ~1200lb Harley Desk Glide reliable with ~145hp, you could keep one alive behind a mild ~100hp 2000 lb Miata, I'm sure. If you had a clutch basket and hub machined to use dry Ducati clutch plates, and the Revtech/ whatever input shaft, and a small bell housing to mate the trans to the back of the Miata block, you wouldn't have to run the clutch in an oil bath, and you could put as many plates in it as you wanted, meaning you could keep the clutch alive forever with all that thermal mass and surface area. That does kind of limit you to a really mild motor or a really light Miata though, unless the trans manufacturers think they're product is a tremendous amount tougher than I do.

3.) If I were going to do any silly bike engine swaps into a Miata other than a Hayabusa, I'd do a Honda Goldwing. Think about it: 1800ccs, enough torque to clutch up wheelie a 900lb bike in 3rd gear, simple, symmetrical, almost centered shaft drive output shaft, reverse... They sound SICK uncorked, they're Honda indestructible, they're light(ish). Just a thought.

I don't personally think a Bike engine Miata would work particularly well anywhere other than on track. The bike engine Locost guys are pretty abusive to the clutches already, and Miatas weigh at least 50% to 100% more than a Locost, depending on how stripped the Miata is. I can't see a sportbike trans/clutch living long doing street duty stop and go lugging around 4-6 times the weight it was designed for.

That said I agree with your Leadwing idea... that flat 6 in a light mid-engine chassis would be sweet.

Boxster did it...

If clutches were the only worry, I'd be all over this. The clutch would face forward on a longitudinally mounted engine, so you would just pull the cover (8-10 6mm allens), pull the 6 pressure plate bolts (6 8mm allens), pull the old plates, stuff in the new plates, reassemble, top off the oil, recommence to boogy. I'd be more afraid of what happens to the shift forks/ drum after a weekend of track days. A careful driver would never hurt them, but KWS's rider last year (Ben Thompson) could smash a set of shift forks to pieces in an afternoon of practice and qualifying. Every missed shift or careless down shift wears on the pads on the forks, and every over zealous shift in either direction wears the drum lugs off the forks, and beats the life out of the drums...