MG-Cars.net

Welcome to our resource for MG Car Information.

Recommendations

Parts

MG parts spares and accessories are available for MG T Series (TA, MG TB, MG TC, MG TD, MG TF), Magnette, MGA, Twin cam, MGB, MGBGT, MGC, MGC GT, MG Midget, Sprite and other MG models from British car spares company LBCarCo.

MG Midget and Sprite Technical - New Clutch Dragging


I'm very nearly back on the road after 2 years of blood, sweat and tears. As part of the rebuild I've fitted a new clutch, roller release bearing (unknown provenance, it was an EBay bargain) and rebushed the release fork. I've bled the hydraulics and am confident there is no air in there, but I'm getting a slight drag on the clutch. It's just enough that it has to be forced into gear and the car moves even with with the clutch on the floor unless you sit with your foot on the brake.

I've had it suggested just to run the engine with the clutch down and the brakes on for a while, but I think I'm more inclined to get hold of a longer push rod. All of this is a long way of asking whether anyone supplies longer push rods, or is it a case of bodging one?

Thanks in advance,
Mark
MarkH1

Have you checked for wear in the hole at the top of the pedal or in the master cylinder rod? its more important to take out the slack rather than just starting in a differant place.

Carl
C Bintcliffe

What Carl said - eg new clevis pin. After that a longer push rod - how long is the one you have fitted - tip to tip?

Note that the pivot point on new and some new roller release bearings is not the same as the original pivot point and thus you need a longer push rod to get the same travel.
Daniel Thirteen-Twelve

got agree with the above


WE should have a national or international day of check your clivis pin....we could solve 46.38% of all the clutch problems in the world on the midget if we did.

prop
Prop

could it be the slave cylinder not fully bled?

Jeremy
Jeremy 3

The length of the slave pushrod makes no difference to the travel of the clutch release fork. The only circumstance where a longer pushrod would hepedal.lp is if the piston of the slave cylinder is coming up against the end circlip before the clutch is disengaged. The travel is dependant on how much fluid is moved for a single stroke of the pedal, and how much of that movement is lost in worn pivot points.

Guy
Guy Weller

If you want to try the longer pushrod thing before you do anything else, I have a solution. Put a socket (1/4
" drive) over the end of the rod, inside the slave. I don't remember what size I used, but it was just laying around worn out anyway. I used one that fit the pushrod without a lot of play, so I suppose it was 3/8th's or something like that. I've been running the car for years like this, found a few replacement pushrods to use instead, but never needed them or had a bit of problem.

No guarantees, but it didn't cost me anything to try, and it worked well. If you think the throwout is contacting all the time, it will probably wear out soon, could be why you had to replace the other one, so check that out.
Tom

I agree about worn clevis pins and the holes they run in. It's surprising how often it happens, and customers were invariably relieved to find it was a 10 min miracle cure rather than a hefty bill. (Though some were disgruntled because they'd already paid someone else a hefty bill and it had mad no difference).

Another cause in the 1275 clutch can be if the release bearing travels too far. There is not much room to spare before the release ring in the centre of the cover hits the driven plate and pushes it forward hard onto the flywheel and creates drag that way. Symptom of this is the the clutch drag is worse when the pedal is fully down on the floor.
Lengthened pushrods can aggrevate this cause by not allowing the diaphragm to return fully and so the full stroke of operation pushes it further than it should.

I also had trouble with clutch drag when a roller bearing was fitted. Couldn't see why, but returning to carbon fixed the problem. Nearly caused divorce on the Stelvio Pass last year when Bron radioed that the clutch wouldn't disengage and I didn't want to believe her (when will I learn ... in mitigation, I was in the Midget ahead of her which was already clutchless - broken cable. I dutifully U-turned - try a clutchless 3 point turn on Stelvio, it's a challenge - only to be then told the #$@#! car is now back into first and "don't bother!!!". So another clutchless 3 pointer later ... )

I digress

Because of the variety of causes, some of which entail removal of engine, it's a good idea to sort out whether the problem is in the hydraulics or the bellhousing. There is a magic number here, being the amount of movement there should be at the slave cylinder. In an MGB it is 11mm but I can't remember what a Midget should be and our ones here have all got non-standard clutches so I can't measure at present - maybe someone else can help.

