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 - Clutch won't disengage

My clutch won't disengage. 1971 Mk IV 1275 Sprite. Here's the story so far:

Used to work fine. Took the engine out last Dec for rebuild. Didn't touch the gearbox or clutch release arm. Fitted new slave and master cylinder seals and fitted a new clutch plate, cover and release bearing. Definitely put the flywheel side towards the flywheel. Bled the system on Saturday. We began with the old undo / push / tighten / release system but it wouldn't push the fluid down so we switched to using an eazibleed. That seemed to go well and after removing lots of air nice clean fluid came out. Pedal felt ok, perhaps a bit light, but not a lot of movement of the release arm under the car. We let it stand and yesterday (Sun) bled it again. No more air came out.

Today I finally got the car on the floor and tested it. It won't disengage. Not even a hint of it. I tried starting it in gear and then pressing the brake but it just stalled the engine. I took the slave off, held it with the pushrod downwards and uses the easibleed again. Still no air came out. Put it back and no change - still not disengaging.

I don't believe the plate itself is seized. It's brand new and only about 3 weeks since I fitted it. It all went together fine and I double checked everything as I went.

Is there any way to move the release arm without the hydraulics to test if the clutch is disengaging when it's moved further? Ratchet strap to the rear axle and tighten it up to pull the lever backwards perhaps?

What should I do next? I really haven't time to pull the engine again!

Ant
Ant Allen

Right - just been under the car with a ratchet strap :o) The good news is that if I pull the release arm back by 13mm then the clutch disengages perfectly. With the hydraulics connected it only moves by 3mm at most.

So it's my seals? Or air?

Suggestions welcome.

Ant
Ant Allen

If it's not leaking, I'd say air.

Best of luck,

-:G:-
Gryf Ketcherside

It's not leaking, but it's not working either. I've already bled it 3 times and only got air out the first time. Hmmm.
Ant Allen

Jam the clutch pedal down with a chunk of wood tonight and leave it overnight - try bleeding again tomorrow. Has worked perfectly for me twice. Just don't ask me why.
Steve Clark

Made a bit of progress. first off I dismantled the master cylinder and discovered that I'd put the piston seal on the wrong way around. Doh. Stupid mistake. Put it back again, pressure bled it, but still wouldn't disengage.

Then I did as Steve suggested above and left the pedal jammed down over night. I also removed one of the slave cylinder bolts and tilted it so the push rod is pointing slightly more downwards than normal. I connected the pressure bleed this morning (pedal released again but slave still raised at the nipple) and one bubble came out straight away, followed by nice bubble free fluid.

I've jammed one rear wheel, and used the other to test if the clutch is disengaging. Hey presto I can turn the wheel when I press the clutch, and can't when I release it. Brilliant! Only trouble is there's still some resistance to turn the wheel so when I put the car on the ground it's dragging sufficiently that I can't actually select a gear with the engine running.

So I'm almost almost almost there. Options I have are:

1) Repeat the above and see if I can get one final bubble out.
2) I've read in the archives about putting a washer on the slave cylinder head. Is that a solution? I only wanted 13mm of travel so I reckon I'm about 1mm short. I can measure that and insert the correct number of spacers. Can this be done without removing the slave from the car and having to bleed it again.
3) Get a new M/C and new S/C. The master has had the seal in backwards, and the slave has a piece missing where the circlip fits so neither are perfect. But if I spend £75 to £100 on new parts and then it still doesn't work I'll be even more annoyed and have less time to do the above.

Suggestions? This is driving me mad at the moment!

Ant
Ant Allen

Another thought. What if I have a mismatched M/C S/C combination? My M/C is the non-groove type. Is there more than one slave variant? Perhaps of a smaller bore?

