Here's the rub; currently I have the ticking time bomb four speed auto with a standard shifter, as well as one arm. Remember that last bit, it's important.

I would love a manual, but they don't come with left arms, so dont try and sell me on a manual conversion. I'm kicking around the following idea - Foot shifter, similar to a motorcycle, for an auto.

In principal its 100% do-able. A bit of measuring/research/sketching and a whole lot of time in Inventor creating working 3D model. This will produce either a ratcheting shifter to link to the standard auto-shifter, or, an extension from of the tippy shifter arrangement. Both allowing gears to be selected using my foot. Make sense?

Anyway, I'm looking at the tippy shifter conversion to make life a little less complicated here. I've searched the bejeesus out of this on the forums - tippy shifter conversion - but I still dont have a clear and concise shopping list. Apart from the obvious (shifter) what else do I need and exactly what do I need so I can roll up at the local wreckers with a spanner and avoid grabbing the wrong parts?

There's talk about a "BEM" and reprogramming such and such... maybe something about keys and immobilisers. It sounds complicated but, once again, do-able.

The upshot of me doing the tippy conversion now is that when my transmission dies (again) I'm halfway there to fitting a replacement five speed auto.

If the outlay is too much I'll revert to making my own ratchet/rail arrangement to adapt to a standard shifter.

Bear in mind, this is purely a "want of control" mod. Not a "go fast mod", or a racer mod. This will just allow me shift gears without taking my one and only hand off the wheel. I'm aware a tippy shifter would limit my gear selection to forward gears - I'm happy with that. Had planned to apply the same limitation to the standard shifter if I had to go that way - retaining the safety feature of not selecting reverse at 60kmph.

A thousand thanks in anticipation. Will reward with good karma, photoshop work or in the creation of any 3D models/2D drawings from Inventor or Solidworks. AND, I'll post pics and plans of anything I create in making this work. Just incase anybody else wants to do other stuff with their other hand while driving.

TLDR: Please supply a CLEAR and CONCISE parts list required to convert a STANDARD auto transmission shifter to a TRIP-TRONIC (TIPPY) shifter. For extra points, explain how to put it all back together again once I have said parts. Double point bonus; step-by-step "tippy conversion for dummies" type instructions.

Anyway, I'm looking at the tippy shifter conversion to make life a little less complicated here. I've searched the bejeesus out of this on the forums - tippy shifter conversion - but I still dont have a clear and concise shopping list. Apart from the obvious (shifter) what else do I need and exactly what do I need so I can roll up at the local wreckers with a spanner and avoid grabbing the wrong parts?

Some fill-in information:

1) the tippy shifter is just a collection of microswitches, along with a mechanism to activate them. You should be able to pick up the black version relatively inexpensively ($50?) which would give you the switches along with the mechanism to study...

2) the tiptronic function is managed entirely by the ECU, so you require a tippy enabled ECU. All 5 speed auto boxes were shipped in cars with TCL except for the AWDs. If your car doesn't have TCL, you're in for extra pain. I believe the TCL can be disabled, but I've seen a post by MadMagna where he stated he'd found that without parts of the TCL harness it still didn't work the way it should. In the 4 speed, I understand the auto box is the same for both tippy and std shift; just the ECU and shifter harness are different provided the car doesn't have TCL (see comment re 5 speed...).

3) all the wiring diagrams are in the downloadable workshop manual.

If it were me in your situation:
- look for a working 4 speed box and find someone to do it up with the revised wave spring;
- get the necessary harness bits and ECU for a TJ 4speed non-TCL tippy;
- put the tippy shifter in the car and strap your extra switches in parallel with the shifter.

This approach avoids the hassle of sorting the TCL mess out, and a rebuilt box should last very well. No 5 speed though :(

Northside Transmissions
Shed 3, 87 Connaught St.
Sandgate, Qld. 4017.
Ph - 3869 0184
Mick Hanley

Good bunch of blokes ;)

You can mention the my car if you want, they will probably remember. I told them I would send any club members their way :)

A good auto transmission mechanic :D

Just to clarify, is the tippy auto 100% electronic driven, or only the tiptronic side? ie., is the "standard" side (P,R,N,D) also just controlling switches, or is that side a mechanical link? I guess if the latter it's a pain, but theoretically you'd only be changing those gears in situations where it's OK to take your hand off the wheel.
Could track down a steering-wheel audio control unit and associated slipring/wiring, should give you enough wires to control the tippy from the steering wheel, or maybe even modify a cruise control stick instead, that's probably the most convenient way and positioning. Not sure if all later steering wheels had all of the sliprings in place and just need additional wiring to implement a cruise or audio link?

