Offset shift linkage--how?

Unfortunately, I don't think so. There are springs and ball detents to overcome in the tranny internals and it's a push/pull for first second, then a rotation to move shift gates, then a push/pull for 3/4, then a rotation to move into reverse gate, then another pull (but who needs reverse, right?) The ball detents are at each gear locking the shift fork into position as well as the syncro hubs into the gear.

However, if you can think of a way to do cables, I'm open to it.

Hmmm. wonder if this might help alleviate some of the binding that I'm experiencing at the end bushing...

http://bearingsdirect.com/store/index.p … amp;p=3812

Hey Spank, the ASP used craftsman 3/8" u-joints and heims-joints about 4" either side of each joint to get the shift from the rear to the driver. It was convoluted, but it worked.

I was thinking of doing a mod from a boat-steering or shifting cable - one cable to do the twist and the other to do the in-out.

I think in the Apollo's case you could do a U-joint and rod for the in-out and do the rotate with a  boat steering or shift cable - the advantage of the cable is that you can mount the arm/lever/cam on the in/out shaft and mount the cable-housing holder 4-6" away and not have to worry about the sliding motion transfer and alignment...

--as in, put a lever on the output of the trans, mount a cable at 90 degrees to car centerline - run the rod pretty much like you have it and then either put the cable to the bottom of the shifter ball or make a lever just like you did on the other end (the latter should be easier with the std Rod-x box than the former)

Luck, wish I was there to cobble that one together w/you.

Another option that I have only thought through but not sketched out yet (I have no CAD software - so bar napkins are my usual level of sophistication for tech drawings) is to put the stock shifter on the trans wherever it goes. Then build a median 'shifter relay' directly forward (i.e. straight forward from the 3-4 shift line) that is a rod with the same height as the stock shifter rod and the same diameter - the pivot for this forward displaced 'relay' is in the same exact horizontal plane as the stock shifter. Then build a third device, we'll call it a 'proxy shifter' with the pivot in the same plane as the other two pivots, and the same rod length/diamter as the other two, this one located 90 degrees off of the axis of the stock shifter to the 'shifter relay'.

Now you have three identical linkages (not hooked to each other yet, that comes next) that all have the same height/diameter lever and pivot on the same horizontal plane. Stock one is where ever it has to be, relay is directly in front of the stock one, and proxy shifter is directly to the side of the relay (all assuming your trans is too far back but basically any displacement can be addressed by two right angles as long as you can keep everything in the same plane).

Now go find a heim joint that fits the rods (as far up as possible to allow for all the errors in fabrication, and you actually want to buy all the rods / modify them to fit a size you can get heim joints for in planning - but you knew that) and buy four of them. You'll cut one of them and weld part of it to make a right angle fitting for the relay rod with it (you end up with a stock heim joint, with an extra female thread portion 90 degress offset, and in the same horizontal plane as the original one). Then you'll mount that to the relay rod at height x up from pivot, mount one of the unmodified ones to the stock shifter x up from pivot, then the last unmodified one to the shifter proxy at height x up from pivot.  Then use tie rods or something similarly crappy to tie them together (start with the lengths between the pivots, and I imagine you'll end up fiddling with it to get it to make it shift). You'll need bracing between the two right angle tie rods - probably a lot of it, and reverse may not be manageable with this goofy linkage.

Caveat: I haven't built this yet, so if it doesn't work for some glaringly obvious (or otherwise)reasons, you should have known better than to listen to some guy on the internet.  If it does work, I'd appreciate pictures of it completed...

Good luck.

If you can't find that, there should also be at least 1 SHO in the junkyard. The 1st gens have a pretty versatile cable setup.

I think most FWD cars have a cable shifter.  They use two cables, one for fore-aft movement, another for left-right movement.  Probably cheaper and easier to find one from a common, readily available car like a Focus or something vs. an MR2.  (unless you need the potentially longer cables from the MR2.... but even FWD cables are likely to be 5-6 feet long)

You could try adding a second bushing on each of your two short sections.

Your problem as you sort of talked about is the linkage is twisting and binding in the 1 bushing. When you push forward (or back) the motion is transmitted along the shaft (straight line), and then to the universal. But the universal, operating at an angle converts the linear F-B motion into a motion that is now 27 degrees off axis. When it does that there is a corresponding reaction in the opposite direction. That reactionary force rotates the first rod counter clockwise. That 1 binds up the rod in the bushing, and 2 reduces the amount of force getting transmitted to the next universal, where it happens again. So with all the loss of force, you get no movement on the other end. All the force applied gets turned into rotational forces in the universals, which just flex because that's what they do, and the result is no movement at the tranny.

two bushings on each rod will limit the amount of rotation to near zero and should allow more force to be translated. You'll still have some slop in the universals which will dissipate some of the force, but it should work.

A better linkage would use one rod for the front to back with pivot linkages on each end (essentially what the shifter is, pivot in the middle). then you need a second bell crank to transmit the side to side.

am i missing something? why cant it be a solid piece of steel rod (ya know, sans u-joints)?

all the good Hondas are link shift, and they are a single piece of metal.

cant you just bend some rebar?

That diagonal piece doesn't need to push in and out-- it's only supposed to (in my eyes) say in the same plane and move fore/aft-- so the two sections of lengthwise rods move in and out and the diagonal just moves front to back. Someone on the team pointed out to me in order for it to work the two forward facing rods need to be in the same alignment: traveling in the same trajectory just split into to different parallel planes. I think the forward is angled more than the rearward one which is another reason there is binding.

I'm not giving up on this design quite yet. Will play a bit and see what happens.

Appreciate all of the advice everyone and will go the cable route if this one ends up sailing out into the middle of the street boomerang style about 3am some time in the near future...

Well, you could make the MR2 cables go forward instead of backward very easy. Just a matter of removing a couple bolts and flipping the cables. But going with a FWD Toyota would probably be a better idea cause then you don't have to do even that.

Extend the stock shift rod straight back under the seat. Build a second shifter offset to the right, where you want it. Mount it on a gimball and use heim joints to transfer the movements.

A picture would be worth so many words right now, but I'd have to do some drawing.