Ant
Ant Allen

Ant,
One thing you could try to get it going is to get a 7/16" nut, pull back the slave dust cover and slip the nut between the cone shaped outer end of the piston and the end of the pushrod. It will stay there quite happily and has the effect of lengthening the pushrod, making it driveable at least. If this works, one can get longer pushrods which have the same effect. The nut won't readily drop out - I have run like that for months so it should be a sufficient fix to see you through the next couple of weeks!

Another thing is when you bleed it, try pushing the slave cylinder as far up the bore as it will go and jamming it there - a miniature G-clamp will do this very well. It just prevents air pockets that can otherwise lurk in the dead space of the slave whilst all the clean fluid is pumped by the short route straight out the bleed nipple.

Guy
Guy Weller

Yes there is a 7/16" bore (early cars) and a 3/4" bore (post about 1970) I think that is the right way around.
Using the wrong combination can mean insufficient pushrod travel
Guy Weller

Lengthening the pushrod won't give more travel will it? But perhaps it will pre-load the arm and the travel I already have will end up in the right place.

I am still toying with getting a new M/C and new S/C. It is just possible that the ones I have only work with the old part-worn clutch. Now I have a new clutch then I guess I need more travel in the stroke.

Hmmm
Ant Allen

You are bright Ant, travel is a function of the ratio between master and slave cylinders, and the amount of fluid displaced with a single clutch pedal stroke. But this does assume that the slave piston has free movement within the central part of the bore and the system has no air. A longer pushrod may help if the lever arm pivots are oval, the lever arm bent, or there is some other lost movement in the system.

New clutch shouldn't need more travel in the stroke. But possibly a different design of cover might have a different ratio on the clutch "fingers". Actual required operating travel at the slave is only about 1/2" or maybe a bit less.

Guy
Guy Weller

Ant

as per the other thread - last year I replaced clutch plate,cover plate, thrust bearing and slave cylinder pushrod. On re-assembly I was also having clutch issues - but I hadn't disturbed the hydraulics at all.

My conclusion was that I had enough movement...it just wasn't in the right place - i.e. too much slack at one or other end of the movement. Assumption was that combined tolerances in "length of pushrod", "thickness of thrust bearing", "thickness of cover plate", "play in fork bush" had combined to give a problem.

The solution was to change the length of the pushrod. No change to total movement...but does change how much slack is in the movement. Can either be done by welding/grinding or make an adjustable pushrod. I found I didnt need very much adjustment in the pushrod - it was a few mm max either way to get it right.

Somewhere I have a diagram of my pushrod...but I cant find it quickly....
Dean Smith ('73 RWA)

I reckon one reason for the spate of clutch issues in recent times is that the release bearings are not made to OE spec.
If one compares known original 1098 and 1275 release bearings the castings are different, the 1275 version has longer lugs whch sets the carbon face nearer the engine. I think this is necessary because the 1275 clutch cover is shallower than the 1098 one. However inspection of a variety of used aftermarket release bearings shows all manner of different size lugs somewhere between the two.
David Smith

I welded a bit on the end of my push rod then ground it down to suit, if i had to do the job again i would make it adjustable.
Brad (Sprite IV 1380)

Guy - I measured how much travel I need. 13mm (= your 1/2 an inch) is indeed sufficient.

OK. So I'll be in the slave cylinder tonight with a thick washer or possibly a nut. Unless anyone can suggest a quick way to make it adjustable???

Thanks for the help so far guys. I can't believe that after a full front end strip down and engine rebuild the thing to stop us using this car on Saturday will be the clutch!

Ant

Ant Allen

Ant

Similar thread from the archive - solution also a lengthened pushrod.

http://www2.mgcars.org.uk/cgi-bin/gen5?runprog=mgbbs&access=&mode=archiveth&subject=97&subjectar=97&thread=2008082716281018439

For mine I started with ...

New pushrod (to get a decent "eye" for the clevis pin)
A bolt approx the same diameter as the pushrod.
Some nuts to match.

