MatraSport Forum

Hi Everybody

The original gear shift is a very poor construction, with lots of sloop & slack.
Has anybody tried to modify this ? - eg. maybe use cables from the steering of a speedboat ?


Best Regards
Henrik

The original isn't bad if all the worn components are changed.  I've driven low mileage 2.2's which have really slick changes but it's quite easy to get older cars up to near that quality.

While the original sockets had nylon liners we find that  solid steel ones are better and you can change the gearbox relay bushes from nylon to brass.

The worst part of the system is the top hat washer under the gearlever. but it's a cheap item and easy to change to get rid of lateral slop.

If you email off-list I can supply all the parts to bring it up to spec.

Hi
Most of the ball joints you can get from the local Valmet tractor dealer (or any other tractor workshop) if you are living in rural areas. This ball joints does not have any nylon bushes.

The tube under the car is flexing quit a lot side wards. I have made a support of this longitudinal gear tube, just in front of the engine. This has improved my gearshift tremendously. The support is mounted on the cross member between the fixation for the rearweel suspension. The mounting can be done without any drilling to or modification of the car.

Regards Hans

Ok, I helped myself. First, take a look at this picture of the inside of the gearbox:

(http://tmp.dinsen.net/cxgearboxinside.jpg)

Picture is from http://www.atspeedimages.com/1988_citroen_cx/clutch_and_gearbox/ (http://www.atspeedimages.com/1988_citroen_cx/clutch_and_gearbox/)

Note the dark grey forks (where the red arrow points) that move the syncros (shown with the yellow arrows). There's one fork for 1-2, one for 3-4, and one for 5-R (5'th gear is obviously not fitted on the picture here).

This drawing is in the workshop manual:

(http://tmp.dinsen.net/murenalinkage.jpg)

G is the one that operates the stuff inside the gearbox.

The long rod B rotates and shifts from side to side as the gear lever is moved from side to side. These two actions will work slightly against each other, but it is the rotation that lifts G up to select the appropriate fork inside the gearbox.

People with experience tells me that the system can be restored to near-new-standard. I'd recommend replacing joints and checking the pivot points.

Hans, I don't quite get that. The tube is supposed to move sidewards.

- Anders

which is true, but I guess the sloppyness can be so dominating, that fixing ANY part og the linkage can produce useable results. I know that more cars have Hans's solution fitted. (I think it was proposed by our own Stig .. well not THE Stig, but close)

However, personally I'll still recommend replacing all the worn parts. Specifically around the gearlever, and the other ball-joints.

On my 1.6 the knuckle-rods in front of the gearbox (connecting the L-shaped lever on the longitudinal rod with the gearbox) were the main reason. The ball-joint at the back of the longitudinal rod is also known to deliver quite some slack.

(Since then I had the engine converted to the XU9JA, which incidently gave me a brilliant gear-linkage, courtesey of  MatraMagic, but thats another story)

/Lennart

So where is it fixed, and how?

If the long rod can rotate around it's own axis, then the rotation will have the opposite effect as the sidewards movement, thus making it far less precise.

Pictures?? :-)

Well, what you use from the movement is the rotation around the longitudinal axis, en the forward/backward movement, so even if the system was designed to move sideways as well, it is not really mandatory for the gearchange (at least on a 1.6, which I know the best)

It was a little triangular thing, with a bearing holding the rod steady, while still letting it rotate and slide forwards/backwards.

It was fixed on the chassis between the trailingarm mountings.

but personally I will still recommend renovating the original system first, and only resort to this modification if everything else fails.

In fact, if you get the ball joints under the gearlever and the ball joint at the far back of the rod just right, I think the above modification will actually inhibit gearselection, as It requires slack in both ends. (This is in fact the slack it "eats", making the gearchange feel better)

/Lennart

Exactly my point! :)
Which means: Renew the system, don't make the modification.
Hwr, I'd still like to see the mod.

My car has a relativ low milage (33K)
I have only been driven it for about 300 kilometers. It seems to be better now, when it has been used a little.
I will maybe try to work it over a little with oil and/or grease in the junctions.

Oil may help to eliminate friction, but make sure it is compatible with the nylon bushings.

I had a very sloppy gear stick and imprecise gear changes.  Also, in first or second gear the stick was under my right knee.