If your cyl/pushrod movement is less than spec by 1 mm or more then the problem will be hydrauilcs (this includes the clevis problem). If it is up to spec, then there's nothing for it but to remove engine and find the cause. More than spec and you may find it is the plate hitting the flywheel as above.

Paul Walbran

I'm going to disagree with Guy for the same reasons I made earlier.

If the pivot point on the roller release bearing is further back than on the correct original item, you've immediately LOST several mm of travel. All the perfect hydraulics in the world won't get that LOST mm back. What will get those lost mm back is a longer pushrod


Daniel Thirteen-Twelve

Daniel,
With the clutch engaged, the thrust bearing is pushed back only by the clutch cover springs so that it is barely in contact with the clutch cover. This is irrespective of how thick or how worn the thrust bearing is. So the distance to be travelled remains the same.

However, the combined effect of different components, differing design of thrust bearing, bent clutch fork, different degrees of wear etc, etc can all result in the slave cylinder piston not operating over the correct portion of the cylinder. It can either be pushed too far back up the bore at rest, or can end up fouling the end circlip when operated. In either case an adjustment to the length of the slave cylinder push rod can be made to compensate and bring the slave piston back into its proper operating range.

Guy
Guy Weller

Thanks all,

I'm as confident as I can be that the system is properly bled - as far as I can tell with a two person 'has it moved yet?' test

although there is a bit of slack, I don't think the wear on the clevis is excessive - I put a new pin in and it seemed a reasonable fit at the time. I'll re-check it at the weekend

Also, don't think it's over-travelling as there's no point in the travel where it's fully disengaged.

Think I'll try Tom's idea and see how it goes.


BTW, attached is a picture of the release bearing I took out (closest to the camera, obviously). Amazingly, the clutch was functioning just fine when it was removed from the car!


MarkH1

...and this illustrates the effect of a machine screw in place of a bolt in the release bush. Really difficult to photograph, but basically has worn a threadform into the bush.


MarkH1

Guy,

I agree with your first statement.

I'm not sure I agree all of your second statement apart from the conclusion.

I'd put it like this: A stroke of the clutch master produces x length of travel of the clutch slave piston. X length of travel of clutch slave piston moves the rod which moves the fork which applies pressure to the clutch fingers and actuates the clutch.

Say that X has to be no greater than 10mm (to avoid overstroking the cover) and no less than say 6mm to actuate the clutch.

Then have a release bearing with a pivot point that is 3.5mm further forward.

The net result is that a stroke of the clutch pedal still produces X but now it needs to produce X + 3.5mm to actuate the clutch.

Your point is that OK the first time the clutch pedal is actuated it takes up the extra 3.5mm but the slave piston stays at that position in the slave body. So every 2nd and subsequent stroke of the clutch pedal still produces a full stroke of X so that in fact there is no problem. On that basis a longer clutch push rod really can't make any difference.

So I see you are correct in theory. In practice though a longer push rod does help so why is that? Is it the case that for a weak hydraulic system (poor or worn master) that since X is never achieved the longer push rod helps by ensuring that the bearing is right up close to the cover such that the slave piston is forced back to the part of the bore where it is operating in a fresh area with little or no wear?

Daniel Thirteen-Twelve

Daniel,
My assumption is that the thrust bearing remains in light contact with the pressure plate because, as far as I am aware, there is nothing to pull it back clear of the cover and there should be no slack to take up. Where I agree this theory falls down is if either the thrust bearing is "jiggled" away from the cover as it rotates, or if the slave piston is "sucked" back up the bore by the M/C return spring.

Modern hydraulic clutch systems (like my concentric slave arrangement) are designed so that the release bearing is firmly held in contact with the release pad so there is no slack to take up before operation.

I am certainly not denying that a longer pushrod may help in certain circumstances - I have used the nut on the end of the pushrod trick myself - but I don't believe that changing the pushrod length can alter the actual clutch travel.