I welded 3 nuts together. Then cut the pushrod down so its just the "eye" end and short part of the rod which I could fit into the nuts. I then welded that piece into the top nut. Effectivey leaving 2 nuts worth of thread inside the tube of nuts with the clevis eye at the other end.
I then cut down a suitable bolt so that it had a rounded end (as per original pushrod) and then screwed into the top half. Overall length can now be adjusted by screwing the rod in and out of the nuts. Finally add a single nut to act as a locknut.

Actual adjustment is then guess a starting point on the bench (a few mm longer than it was originally probably) then trial and error to put it on and test - then screw the threaded portion in and out to lengthen or shorten.

Its quite tight to have enough space to allow the slave cylinder seal to fit and ensure you dont foul the clutch fork around the clevis pin.

I'm sure someone could produce a much neater version if they had the necessary machine tools. I think Daniel was looking at that once but I cant find the thread in the archives quickly.
Dean Smith ('73 RWA)

try a 1/2 thickness M8 nut behind the pushrod.
Brad (Sprite IV 1380)

just doing the same job on the same car and am anticipating similar problems, the reason being I have bought a new slave rod (MGOC) and its about 10mm too short so will probably end up putting the old one back in!!!!
Also, another parts supply problem (MGBhive but via an e-bayer!!!), the new clutch kit came with a rel. bearing from a 1098, luckily I spotted the difference before fitting so was able to order another, I would guess either or both of these would cause your problems, just hope its not your release bearing.
Mike
mike storey

Ant,
The reason I suggested the 7/16" nut (an M8 would do just as well) is that the hole in the middle is not large enough for the pushrod to slide through, but is large enough to provide a "locating" depression so it remains aligned with the pushrod and doesn't drop out. A standard half nut is also about right to lengthen the rod without risk of overthrowing the clutch.

Other methods like buying a new pushrod, welding bits on and making up adjustable versions may well all work but lack the essential "quick fix" of my suggestion. If the quick fix works, then it may be enough for now, with perhaps a more permanant solution when you have the time. If it doesn't you have at least eliminated that as the problem and can go back to bleeding attempts again!

Guy
Guy Weller

"welding bits on and making up adjustable versions may well all work but lack the essential "quick fix" of my suggestion"

No argument there !.
Dean Smith ('73 RWA)

I'm definitely going to try a nut behind the push rod later tonight. I'm still struggling with the conecpt a bit though. I've already tested the release mechanics and found that it fully released with 13mm of arm movement. Less might be sufficient but I didn't try varying that. I'm concerned with what Mike has written above about the release bearing. I (foolishly) trusted the parts that the MGOC sent me and fitted the new clutch so I don't know if the bearing is correct or not. But if it weren't correct would the mechanics still work?

Push rod length won't give extra travel, unless it has a side effect of putting a pre-load onto the release arm. Then the travel I have will reach disengagement. But the pre-load will be constantly rubbing on the release bearing so it will fail quite quickly, right? Or it will wear down until no pre-oad is needed.

Anyhow I'll try the nut and see what happens.

Ant
Ant Allen

Ant,

IIRC one of the things that can happen if the pushrod is short or the wrong release bearing is used is that the slave piston may reach the internal circlip in the slave cylinder bore and prevent further movement. Extending the pushrod or adding the temporary nut pushes the piston further back into the bore allowing more potential travel in this case. Maybe you can pop the rubber seal off the end and have a look to see what the piston position is like and if it might be the cause of your short travel.
David Billington

I knew the circlip would become an issue at some point. This is where I admit that mine is missing. I only discovered that it was missing when I took it apart to fit new seals. At that stage I figured it's only a problem when it's not fitted on the car as the release arm prevents it popping out in normal use. So I put it back as was with a "must sort that one day" lodged on my "to do when I have lots more time" list. But given that my circlip is missing I know for a fact the piston wouldn't hit it because there would be a puddle under the car by now.

