Shift Return Spring Example Needed

svsgt1 · 07-02-08, 03:22 PM

I built a shift setup for my Audi 01E box using a cable shift system. When I hook up everything the audi return springs are not strong enough to bring the shift back to the middle position for the left to right motion. (Front to back works just fine) Have any of you guys engineered a return spring for their shift setup? Can you please take some pictures so I can get a clearer idea of how to make mine? I have a few ideas but some examples would really help.

Thanks,
Adam

andys · 07-02-08, 11:28 PM

Adam,

Go down to your local pick and pull and find a mid '80's Toyota front drive shifter (very cheap). They utilize a mousetrap type spring arrangement for centering the gate (left to right). I have photos, but nothing that shows the spring as it's well hidden behind the gate actuator. You can also consider the arrangement used on Porsche shifter towers (on cars that utilize the shift rod), a kind of an opposing compression spring configuration. I may have a photo of the underside of one of those somewhere.

Andy

Andy Hills · 07-03-08, 02:16 PM

The return spring on the Audi gearboxes is a bit weedy as the linkages are much shorter on the OEM application and the lever assembly also has return springs.

I used a torsion spring on my linkage.

Even with this it is still not quite enough, so I intend to add another onto the lever assembly.

Hilly

svsgt1 · 07-08-08, 03:09 PM

Andy,

I tried the torsion spring and it doesn't work so well. I did a little research and found out why. Springs are linear in their force so if you only have 1 spring to act in each way it must be at rest for the center position. Therefore it will generate less and less force as it gets back to the center position (rest position). So it will start to move it back but will eventually run out of force. So the spring must be preloaded essentially. So 2 springs must be used. BUT they can NOT act upon each other otherwise they will negate each other when the position gets close to the center/rest position. So basically you need two springs (one for each shift direction) that are preloaded and that have stops so they do not act upon each other when in use. A bit cumbersome but that is the only answer I can come up with to get it to work properly. Someone please correct me if I'm wrong on this.

So I'm going home tonight to open up the shifter on the box to see if I can take the stock springs out and replace them with something stronger. I'll keep you guys informed. Any info on what the inside of the shifter looks like would be helpful but I'm sure I will just figure it out as I do it. How hard can it be right. (Famous last words) LOL

Adam

andys · 07-09-08, 12:56 PM

Quote:

Originally Posted by svsgt1

Andy,

I tried the torsion spring and it doesn't work so well. I did a little research and found out why. Springs are linear in their force so if you only have 1 spring to act in each way it must be at rest for the center position. Therefore it will generate less and less force as it gets back to the center position (rest position). So it will start to move it back but will eventually run out of force. So the spring must be preloaded essentially. So 2 springs must be used. BUT they can NOT act upon each other otherwise they will negate each other when the position gets close to the center/rest position. So basically you need two springs (one for each shift direction) that are preloaded and that have stops so they do not act upon each other when in use. A bit cumbersome but that is the only answer I can come up with to get it to work properly. Someone please correct me if I'm wrong on this.

So I'm going home tonight to open up the shifter on the box to see if I can take the stock springs out and replace them with something stronger. I'll keep you guys informed. Any info on what the inside of the shifter looks like would be helpful but I'm sure I will just figure it out as I do it. How hard can it be right. (Famous last words) LOL

Adam

Adam,

The Toyota moustrap (torsion) spring is pre-loaded at the levers' center position. While it is true that the spring force is lowest at this point, the pre-load is high enough to be effective Also the spring is over-wound in its normal position, if I remember correctly.

The attached illustration is of the opposing spring configuration I was trying to describe. With this setup, there is no cancelling of force by springs working against each other. What I recall (from poor memory), is that there are four compression springs on the underside, two on each side of the shift lever. There are two pins about 3" (76mm) or so apart anchored to each side of the shifter housing. The springs ride over these pins with two bars retained by clips that can deflect when the lever moves from side to side. Hope this helps.

Andy

Jim c · 07-10-08, 08:39 AM

Where the shift shaft goes into the box and connects to the selector forks I am certain there are 2 coil springs in there that cancel each other out at rest.
You maybe able to increase the tension.

I have an O1E on a rod set up and it centers strongly, maybe your cables are dragging.

Jim

svsgt1 · 07-21-08, 01:08 PM

Quote:

Originally Posted by andys

Adam,

The Toyota moustrap (torsion) spring is pre-loaded at the levers' center position. While it is true that the spring force is lowest at this point, the pre-load is high enough to be effective Also the spring is over-wound in its normal position, if I remember correctly.

