[Brady_Hill27]

Yesterday I was driving around and my clutch started feeling weird. The car was still shifting fine even though the clutch felt 'empty'. About 10 minutes afterwards I found that I couldn't shift at all. There is little action from the clutch, just seems like it's not engaging all the way. I checked the reservoir and it looked fine. Any ideas?

P.S. the tranny is fine, i can put it in gear when the engine's not running, 180k miles on the car. Thanks

Edit: I can shift when the engine is running, but it needs a lot more force to do so. Clutch still has significantly less resistance than normal

 

[unclemiltie]

mileage?

Are you leaking fluid? If so, check the clutch master cylinder to see if that's what is leaking. If so count yourself lucky and replace/rebuild it.

If not, there is a hydraulic slave cylinder, unfortunately it is inside the bell housing and the only way to get it is to take the transmission out.

Those are the only two components to the hydraulic system so it's likely that one of them is your issue.

 

[stuf]

I know him. I think his mileage is around 180k. I don't believe the clutch has been replaced yet either. The miles are mostly commuting to work long distances tho. Regarding the master cylinder leaking...how much fluid would have to leak to make a difference or is it more of the issue of a bad seal allowing air into the system?

If it is the slave, is it a hard fix after removing the tranny? Or pretty straight forward after we get past the removal? And to confirm, he fluid is supplied via the brake master cylinder, correct?

 

[BikePilot]

Yes fluid is supplied from the brake MC. The fluid is taken from about midway up the brake MC so if the clutch springs a leak it won't take out your brakes. Of course if the brakes leak your clutch will be the first thing to stop working.

I don't understand the description of the symptoms.

Does the clutch pedal itself feel normal? Similar freeplay and resistance?

Does the clutch (the bit that connects the engine to the transmission) release fully? (can the car roll easily in gear with the clutch in?)

If it has plenty of fluid and the clutch pedal is overly soft then the issue is most likely with the clutch pedal hydraulics. This can be the clutch line leaking, slave cyl leaking, or master cyl leaking. The master can leak internally or externally. An internal master cyl leak won't cause fluid to leave the system and drip on the ground, rather the seal on the master cyl piston will not seal totally and so it won't move as much fluid as it should.

If it is low on fluid this could be due to brake pad wear, a leak in the clutch system, or a leak on the brake system.

 

[JohnLear]

Yesterday I was driving around and my clutch started feeling weird. The car was still shifting fine even though the clutch felt 'empty'. About 10 minutes afterwards I found that I couldn't shift at all. There is no action from the clutch whatsoever, however it feels fine. Just no action coming from it. I checked the reservoir and it looked fine. Any ideas?

P.S. the tranny is fine, i can put it in gear when the engine's not running. Thanks

Edit: I can shift when the engine is running, but it needs a lot more force to do so. Clutch still feels empty

Brady, you say; "the clutch felt 'empty' ", then you say "however it feels fine". These are mutually exclusive statements, so it's confusing.

At any rate, from your description of the symptoms it sounds like it's quite possible there is air in the hydraulics. There may well be fluid leaking where you can't see it, but it's possible that air can get in with zero or a very minor amount of fluid leaking.

When the pedal is pushed this creates a pressure inside the system that forces the seals outward against the cylinder bores creating a good seal (but can still leak if the seal edge or cylinder bore is damaged). When the pedal is released the clutch spring (i.e. diaphragm) pushes the fluid back the other way and keeps the seals under pressure and thus sealing. When the pressure drops to zero (foot off pedal) the seals only have their inherent over size to prevent fluid leaking or air entering the system (i.e. the relaxed seal rubber would be a larger OD than the cylinder ID if the seal were not 'compressed' by the cylinder bore, so the seal 'pushes' lightly against the cylinder wall even when no pressure exists within the system).

But, if the seals are old and have taken a permanent 'set' so that in a relaxed state they are no bigger than the bore ID, then when no pressure exists in the system they could slowly seep fluid in one direction and / or air in the other direction. For pressure to rise in the system (and force the seal lip against the bore wall) the master cylinder piston has to start moving, and if the seal is not 'naturally' pushing against the bore wall then this is the moment when air is most likely to enter, i.e. just as the piston starts to move but before internal pressure forces the seal lip outward against the bore wall. The amount of air may be tiny, but over many pressure cycles can build up to a quantity that affects system operation.