As for the mechanical link, might want to include a safety-interlock button on the wheel, maybe driving a solenoid, so you can't use the foot pedal to change P-R-N-D unless you are holding the button. Save an accidental kick from drive to neutral (or much worse, straight through to reverse!)

- look for a working 4 speed box and find someone to do it up with the revised wave spring;
- get the necessary harness bits and ECU for a TJ 4speed non-TCL tippy;
- put the tippy shifter in the car and strap your extra switches in parallel with the shifter.

Excellent. I think I may have seen a series one TJ - 4spd? tippy - at the wreckers. Should have all the parts I need...

As for switches, I'm going to do this old school - mechanically. Linkages that are an extension, not a replacement, of the tippy shifter. As in, shifting using my foot would only be available once the shifter was pushed over into the "up/down" part of the shifter. Saves me building a ratchet.

I had some rough sketches of using the original shifter, but it meant pulling the actual shifter apart and reshaping the "teeth" that limited shifting through 1,2,3,4 without depressing the button. Do-able. Might still do it as a proof of concept.

As for biodegradable wave washers and four or five speeds; the 3 piece washer may not be in the transmission I have now? Its an earlier auto (has external spin on filter) and I gather it was the earlier ones that still had the single piece washer. Having four or five gears isn't such a big deal, I'm not speed racer.

Appreciate the help so far.

Just to clarify, is the tippy auto 100% electronic driven, or only the tiptronic side? ie., is the "standard" side (P,R,N,D) also just controlling switches, or is that side a mechanical link

I believe that both tippy and standard run a cable to an inhibitor switch. The tippy just has the circuitry to select from the four forward gears once the inhibitor switch is in "D".


Could track down a steering-wheel audio control unit

Possibly... Electricity and circuitry is white man magic though and I have a habit of letting the magic smoke out of devices. A mechanical linkage is easily built/modified and repaired/replaced.


As for the mechanical link, might want to include a safety-interlock button on the wheel

My plan was to retain the existing "safety" of having to depress the button between P,R,N,D. Just reshape the teeth on L,2,3,D and limiting the range of any ratchet I made to those four gears. That's if I head down the way of using the standard shifter.

The tippy selector has two micro switches to tell the TCU to shift down or up. It a standard transmission, unsure why you want a mechanical set up. When I get home from work I will upload the tippy TCU wiring diagram. The transmission on magna has mechanical setup to select PRND but when you select D the TCU take over and sends electrical pluses to the solenoid to select the gears. The tippy let's you down shift or up shift a gear. So if you don't know about electrical system let an auto electrician to do work for you :)

I believe that both tippy and standard run a cable to an inhibitor switch. The tippy just has the circuitry to select from the four forward gears once the inhibitor switch is in "D".

They are both electronic in that all operation, including the inhibitor, is by microswitches.


Possibly... Electricity and circuitry is white man magic though and I have a habit of letting the magic smoke out of devices. A mechanical linkage is easily built/modified and repaired/replaced.

Still worth considering setting up your foot controller to directly operate a pair of microswitches wired in parallel with the +/- microswitches. Would save a lot of fiddly fabrication and adjustment, as the it would be easier to control the correct positioning of the microswitches for reliable operation.


My plan was to retain the existing "safety" of having to depress the button between P,R,N,D. Just reshape the teeth on L,2,3,D and limiting the range of any ratchet I made to those four gears. That's if I head down the way of using the standard shifter.

I know you'd prefer to avoid electronic solutions, but another idea is to repurpose the cruise control switch (if the car is so fitted; add if not) when cruise is off. Would require a moderate amount of electronic finesse though.

BTW, do be sure you know your position with respect to any rules regarding control modifications to a vehicle - I gather hand controller systems for paraplegic drivers have certain approval requirements, and you wouldn't want to invalidate your insurance...

They are both electronic in that all operation, including the inhibitor, is by microswitches.

So "Park" is engaged electronically? Well, I'll be damned.


foot controller to directly operate a pair of microswitches wired in parallel

Okay, that's starting to sound a little more appealing. Have wiring diagram, can solder...
I would need the shifter, a suitable ECU - from a series one TJ with a tippy shifter - and a wiring loom? My immobiliser isn't going to wig out if I switch ECU's? And this "loom" of great mystery, its between the shifter and the main wiring harness?