I'll try the nut idea tonight, maybe give it one more bleed, and if neither of those work I'm taking it all apart and ordering a new MC and SC on next day delivery.

Ant
Ant Allen

I think of the clutch mechanism starting from the surface of the clutch fingers (or thrust face if the clutch cover has one of those). There is mnothing in the system to actively move the thrust bearing away from the clutch other than the clutch cover springs themselves, plus a degree of "jostling" of the bearing when the clutch is rotating. So the "at rest" position of the bearing is going to be just lightly touching the clutch fingers. Working back from that through the lever arm and pushrod, this determines the "at rest" position of the slave piston. In theory this gives the ideal situation where there is no lost movement when the slave piston begins to move. The design is cleverly self adjusting.

The distance moved by the slave piston is a function of the fluid displaced. If the start position of the slave piston is too close to the end circlip it will reach this before full disengagement of the clutch, as David explains. This start position can be down to the selection of thrust bearing, and clutch cover, bent clutch lever arm (not uncommon) or short pushrod. The easiest of these to "adjust" is the pushrod length.
Guy Weller

The 'lack of circlip in the slave' comments are interesting. I had a leaking slave cylinder and when I stipped it down there was no circlip as shown on any parts diagram I could find. As the inside of the bore was scored I ordered a new slave cylinder, which came with no circlip! No sign of a mounting point for one either.
Is there perhaps a different design sans circlip?
B Slater

Update - I tried putting a nut behind the push rod but couldn't get it to sit there whilst I reassembled it. I then found a washer of appropriate sizes which sat firmly on the end of the push rod so I assembled it with that in there. At first I thought ti still wasn't working but after a few pedal strokes (possibly repositioning the slave and taking slack out of the system?) it began to work. With the handbrake on (car in the air) the gears are turning witht eh clutch pedal in but as I go to select a gear they soon stop. Engaging with a firm action will be fine.

So, for now at least, I'm leaving the clutch as is and moving on to jobs anew :o) Any more problems in this department and it gets a new pair of cylinders and a new pushrod too!

Thanks for all the help guys.

Whilst we're on the topic, on the web last night I came across a few adjustable push rods - mostly for minis. Does anyone sell them for spridgets?

Ant
Ant Allen

another possibility if the "nut" thing doesnt do it, is to fit a small (1/4" drive) socket over the pushrod inside the slave. Fit the socket bit over the rod and the extra 3/8" inside will give a longer pushrod, that fits in the hole and doesnt fall off when you try to insert it.
Bill sdgpm

Thanks for that Bill. With wedding using wedding cars sevenoaks car duties less than 3 days away I'm not touching it again. I have lights, indicators and some wiring behind the dash to attend to. The clutch lost me about 2 full days over the weekend so there's lots to do. Oh and apparently I have to take seriously the statement "that car needs a damned good cleaning before I get in it with my dress on". So that'll keep me busy too.

Ant
Ant Allen

I had a similar problem years ago, that was due to the fact that the aluminium flywheel was secured by allen head bolts which were hitting the fingers on the clutch. Changed to normal bolts, and it cured it. Took me months to figure it out, and drove me mad!

Neil
Neil Williams

...so grant her permission to clean it then...
David Smith

<< ...so grant her permission to clean it then... >>

Yeah - I suggested that. I'll clean my side for my hired black suit :o)
Ant Allen

Well, if she won't get in with her dress ON...
David Lieb

OK, I think we can stop this thread digression at that point!

Adjustable push rods, like those available for the mini, anyone?

Ant Allen

My suggestion would be to take some epoxy putty and clean the end of the rod really well then put a bit on the end of the rod. After it set up nice and hard just grind it to shape and then fit it in place and grind it down till it had enough to disengage the clutch without preloading it too much. Nothing to fall out or dislodge that way and completely removeable if needed by just sanding it off
Paul Ruszczyk

This thread was discussed between 13/04/2009 and 19/04/2009

MG Midget and Sprite Technical index

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