Such a leak could be at the master cylinder or the slave cylinder, or both. If I were betting on it then I'd lay some money on it being the slave cylinder, since this will be exposed to higher temperatures than the master cylinder, which is likely to have an affect on the seal.

The short term solution would be to bleed the system. This will probably help, at least for a while (until more air gets in). If / when the symptoms return, since it appears necessary to remove the gearbox to access the slave cylinder, I'd replace the master cylinder first and see if that fixes the problem longer term. If not then the slave cylinder probably needs changing.

If so then I'd also be seriously considering replacement of the clutch and the rear crankshaft oil seal while the gearbox is off. Even if these are currently in OK condition, if they're 'getting on' then they could start failing at any time, might as well replace as preventative maintenance to avoid having to remove the gearbox again in the possibly not too distant future...

This doesn't mean that the problem is definitely the hydraulics, there could be some problem with the clutch itself, or the release bearing / mechanism. And of course, make sure the system hasn't 'sucked air' due to low fluid level in the reservoir...

Regards,
John.

 

[Brady_Hill27]

I did a little test this morning: I put the car in 2nd (it didn't want to go in 1st) and held the clutch all the way down and accelerated, at about 2500 rpm the car started to move forward slowly. I also tried pumping the clutch for awhile but had no results.

 

[justin.graeff]

To me, it sounds like there is not enough pressure to fully engage the clutch. That happened to me once when my slave cylinder was leaking. After a while I wasn't able to shift easily in lower gears, then eventually in all gears. While the pedal still had resistance to it, it wasn't allowing enough for me to shift gears.

Got the slave and lines replaced and things went back to normal.

Fast forward to this weekend (about 2,000 miles) and the same thing happens. After a night of letting it sit, clutch goes back to normal. Thinking it's the master this time around. It's at the shop being determined as we speak.

I think with the clutch system, it's hard to diagnose unless we can physically see a leak coming from somewhere. If the master leaks, it can happen internally which can cause problems too (which is what I think is happening on mine). It takes a better mine than I to figure those issues out.

 

[unclemiltie]

If so then I'd also be seriously considering replacement of the clutch and the rear crankshaft oil seal while the gearbox is off. Even if these are currently in OK condition, if they're 'getting on' then they could start failing at any time, might as well replace as preventative maintenance to avoid having to remove the gearbox again in the possibly not too distant future...

This is good advice. I'll add that the differential seals and the shift seal are also both cheap and much, much, much easier to change with the transmission out of the car.

So if it were me the list would include:

Clutch, clutch cover and slave cylinder
rear engine main seal
transmission shift seal
both differential seals

And depending on if you take the engine out with transmission from the top or the transmission out the bottom I would:

Top: change the timing/balance chains. Much harder to do with the engine in the car and only a $200 kit from eeuroparts

Bottom: change the 6 subframe bushings. Roughly $10 each and the subframe has to come out to do it, but it's already out to take the trans out the bottom.

 

[JohnLear]

To me, it sounds like there is not enough pressure to fully engage the clutch.

There is not enough pressure to dis-engage the clutch. That's the problem, the clutch is remaining partially engaged when the pedal is depressed. This causes rotation of the input shaft when it's supposed to be stationary (when selecting first at a standstill). The synchromesh isn't strong enough to force the input shaft to stop rotating when the driven plate is dragging on the flywheel and pressure plate faces. Keep in mind that the main shaft is directly connected to the wheels, which are stationary when the car is at a standstill, so it's the input shaft that has to be prevented from rotating to achieve an easy shift from neutral into first gear, which is harder to do the more the clutch plates are dragging.