Actually, just to clarify my my model; Series One TJ (September 2000). 3.5l, 4 speed auto on a standard shifter. No TCL, no ABS, no power windows... no power anything. But the car does have cruise control.

No, Park is not electronic, the only electronic function is the gear shifts 1,2,3,4,5 etc

Your problem has a few VERY easy solutions

1. Get your trans fixed, new wave spring etc Get the following
Engine harness,
ECU and BEM
Shifter
If you want the dash to show gears also get an upper dash harness and cluster

Get it all out of a Series 2 TJ Exec or Advance thus no TCL

The Keys, cheapest way is to get 2 fresh blanks cut from Mits, programme these into the car and you will then retain your Immob etc


Doing the 5sp conversion is also doable with NO tcl, I have done it and eventually got it working ok, but personally I dont like the 5sp auto as it tends to shift straight into 5th at lower speed, you hit a hill and you have to down shift as the car labours

As for reomote control, you idea of a mechanical set up is certainly the best way to go, the switching you could use a rocking pedal with microswitches, the actual cable system, is there any reason you can not make up some sort of column shift on the RHS of the column or dashboard, being a cable you can pretty much put this where ever you want

Is much more cost effective to flash the new ECU with your existing immob code and retain your factory BEM (will be doing this for mine). Cost is approx $120 for the cable and can be done with all TH onwards magnas (earlier models do not have a reflash connector).

Madmagna

is there any reason you can not make up some sort of column shift

Still means letting go of the wheel.

Thanks for the parts list though, that's the important part. I just have to find a day off work (working two jobs now) for a visit to the wreckers. Sunday I'll get under there console and take as many measurements as possible, start drawing something up. I have access to lathes, milling machines, welding equipment and a source for sheet metal work.... just have to use my manners.

Life

Is much more cost effective to flash the new ECU with your existing immob code and retain your factory BEM (will be doing this for mine). Cost is approx $120 for the cable and can be done with all TH onwards magnas (earlier models do not have a reflash connector).

That's not the Evoscan cable is it? Cables a version 1.3 or something, I have one of those. But, from what I've read (workshop manual) I need the MUT-II tester to re-teach the BEM. Your idea sund like just the ticket. Got any other details?

Use a Evoscan 1.3U (rebranded Openport) cable to flash a TJ onwards (SH2 based) ECU.

Basically download the magna definitions (google them) and read your existing rom using ECUFlash under the vehicle type "Evo 8", if you do not get a definition hit try the type "Evo 7 GTA + Magna AT" (this is included with the magna defs you'll download).

Once you have it up, expand the "Immobilizer" setting and copy down your immob code somewhere safe.

http://imageshack.us/photo/my-images/31/ecuflashimmobvalue.jpg/

Once the new ECU is installed, repeat the read procedure, but this time open the immob section, select the existing value, press the "=" key and enter "0x####" where #### is your immob value from your original ECU.

Once this value is updated, select the "ECU" menu, then "Write to ECU".

You will now be able to start the car.

The BEM's immob is a single-write device, it reads the Immobilizer code from the ECU when first installed, once this has happened the immob value is permanently locked so the value on the ECU needs to be changed, not the BEM. In the event of a failure, Mits will need to replace the BEM (at a cost of over $800 to you).

Use a Evoscan 1.3U (rebranded Openport) cable

Damn. I have 1.3D


The BEM's immob is a single-write device, it reads the Immobilizer code from the ECU when first installed, once this has happened the immob value is permanently locked so the value on the ECU needs to be changed, not the BEM. In the event of a failure, Mits will need to replace the BEM (at a cost of over $800 to you).

Shiiiit....

Just an idea.. Use a cruise control stalk as the gear shift lever. You can always tap it with your fingers without removing your hand. I'm sure you can get a little jaycar kit or something to convert the different resistances of the different switch positions of the stalk and parrallel them to the shifter micro switches.

Of course this would be ideal if your car doesn't have cruise control - as you won't lose that functionality.

Basically download the magna definitions (google them) and read your existing rom using ECUFlash under the vehicle type "Evo 8", if you do not get a definition hit try the type "Evo 7 GTA + Magna AT" (this is included with the magna defs you'll download).

Don't waste time with the Evo 8 vehicle type when the car obviously has an auto ECU.