Guy Weller

Guy, there's a return spring in the slave cylinder too, but it's very light and I reckon that too much stiction means the arm will stay where it's put anyway. I'm following this with interest as seemingly like many of us I don't really understand the clutch throw mechanism. i *think* it is important when installing and bleeding, to first move the clutch arm so the bearing is in contact with the thrust plate thus putting the slave piston in the right position from the outset. It is possible that if the release bearing is too 'short' then the piston may reach end of travel before the clutch is disengaged, this is when a longer rod may be useful. There have been several different versions of carbon bearing over the years and one theory is the aftermarket ones are not made to original dimensions. Also I think maybe 1098 ones are mixed up with 1275s (the cast carrier is a different size) leading to this type of problem. Roller release bearings of several types and sources, together with removal of the thrust plate all go to making the stroke position and distance required quite hard to calculate. Sorry for the ramble.
David Smith

As an aside, once a flywheel has been skimmed this also moves the release bearing further from the cover in its initial position and so again while the slave ought to cope with this, if it retracts after operation [I'm pleased to agree with Guy that the concentric doesn't] this can also contribute to not only a very long sentence with little punctuation, but running out of travel for actuating the cover.
Daniel Thirteen-Twelve

The spring in the slave cylinder is not a return spring, but acts in the opposite direction - to keep the piston pushed out so that the release bearing retains the "light contact" which Guy describes.

This is fundamentally different from slave cylinders such as the Mini which have a return spring (external) which pulls the slave cylinder piston fully home each time the clutch pedal is released. Such systems require adjustment from time to time as the clutch wears and the centre of the diaphragm sits further out from the flywheel in response.

The system in the Midget (also used in the MGB) is automatically adjusting as the internal slave cyl spring maintains a zero clearnce situation. The available travel in the slave cyl is much greater than what is required to operate the clutch, so the piston strokes from where it happens to settle once zero clearance between release bearing and cover is established. As the clutch wears, the release ring "off position" moves back from the flywheel, pushing the piston further into the bore to achieve the zero clearance position.

All a longer pushrod will achieve is to push the piston further up the bore. This won't be a problem unless it goes so far that it bottoms out, which may not happen intially but as the clutch wears and pushes the piston back it may do so. If this happens, the clutch is not fully released resulting in premature release bearing failure and/or clutch slip.

Back to the original problem:

I'm wondering whether Guy's jiggling comment is a clue. As I noted above, we had similar problems with the roller release bearing we had fitted. One of the symptoms was a slight shimmy in the pedal, and certainly the release bearing had worn significantly on the back of the lugs which engage with the fork (so had the fork, which needed replacing. It occured to me that the weight of the bearing offset so far in front of its fulcrum might be contributing to the problem. Whatever the cause, the effect was plain to see, and the clutch now works properly with no other change than the carbon release bearing.

I can vouch for David's comment about 1098 and 1275 bearing being mixed up. We have had wrongly bearings on a number of occasions, and from more than one supplier and it is something we need to keep a careful eye on as the 1098 bearing in a 1275 definitely causes problems - premature failure at least.
Paul Walbran

Paul,

I agree with what you say so why should it matter if a 1098 bearing was used instead of a 1275?
Daniel Thirteen-Twelve

Two reasons:

1. The clutch fork has to move further forward to compensate. Since the bearing swings through an arc about the fork's pivot, this means that the bearing becomes significantly off-centre compare with the gearbox first motion shaft. Typically, the shaft contacts the release bearing and wears rapidly it it. Also being significantly off-centre on the release ring is never helpful for good bearing life.

2. Under certain circumstances (most commonly with a new clutch, especially of the linings are slightly over-thick), the fork has to go so far forward that the piston hits the circlip and runs out of travel. result is failure to disengage fully. This is when a longer pushrod would be a useful fix. However, as it doesn't address the release bearing wear, it should be regarded as an interim measure to delay the major work to a more convenient time.

We found the wrongly labelled bearings the hard way, by way of a claim from a customer for supply and fit another clutch. On looking into it I was convinced that it was a fair claim as per 1 above. Unfortunately for me the broken bearing had taken out the cover too.
Paul Walbran


Just to close this one out for anyone trawling the archives, I sorted it out today by lengthening the pushrod. Actually created an adjustable pushrod by shortening, drilling and tapping the slave cylinder end of the rod so a screw and locknut could be fitted. By my measurements, it's now about 3 or 4mm longer than it was before, and the clutch is behaving beautifully.

Thanks all for your help.
MarkH

This thread was discussed between 27/08/2008 and 06/09/2008

MG Midget and Sprite Technical index

This thread is from the archive. The Live MG Midget and Sprite Technical BBS is active now.