As a result of the dragging clutch the speed differential between the male synchromesh cone and the corresponding female cone is too great (one rotating, the other not rotating) and the 'blocking ring' will refuse to allow the dog teeth to engage easily. Significant force will be need to be applied at the selector fork to press the conical synchromesh faces hard enough against each other in order to create enough friction to overcome the clutch drag and bring the input shaft to a standstill (or close enough to it, but even if the input shaft is still moving slowly enough to achieve gear selection, if it's still a bit too fast the dog teeth may grind against each other).

That happened to me once when my slave cylinder was leaking. After a while I wasn't able to shift easily in lower gears, then eventually in all gears. While the pedal still had resistance to it, it wasn't allowing enough for me to shift gears.

Because eventually (with enough compressible air in the system) the clutch hardly disengages at all, and the synchromesh can't synchronise the dog teeth, regardless of which gear you're trying to select. When this happens the 'blocking ring' will prevent gear selection, or allow it but with dog teeth grinding against each other (depending on how strong the synchromesh action is).

Note that with practice it is possible to change gears without the clutch disengaged. This requires correctly timing the shift so that you're only attempting to get the dog teeth to engage when they are at similar enough rotational speeds to engage without the synchromesh interfering with the shift action.

When shifting from a lower gear to a higher gear you need to move the shift lever so that the dog teeth disengage just as you lift off the throttle pedal and rotational torque loading is momentarily removed from the teeth (if not then the teeth will refuse to disengage), then wait momentarily in neutral as the rpm drop before then selecting the next higher gear (at just the right moment). The same applies for selecting a lower gear, but in neutral you need to give the throttle a 'blip' to bring the rpm up (not down). With practice you can achieve fairly clean shifts in this way, if you have a lot of talent then you can achieve consistently perfect gear changes. It is a bit tricky, timing is everything. I can do it, though I'm not great at it.

But even if you can do this, if the clutch isn't disengaging, you're still going to be struggling getting it into first gear...

Regards,
John.

 

[BikePilot]

Yep you have a leak and/or air in the system. Do not drive the car as it is, it will wear the clutch very quickly AND will destroy the syncros in the gearbox.

If it's not the slave cyl it'll be an easy fix. If it's the slave it'll be straightforward, but a bit of man-hours to turn all the wrenches.

So here's how I'd proceed: check fluid level in MC, is it low? If yes, find leak and fix. If no, inspect clutch line, any bulges? If no, replace clutch MC.

FWIW although it seems that the slave should fail most often I've had MCs fail more often than slaves. My wife's civic and my porsche have both never had a slave fail and both had MCs fail. No idea why, maybe they aren't as rugged.

 

[Brady_Hill27]

Thanks for all the advice. I'll order the MC and try that out, let you guys know how that goes. I would also like to bleed the fluid just in case there is air in the line. How do I do this?

 

[unclemiltie]

There's a bleeder right on top of the transmission where the hose goes in.

Get a piece of clear hose, put it on the bleeder and put it into a jar.

Get a helper, open the bleeder and ask them to push the clutch and hold it. Close the bleeder, have the helper release the clutch. Repeat until there is no more air coming out, you can see it in the hose.

make sure the reservoir doesn't go empty while you're doing it

 

[JohnLear]

Do not drive the car as it is, it will wear the clutch very quickly AND will destroy the syncros in the gearbox.

I'm not convinced that this problem will add (much) to clutch wear (the forces and friction involved when attempting to select a gear are substantially less than occur when the clutch is engaging in order to move the car from rest), but totally agree that it's very hard on the synchromesh.

FWIW although it seems that the slave should fail most often I've had MCs fail more often than slaves. My wife's civic and my porsche have both never had a slave fail and both had MCs fail. No idea why, maybe they aren't as rugged.

It might be to do with push rod angles etc. on particular cars? Bigger angles as the piston is 'stroked' might cause greater side thrusts within the cylinder? Or, maybe if the master cylinder is a smaller ID than the slave cylinder then a greater stroke will be required at the master cylinder than at the slave cylinder? (greater stroke equalling more wear on the seal).

My experience differs, it's more often been the slave cylinder that has caused me grief (though my statistical sample size isn't huge...). It's not unusual to fix the slave cylinder only to have the master cylinder fail not too long afterward.

Regards,
John.