The attached illustration is of the opposing spring configuration I was trying to describe. With this setup, there is no cancelling of force by springs working against each other. What I recall (from poor memory), is that there are four compression springs on the underside, two on each side of the shift lever. There are two pins about 3" (76mm) or so apart anchored to each side of the shifter housing. The springs ride over these pins with two bars retained by clips that can deflect when the lever moves from side to side. Hope this helps.

Andy

If the toyota torsion spring is preloaded then it will work but it will only be for one direction, so two will be needed and they will have to have stops so they do not act upon each other. The illustration with the compression springs will also work but would also need to have something so they can be preloaded.

Don't get me wrong, you don't need to have a preloaded spring for all circumstances. It is just that if the mechanism won't go back to center with some force then it will need to have a preloaded spring. The springs force will always go to 0 at the rest position.

svsgt1 · 07-21-08, 01:12 PM

Quote:

Originally Posted by Jim C

Where the shift shaft goes into the box and connects to the selector forks I am certain there are 2 coil springs in there that cancel each other out at rest.
You maybe able to increase the tension.

I have an O1E on a rod set up and it centers strongly, maybe your cables are dragging.

Jim

Jim,

Actually the way the 01E returns is very smart and simple. It does NOt use 2 coil springs. It actually uses only 1 spring for the left/right motion. The shaft itself has a V shaped grove. There is one preloaded spring with a ball point tip on it. It acts perpendicular to the V. The center/rest position is the middle of the V. When you move the stick left the shaft and V moves left and the spring compresses. The spring force acting downward on the V forces the shaft to move back to the center position.

I was debating whether or not to make my own spring here with a higher spring rate, but decided it woudl be easier to make another system with 2 pull springs. The stock spring thingy is about $70-80 so they are not cheap. But they are pretty cool as they are threaded and just screw into the box. Very clean and simple and work great.

Adam

andys · 07-22-08, 01:18 PM

Quote:

Originally Posted by svsgt1

If the toyota torsion spring is preloaded then it will work but it will only be for one direction, so two will be needed and they will have to have stops so they do not act upon each other. The illustration with the compression springs will also work but would also need to have something so they can be preloaded.

Don't get me wrong, you don't need to have a preloaded spring for all circumstances. It is just that if the mechanism won't go back to center with some force then it will need to have a preloaded spring. The springs force will always go to 0 at the rest position.

Adam,

You really need to look at the Toyota shifter. The torsion spring has a positive stop in the center of its rotation such that it allows the spring to wind up in either direction yet provide pre-loaded self-centering at the lever's rest. Though it may not provide enough force to return-to-center for your particular application, the principle is a sound one.

My illustration was a best effort while using MS WORD, which is not the worlds best sketch tool. The springs are pre-loaded against the bars retained by clips on the ends of the pins. There need not be a gap between the lever and the two bars. I hope we're "connecting."

Cheers,
Andy

svsgt1 · 07-24-08, 10:51 AM

Andy,

Thanks for the info. I'll take a look at it. It sounds like a good solution. I'm using a custom cable shift unit from Brandwood cars. I was thinking about adding the centering springs on the shifter itself but I didn't like the way I came up with which would allow the springs to work in the left/right directions but then the shaft would rub against the spring plate when shifting forward/backward. SO I decided to go right to the gearbox side of things where the 2 shift directions are seperated. I finished the springs yesterday and it works great. It is a bit heavy at the moment but fortunately I designed it with enough adjustablility that I can lower the tension of the system without any trouble. Maybe I should post some pictures of it some time. Its a nice piece but a bit cumbersome for something that the factory did so simply. lol

Thanks,
Adam

andys · 07-24-08, 12:20 PM

Adam,

Glad to here it's working out for you. Yes, please share your photos.

My only apprehension with adding additional spring force to the transaxle end of things, is the additional load is transfered directly to your cables subjecting them to increased chance of possible failure (theoretically). When the solution is configured at the shift lever end, then the cables realize no additional loading.

If I recall correctly, the Brandwood shifter has no spring centering provision. A local shop of some repute, chose a modified Ford Focus shifter over the Brandwood shifter for a GT40 replica they were re-fitting. I wish I had my camera at the time to take a shot; it looked very nice indeed.

Andy

Big-Foot · 07-24-08, 12:32 PM

Quote:

Originally Posted by andys

If I recall correctly, the Brandwood shifter has no spring centering provision. A local shop of some repute, chose a modified Ford Focus shifter over the Brandwood shifter for a GT40 replica they were re-fitting. I wish I had my camera at the time to take a shot; it looked very nice indeed.

Andy

You are correct Andy - No spring centering.. Frankly I am not overly impressed by the Brandwood shifter. Although I am using a slightly different version of the transmission that the shifter was meant for (see my build site for more info) the feel of the shifter and the gates is more truck-like than SportsCar... I am also not crazy about some of the smaller hardware used such as #10 screws/bolts where they should be far more substantial...