I started off by changing the rear ball joint.  This didn't really make much difference.  Changing the top hat bush at the bottom of the gearstick made no difference at all.

I then replaced the three threaded link rods and all of a sudden things are a whole lot better - much more precise changes but the slop in the gearstick is still there.

Last night I removed the gearstick and found there was an awful lot of play in the nylon ball and cup assembly.  I ground the cups down until the play was gone, made some cardboard spacers (all per Roy's site), lubricated and reassembled it.  At the same time I welded together the two sections of the gearstick that slide up & down one another (for reverse gear selection) and cut off the reverse gear gate.

The difference these last two steps have made is fantastic.  It works like a proper gearchange!

Only problem is that the stick is still under my right knee.  Not as much as before but it's there.  It looks like this is how it should be too as the stick is mounted angled slightly towards the driver.  So do I get my adjusting hammer out?  Or is there a more delicate fix?

Cheers,
Roly

It's great to read about your experiences, thanks for sharing it! :)


Maybe the long rod B is bent towards the right (i.e. find the adjusting-hammer). Or it's turned so that A is not vertical. Or rod D is too long. I'm referring to the diagram (far) below :)

A short note to inform that I have changed my description of the system below because I now understand that it's the rotation of the long rod that lifts the selector in the gearbox. As Lennart correctly pointed out.

Hi all,

A vey old topic I know, but I thought as no pictures were ever provided of gear linkage modifications, I would attach some of the very simple modification I did some years ago to my standard gear linkage. (Having serviced and adjusted the linkage for optimum performance first!)

N.B. If you do not have the lower "Cross Brace"  fitted to your car, obviously you cannot carry out this modification.

I utilised an E.O. 22mm Gr3PP Plastic Pipe Clamp (Available in any good Norwegian Shipyard when the storekeeper is not looking  ;) )

The 22mm plastic hole was drilled out to 25mm  + to give a loose fit around "Rod B".
Two 6mm holes 33mm apart were drilled centrally in the cut out in the "Cross Brace" above "Rod B".
The assembly was then fitted around "Rod B" with a good application of grease and bolted to the "Cross Brace" with longer M6 x 60mm bolts and nyloc nuts.

It had great effect in minimising sloppiness and firming up the gear lever operation.

Question:

I was told recently by Titus Taylor that he had seen a car which had the outer connection point of linkage "Rod K" moved closer to the pivot point of "Arm G" which sits on top of the gearbox. This effectively created a shorter throw of the gear lever to change gear, ie. a quickshift. (A bit more arm effort may be required though, which I am not bothered about)

Does anyone have any further information on this, a picture or dimensions about this modification, before I set about driliing holes everywhere, failing miserably and go down the pub instead?

Cheers,

Graham

I am thinking about replacing the entire assembly with a cable activated one.
My Xantia has cables instead of rods and it makes a huge difference. Think the leaver from a mr2 will fit, custom brackets shouldnt be to difficult to weld up.

Hi Graham

Your modification is neatly made, but it must be working against the geometry of the gear linkage. As your photos also show, rod B is not straight, so by fixing it like you have done, something will have to "give" when you move the gear lever from side the side. Only if your fixture was vertically in line with the line going from the gear lever pivot point and the pivot point of joint I, it would work without "give".

I have done a bit of editing on the linkage drawing. I'm not sure about the actual length of the rod B, but this should illustrate the problem. The red dashed line is the line around which the rod swings from side to side, and either ends of that are (supposed to be) fixed. With your modification, either the rod is bending, or one or both of the joints are be loose.

(http://tmp.dinsen.net/murena/gearlinkage-pivot.jpg)

Regarding your "quick shift" modification - it sounds like a good and simple idea. I think the easiest way to implement it would be to drill two extra holes a few cm apart on arm G and weld-in nuts on the rear where you can fit custom (longer) links K. I think you would have to experiement to get the desired effect.

/Anders

Hi Graham.

Check out the topic that I started on mechanical challenges.
http://www.matrasport.dk/forum/index.php/topic,1358.0.html

By reducing the angle on the arm "A" to half of what it was, I got more lateral movement of the linkage for the same amount of lateral movement of the gear shifter. Reducing the length of arm "G" will do the same for forward/backward motion.