 

[justin.graeff]

There is not enough pressure to dis-engage the clutch. That's the problem, the clutch is remaining partially engaged when the pedal is depressed. This causes rotation of the input shaft when it's supposed to be stationary (when selecting first at a standstill). The synchromesh isn't strong enough to force the input shaft to stop rotating when the driven plate is dragging on the flywheel and pressure plate faces. Keep in mind that the main shaft is directly connected to the wheels, which are stationary when the car is at a standstill, so it's the input shaft that has to be prevented from rotating to achieve an easy shift from neutral into first gear, which is harder to do the more the clutch plates are dragging.

Thank you for straightening that up for me. I had them backwards. Also, thank you for explaining it better than I ever could! Even though some of it is over my head.

 

[BikePilot]

John that may well explain the differences. My experience is also not all that statistically significant.

 

[unclemiltie]

John

I would think that if the clutch is partially disengaged AND the transmission is solidly locked into some gear while the car is not moving that would cause a good bit of clutch wear and would not damage the syncros since they are no longer really involved in the drive chain.

When the clutch is partially engaged and you're trying to engaged one of the gears that will cause significant wear on the cones on the synchronizers.


It's been a long time since I took a transmission apart, and never one of "our" transmissions so they may be difference but the syncros are there to match the speeds of the gears during shifting by dragging a bit on the cones. But once the gear is engaged there is no speed difference and the syncros just sit there and rotate with the rest of the assembly.

So the point of this is that you can damage both the clutch and the synchronizers by driving the car a lot in this condition, each at different times of the driving process.

 

[JohnLear]

John

I would think that if the clutch is partially disengaged AND the transmission is solidly locked into some gear while the car is not moving that would cause a good bit of clutch wear and would not damage the syncros since they are no longer really involved in the drive chain.

I agree that as described that situation (i.e. with the gear already selected with the car stationary) won't affect the synchromesh components because when the dog teeth are already engaged the synchromesh won't be doing anything.

I also agree that this specific situation would also be causing the partially engaged clutch plates to be rubbing as the flywheel and pressure plate rotates and the driven plate doesn't. This will cause clutch wear, but I doubt it will be a lot of wear, mostly because the clutch surfaces are not being forcefully pushed against each other (just rubbing with a relatively light force). If the clutch is failing to disengage more than to a fairly small degree, then you'll have zero chance of selecting first gear at rest, the baulk rings will prevent selection if the rotational speeds are too disparate (you'd have a far better chance with no synchromesh at all, just force it in and grit your teeth for the associated grinding noises...).

I would expect that each time the car normally moves from rest the forces and resultant wear (as the clutch engages) would be substantially greater than the wear caused by failure of the clutch to fully disengage, considering the relatively light clutch loading and the relatively short time it takes to force the dog teeth to engage when the clutch is dragging only enough to inhibit gear selection (but not to completely prevent it).

The situation as described would be unusual, since you don't often sit for any significant period with the car in gear and the clutch pedal depressed. The problem arises when trying to select a gear, which even if a bit difficult will take a relatively short period of time, or fail to happen...

I think this might be related to what happens when you see racing cars 'creep' on the start line. My understanding is that racing clutches are often set up with minimal clearances and some rubbing 'normally' occurs between the plates when the clutch is 'disengaged' (i.e. some racing clutches don't 100% disengage). As a result, if this continues for too long (say the start lights take a longer than usual time to go out) the driven and pressure plates may get a bit hot and the friction material and pressure plate metal may expand significantly enough that the clearance gets even less and the clutch starts to drag with greater force, to the point that the driver has trouble preventing the car rolling forward...??? This affect may well be different for a racing clutch than for a 'road' clutch because a road clutch tends to grip less well as it gets hotter, whereas a racing clutch tends to grip better when it gets hotter, a function of the friction material used in different applications.

When the clutch is partially engaged and you're trying to engaged one of the gears that will cause significant wear on the cones on the synchronizers.