The BEM's immob is a single-write device, it reads the Immobilizer code from the ECU when first installed, once this has happened the immob value is permanently locked so the value on the ECU needs to be changed, not the BEM. In the event of a failure, Mits will need to replace the BEM (at a cost of over $800 to you).

Don't think so; I understand a new BEM has to be programmed with a MUT so that the ECU and the BEM have the same code. A dealer can do this because they can access the factory records of which code was assigned to each VIN.

Okay, I got held up with work but here's my concept design - mechanical shifter on existing "standard" shifter. Required a ratchet and I needed to make that ratchet as easy to fabricate as possible. Next step is measuring up the car and working around any issues I find in there. There's a surprising amount of room directly below the shifter now, and clearing space toward the firewall shouldnt be a big issue.

http://i.imgur.com/XxMOq.jpg

Mal,

Please explain this for me;

The Keys, cheapest way is to get 2 fresh blanks cut from Mits, programme these into the car and you will then retain your Immob etc

Does this mean all of my cars mechanical locks (doors, ignition and boot) need to be removed and replaced? And, how do I explain this process to the guys at Mitsubishi, what do I take in with me and what work can I expect to do after this stage is complete?

As for the above, yes it does mean that. Flashing the ECU = Much more cost effective.

Now, as for your design, it looks a little overcomplicated. The tiptronic selector activates 3 electric switches when pushed (one up, one down, one sideways to activate "tiptronic mode")

It is much easier to simply splice the wires at the shifter for tiptronic "up" and "down" and replicate these switches at your foot pedal.

it looks a little overcomplicated

Yup. Well, its mechanically simple in theory. Getting it to work exactly how I wanted has proven to be difficult.

Anyway, Mal is sorting me out with the right parts to switch to a triptronic shifter. After that its up to me to splice in as you say and fabricate the foot operated gear lever. The gear lever should be a piece of cake. Old motorcycle gear lever, a bracket to support the required parts and a few other tweaks to make it work.

I'm probably just going to change out the locks to suit a replacement BEM/ECU package... that's not too difficult right?

I just keep reviving this...

I work as a draftsman with a hydraulics and pneumatics engineering company. Essentially I CAN drive a manual transmission. Converting a Magna manual trans to a push button shift (but retaining the original clutch) is the easiest thing to do.

The draw back is the cost. pneumatic cylinders, plumbing and suitable compressor are all relatively inexpensive (a shade over $1000). Where it gets expensive is the programming of the shift pattern - electronic communication between the switches and the cylinders. Even a simple pattern like this would cost around $5000.

Its doable if funds weren't so much of an issue.

If my boss wasnt such a Ford fanatic I could probably talk him into doing it for shits and giggles.

Where it gets expensive is the programming of the shift pattern - electronic communication between the switches and the cylinders. Even a simple pattern like this would cost around $5000.Honestly Skapper, this should be a piece of piss; check out freelancer.com (http://www.freelancer.com/job-search/microcontroller-programming-outsourcing/), you could probably get a uni student or someone in India to code and assemble it for a coupla hundred tops.

Oh no no. Not programming in the computer sense. Pneumatic valving is controlled via a PLC/comma unit and for want of a better term is solid state electronics. Certain electronic inputs are interpreted into pneumatic signals/actions. Like a combination of electronics and pneumatic manifolds.

Nothing "off the shelf".

Google Festo, even better try YouTube. These are the guys on the cutting edge of employing pneumatics in industry.

I'll lean on the Festo rep when he's in next. See just how much it would cost to add puematic controls to a manual trans.

Don't see why you couldn't use a combo of two hydraulic or pneumatic cylinders (horiz and vert) and position sensors, feeding a simple microcontroller; position-sensing hydraulics maybe, dunno the cost, or just a bunch of lever microswitches for example. Each cylinder only needs to go to one of three positions to give the full gear selection range, you could have one "shift up" button (manually doing the clutch with your foot of course), one "down shift" button, and a "reverse" button seperated (on dash maybe) with say a 3sec hold-to-activate safeguard. The programming for such a setup should be dead easy, just a matter of the physical control and sensing.
Wouldn't hydraulic give more precise control and more oomph? eg., x amount of fluid displacement should always result in y amount of cylinder movement?

Maybe I'm being naive having not worked directly with pneumatic or hydraulic control, but I do have a bit of experience with programming for physical interaction, and this seems like overkill.

The mechanics of the pneumatic arrangement is incredibly simple. The contro of it is where things get tricky. Micro-switches and position control sounds very simple - and for the better part it is - but putting it into operating is much more expensive than it sounds. Getting these cylinders and switches to work in harmony and according to driver input is a lot more than just slapping a few switches together.