Hi Graham,

The car you refer to with the closer pivot is Richards 1.6. If you are going to the Wobbly Wheel meeting tomorow he may be there, although he has had the car for sale for sometime now. If so I will take a photo showing the mod. As I recall it was a simple and effective solution. That being said your slick change seemed all but perfect to me whilst driving your car down the strip at Santa Pod!

http://www.youtube.com/watch?v=9of98pahiUs

Hi all,

Anders, yes I agree absolutely with you,  it does work against the geometry as your red line indicates clearly. You have to spring the "Rod B" into the clamp if not in the neutral gate position. The "Rod B" does though flex easily though between the gear lever position and the fixed clamp position as you change gear.

The parallel in line position of "Rod B" at the fixed clamp as it goes forward and aft and rotates, just seemed to make a vast improvement to my set up. This has given me a very crisp precise gear change, which in sprints and hill limbs has been greatly to my liking.
-----------------------------------------------------------------------------------------------
Jon, thanks for the link to the "Mechanical Challenges" which I was unaware of and the advice. Very good information indeed which I will use.
-----------------------------------------------------------------------------------------------
Titus, sorry I will not be attending the Wobbly Wheel meeting. But if Richard is there, I would appreciate some closer inspection and information.
-----------------------------------------------------------------------------------------------

Cheers,


Graham

I have noticed that the rod is rather soft and can twist quite a lot, and that's probably one good reason for the sloppy gearchange the Murena is known for, but your modification must also put the rear joint under a good deal of pressure?

Anyway... this has made me think and there's a way to redesign the linkage using a universal joint in the middle. The rear link is preserved, your "sliding retainer" is too, and the rubber ball and top hat bush in the bottom of the gearstick. But the long rod is broken in two with a joint in the middle:

(http://tmp.dinsen.net/murena/gearlinkage-pivot-idea.jpg)

/Anders

Hi Anders,

I like that idea a lot. It seems to be a very good and simple development to my half baked effort. I will look around for a suitable U.J. and try it out.
-------------------------------------------------------------------------------------------------

Hi Krede,

How about you trying this out, before going for the more complicated cable solution and giving us a second opinion?

-------------------------------------------------------------------------------------------------

All the best,

Graham

I like it too ;) I have been thinking about something like this for years, never able to find a solution ::) - but now it just seems to be there!

Car Builder Solutions have a nice UJ:
http://www.cbsonline.co.uk/gear-linkage-universal-joint-ujg-922-p.asp

Note, that the rear rod piece needs to be straight for this to work, so while the front part can be made with a shortened old rod, the rear will have to be made with a new one. So it might be better to renew both.

The change will cause the gearstick to rotate slightly more around its own axis than it does with the original system, but I don't think that's going to cause any problems.

Don't forget to check the top bush in the gearbox, Graham. It's often worn and allows quite a bit of rocking of the rod down the gearbox.


I agree, the cable solution is going to be far more complicated, and I think this will be just as good. Cable is preferred on many new cars because it's cheap.

/Anders

Hi Anders,

A perfect choice for me.  The 22mm diameter of this U.J. should fit inside “Rod B” with a slight reduction on the lathe. I will fit the Neoprene Boot also.

Car Builder Solutions is just down the road from me. I think I'll treat the girlfriend to a nice innocuous drive out to the countryside on Monday.  ;)

 (She has been there before, usually via Maidstone Machine Mart!)

All the best,

Graham

Hmmm, trying to reduce the sloppy gear change by introducing yet another link???


This is the way to go.
And fix the play of the ball joint of the gear lever itself.....

(http://i282.photobucket.com/albums/kk242/bileau/Afbeelding220.jpg)

I am very interested in what you guys work out. I have had the gear shifter out a few times trying various things to reduce the play in the ball.
No success yet, but my next attempt will be either to cast some additional plastic onto the existing ball and then turn it on a lathe to the right size.
Alternatively I will turn a whole new lower shifter part that includes a ball in metal.
My third option involves a more complex solution, replacing the ball with a gimball assembly. If that is possible it should be much less susceptible to the wear that the ball solution has.

Yes, but not just a link - also the fixture of the rod.

The sloppyness is due to several factors. Worn parts being just one of them, but even when restored to perfect condition, the gearchange is not very precise in the lateral movement. The reason is that even a very small amount of flexing of the rod will allow for a lot of movement on the gearstick. And since the rod is just soft iron, it bends easily. A rod like that is best for transmitting rotation and longitudal movement.