Yes, because the diameter of the synchromesh rings etc. are small compared to the diameter of the clutch plates, and the synchromesh is lubricated and supposed to partially slip, while the clutch (usually) isn't lubricated and is supposed to mostly grip. The synchromesh can exert far less force against the clutch than the clutch can against the synchromesh, so the synchromesh has to work very hard to overcome a dragging clutch.

In normal usage with a correctly disengaging clutch the synchromesh only has to deal with the forces involved in the rotational inertia of the gears and shafts, and that of the clutch driven plate. This is an order of magnitude less than the forces that may be involved with engine torque transferred through a dragging clutch (dependant on how badly it's dragging).

It's been a long time since I took a transmission apart, and never one of "our" transmissions so they may be difference but the syncros are there to match the speeds of the gears during shifting by dragging a bit on the cones. But once the gear is engaged there is no speed difference and the syncros just sit there and rotate with the rest of the assembly.

Yes.

So the point of this is that you can damage both the clutch and the synchronizers by driving the car a lot in this condition, each at different times of the driving process.

Yes, and no. At least (IMO) not much damage will be done to the clutch, unless the situation continued for a very long time, but I really think the synchromesh would die a long time before serious clutch wear occurred. This doesn't mean that this sort of (I think minor) abuse might not tip an already worn clutch over the edge.

Regards,
John.

 

[Brady_Hill27]

I wanted to see how the car felt to drive so i took it around the block babying it. When i would accelerate after shifting gears (up or down) i would feel the car holding back, requiring more time to get to speed. It feels to me (from a newb's perspective) that the clutch isn't engaging, or disengaging all the way. I think this thread has more than enough information to get me back on the road though, and I thank everyone who contributed to sourcing information. Saab owners are the best kind of people haha. Thanks all

 

[JohnLear]

When i would accelerate after shifting gears (up or down) i would feel the car holding back, requiring more time to get to speed. It feels to me (from a newb's perspective) that the clutch isn't engaging, or disengaging all the way.

With the car in gear, the only clutch associated thing that that would be inhibiting acceleration would be if the clutch were slipping under load. When this occurs the rpm will rise abnormally relative to the road speed. You should be able to both hear this and see it on the tachometer. It's usually more evident in higher gears at / near maximum torque output (which is typically when it's first noticed, but it will tend to get worse and worse until the clutch slips even with light throttle in any gear).

There are two possibilities that would cause the clutch to slip, most commonly the clutch is worn out, or, there is a blockage in the hydraulic system. Blockage can be caused by internal delamination of the rubber hydraulic hose, the internal layer coming adrift at some point inside the hose. A third possibility might be that the clutch release mechanism is partially stuck, i.e. the release bearing isn't freely sliding on it's shaft.

If this occurs (partial hose blockage) it can prevent the fluid freely flowing when the pedal is depressed, but you will feel more than usual pedal resistance. The fluid could also flow freely in one direction and then fail to flow freely in the return direction (the blockage acting like a one-way valve, i.e. the delaminated rubber acting like a heart valve that flows freely one way and closes the other way).

If so then the pedal may feel more or less normal and the clutch will disengage when the pedal is pushed, but then the slave cylinder pressure won't fully release when the pedal is released, resulting in residual pressure in the slave cylinder and failure for the clutch to fully engage, with resultant slipping under load. A restriction caused by internal hose delamination could also affect fluid flow in both directions. This will usually be a short lived thing, in that given enough time the pressure will escape the slave cylinder and the clutch should eventually fully engage.

I had a similar issue with one of the brake hoses on my Accord. It caused unequal brake force wheel to wheel at the front axle line (pressure less at caliper A than at caliper B), then for caliper A to drag once the pedal was released. The hose looked perfect externally, but was far from it internally.

Regards,
John.

 

[Brady_Hill27]

I replaced the Master Cylinder with depressing results. Installation went well, as well as the removal. However after all that work bleeding and replacing, I finished just how i started, with a half-working-clutch. I have neither the resources nor the know-how to replace the slave cylinder so I guess I'm going to have to resort to the dreaded mechanic shop. I appreciate all the help you guys have been with this dilemma and I know when something else goes wrong with my car the guys at saabcentral will help get me through. Thanks to all and have a good day.