Pneumatic is the way to go for this sort of thing - if you were to do it. Its easier to work with and avoids heat and oil leaks.

On the positive side you've spurred my interest; I'll get some ideas onto paper and maybe into a 3D model.

Currently - tiptronic and repositioned micro-switches is the winner.

Just as an example - two cylinders work two shift cables. I'm only making an assumption on the exact movement of the shift cables for 3rd/4th. But for this example it'll do.

So, two cylinders using three positions each over seven set modes. Working from only two interactive switches (shift up & shift down). I'll chat to the boss and the Festo guy about the PLC required, but I dont expect it to be cheap.

http://i.imgur.com/dD6mO.jpg

This is all for fun guys, don't take all this too seriously. I'm just having fun with it. If it turns out to be five hundred bucks to convert a manual to full pneumatic controls THEN we'll get serious.

** I forgot to label the shift cables - SORRY! Essentially the manual trans runs two shifter cables - each cylinder would connect to one of these cables. Cylinder one working cable one for example.

Hell, I was imagining a 2 axis gymbal directly driving the gearstick! A lot bulkier though, I would imagine.
Roughly speaking, is three positions a simple feat, or does it rely on external sensors to stop it at a certain position? Again, for shits'n'giggles; just been working on pseudo-code in my head this arvo... but seriously, I can't see any reason why a $2 microcontroller, properly coded, couldn't control this. AND give you audio shout-outs of the gear change at the same time (OK, $4 for that)!

Ha! Yeah, we're getting carried away now.

Like I said, I don't know exactly what cable positions equal which gear. I could be completely wrong. Just me interpreting (guessing) based on what (little) I know about car transmissions.

Positions would be controlled by a reed switch essentially. Well, that would signal the PLC when to stop for each program.

For gear indication I'd just reuse the dash out of a 5 speed tippy.

If there were a solenoid - once again three position - strong enough to do this it would eliminate the need for pneumatics. You could use your micro-controller.... [starts googling solenoids]

I have a feeling that if you were to implement any sort of control like we're discussing here you couldn't retain the original gear stick.

Not sure how small microcontrollers work, but surely a small something program to click relays, to change the solenoids to have the air go into the cylinders to set to either foward or back, not sure how you would get the central neutral though. It has to be do-able without an electrical box full of PLC's.

A small microcontroller can only handle a smallish amount of current, you'd just use that output to trigger a power transistor, and depending on how much voltage/current that's capable of handling, it could either drive the valve solenoids directly or in turn trigger a relay to drive the solenoid.
Optical switching could be more precise and reliable, essentially have a series of break-beam detectors at select positions along the cylinder's throw, so you have some sort of feedback as to where the cylinder is. Again, not knowing quite how these pneumatic cylinders work as far as positioning goes, it's hard to say how easy it is to make it stop in a certain location and keep it there. One issue I'm imagining is if it needs to apply a bit of force to disengage, then all of a sudden there's a lot less resistance and bang it goes too far; hence why I thought hydraulics would be more controllable, as the medium isn't compressable.

Ooh, side thought; how about a stepper motor and rail system? Precise control, plenty of oomph... hell, could even use a driver's power seat assembly, just a question of whether you could mod it to provide a fast enough movement...

ive got a customer who owns 3 ford transits (ex-auspost) that are robo-manuals.
the system employs a 3 button switch on the dash for d/n/r and up/down switches on the wheel.
to select drive the brake must be applied and D pressed once.
to select neutral the vehicle must be stopped with the brake applied and N pressed once.
to select reverse the vehicle must be stopped with the brake applied an R pressed twice.
the system is designed to be used as an auto but if the up/down switches are applied then it goes into tippy mode until the vehicle comes to a stop (or the engine is labored/over revved) where it reverts back to auto mode.
possibly another thing for you to consider is throttle modulation on gear change.

Well. The tippy with a foot shifter will do for now.

Pneumatic shift adaption to a standard manual trans, retaining the need for a clutch to be used, seemed like the next step up from this.

Keeping it simple sort of thing.

I think at the time I bought this up I was on a bit of a bent for a manual trans in my magna.

Bending up some laser cut plate and cobbling together old motorcycle parts along with two simple microswitches will be awesome enough.

Not that I won't press the issue with festo, to get an idea of price and what's required. My boss thinks I'm crazy - he's not a magna fan at all.