By fixing the rod in the middle like Graham has done, he has prevented any flexing of the rod on the middle. It can only slide and rotate. This is a modification of the way the original system works, where the rod rotates AND swings from side to side. He also moves the pivot point of the arm A down to the rod itself instead of around the middle of it, thereby "amplifying" the motion of the gear lever's movement from side to side. Since Graham's car is race prepared, I am sure that is a desired effect.

The fact that Graham has made this system work at all without the link in the middle of the rod shows how much the long rod actually flexes!

A high quality UJ link in the middle will not cause any additional play in the system - these types of joints have been used for years on race cars, where you wouldn't tolerate any play.


Lucky you! I would love to visit that shop one day - they have a lot of great stuff!

Enjoy!


I have seen some older formula race cars or kit cars (not sure, and can't find a photo), where the gear lever has a small rod welded on with a spherical bearing at the end, which is then bolted to the chassis:

(http://www.accentbearing.com/bearings/rod%20ends/41338.jpg)

Of course the gear lever will then pivot around a point somewhat off its axis, but I don't think that would be a big problem and all parts would be renewable.

/Anders

Here is Richards mod.

That looks like a very simple modification. I didn't think about the fact that the arm has a slightly bent shape so the control rod can still reach to the other hole... great idea!

/Anders

Yo Krede, if you do that, please update this thread while you go along. I would love to get rid of that multi-jointed system, in exchange for wires. So sad to have a car, that looks cool, drives cool, but has a gearshift system, that makes it feel like a tractor from ww1.  ;)

Best,
Jesper

The reason that we (Hans is too) are thinking about replacing the linkage with wires is really more of a space issue, as the standard system gets in the way when dropping in a turbo engine.

However. As I have put my turbo project on hold for the moment (Cost , time, facilities) so is the wire linkage stuff.

Mr.Dinsens solution looks very sound to me though (I expected nothing less from him ;) ).
Cutting and Fixating the "B" rod is surely the way to go about it, and I think, that if done right it could make a HUGE difference at very low cost and effort.
I would fabricate the fixing point as a rod sliding inside a decent length of tube with just enough tolerance to allow lubrication by fitting the tube with a couple of grease nipples, and then fit each end of the tube with a small gaiter to keep any excess grease from going everywhere. :)
  

Hi all,

Before I start modifying my existing "torsion bar" gear linkage, I thought I would try out Richards modification (see previous picture from Titus) to reduce the throw of the gear lever by shortening the pivot point on arm "G" and actually take some measurements at the gear lever throw, using a calibrated piece of 4" x 2" timber!  :)

I found a suitable Ball and Socket Joint to bolt onto the arm "G" at a firm in Dorking U.K. called Automotive Components, at the grand cost of £1.54 each. See attachment. However I decided as arm "K" is not actually horizontal, I would make up an assembly using a female M8 balljoint from Car Builder Solutions and an assembly of track rod Go-Kart components utilising a 55mm high tensile bolt to bring rod "K" into the horizontal plane, thus increasing the applied force on arm "G" as much as possible.

I had an educated guess of where to drill the hole in arm "G" and decided on 60mm from the existing socket. The water reservoir was moved to one side and the drilling of the hole was possible to do in situ. Just make sure to cover up the drive shaft and gearbox vent below first though.

The assembly was bolted on and road tested out O.K. The extra force required is noticeable, but with a firm action which is quite acceptable to me.

The actual distance of measured gear lever throw before modification was 140mm. The resultant measured distance of gear lever throw after modification was 100mm.  Which is approximately a 30% reduction of gear lever movement for very little effort and expense.  Well done Richard!
 
See attached images for further clarity and general interest to all I hope.

Cheers,

Graham

NICE!.... I bet quite a few people will be doing this modification soon.
And Good work on the "special" calibration tool! :D ...

Good solution.
This will make it almost quickshift :P
Now a few new synchromesh,  and I wil be undefeatable at a trafficlight ;D

I may try this mod myself. But while I am waiting for my new cooling pipes to come back from electro polishing, I thought I would put the time to good use by trying to fix the play in the ball on the shifter arm. On mine you could pull the shifter up/down about 1mm due to wear of the ball. The nylon housing pieces were pressed fully together so there was no adjustment possible.

After thinking of how to replace the ball I decided that it was not practical and, save from making a whole new assembly in metal, I had no really good solution. (I'm too cheap to by a new one  8))

That is when I remembered a teflon sheet that Jan had given me years ago for some project. It was 1/10th of a mm thick and there was enough left to wrap the ball. Worth a try. I cut a piece big enough to wrap the ball and extend through the nylon housing parts. (photos) This proved to be exactly enough to remove the slack and make the ball rotate nice and firm in the housing, with no sticking. The housing squeezes a small fold on the sheet on two sides, so it should stay in place.

The metal plates were cleaned in phosphoric acid and painted with Hammerite. The whole assembly feels like there will be no play and the teflon is nice and smooth, so it should hold up nicely. If not, hopefully it will last long enough to be worth replacing every so often.

Looking forward to feeling if it makes a lot of difference. ;D

PS. I am sure you notice the nice clean work bench.  :D That is a coming project in itself. ;D

That is a nice solution.
Hope it will last.
My murena has a steel ball in it, so I have no problem with this.
Somewhere in 1982 the plastic ball made place for a metal one.

Does your metal ball version aslo have the reverse lock-out pin seen in my bottom photo?

Looking in my list of modifications I found a bulletin dated juli 1981.
I quote this translated.
Since chassisnumber 5491 the reverse locking is placed in the gearbox.
March 1982 when old stock was finnished they changed all gear levers to the metal ball without locking pin.

Part 1 of this modification I have now on my car to.
I used a holder for a telescope from a rifle .
Tapped for M10 and made the hole  wide enough to fit.
(http://img411.imageshack.us/img411/5956/p1020076medium.jpg)
(http://img682.imageshack.us/img682/3387/p1020080medium.jpg)

It is very short shifting and a bit more heavy to change gears, but I like it.

Movie.

http://www.youtube.com/watch?v=5Si49t-haeo

In a short while I will do the other part to.

Is this the right way to go?

The front end of the rod is moving sideways, because that's what the stick makes it do.  To operate the levers on the gearbox, the rod only have to rotate and move forward/backwards. When the stick rotate the rod, the stick also move the rod sideways. That's the way it is. If this sideways move is not wanted, the rod have to be attached to the stick in the center of the ball!

So fixing the rod in the middle..... and the stick moving it sideways.... Stress!  The stress will try to bend the rod, and put tension on the gearbox levers, causing excess wear.


I think the only right way to go is taking the slack out on every joint, especially the ball of the stick.

Disclaimer..... personal opinion

I think you made a point, but 1 mm play in the middle will give it cm's on the downside of the stick because of the long distance.
Because of no play at all on my ball joints I am very short shifting now.
I am thinking about giving it a tiny bit more play.

With the gear lever end of the main rod clamped in a vice I measured the deflection achievable at the point of the new bearing assembly 110 cm further down. Applying a small force with one finger at the furthest rear end of the rod produced easily 4 cm. of sideways deflection.

These are the bits I used to do the other part.

 ???
Try this again, but now with the rod clamped in the vice at the point where the additional bearing will sit. (behind the upright lever) And try to bend the gear stick end of the rod.

Because that's what is happening. The additional bearing is fixing the middle of the rod, and the gear stick is moving the rod sideways.

Hi Bart,

Your right. A few too many rum and cokes last night I think!

With the bearing area in the vice and a similar force applied 2 cm. of deflection was observed at the gear lever end.

Cheers,

Graham

quote
--------
With the bearing area in the vice and a similar force applied 2 cm. of deflection was observed at the gesar lever end.
======================================================================

So making the new bracket a tiny bit oval (about 1mm) and there will be  more then enough play to have no stress. ;)

A little play added, less stress.
Movie.
http://www.youtube.com/watch?v=m0fY4qvUvOY

Part 2 modified.

Used this balljoint.
Movie.
http://www.youtube.com/watch?v=LMWpYH8-MLM&feature=player_embedded

Mounted at 4 cm from the original.
(http://img687.imageshack.us/img687/8903/p1020125medium.jpg)

The final result.
Movie.
http://www.youtube.com/user/pikvanmanus?feature=mhw4#p/a/u/0/e4-u6OX2-nw

Very pleased with this modification.