Clutch Bleeding

[georryan]

So I just rebuilt my Clutch master cylinder, and I'm pretty sure I've replaced the slave cylinder with in the last couple years.

I bled the clutch and drove it around. It wasn't bled 100%, and after a couple days of carefully driving it, I got home and couldn't shift gears. I had to turn the car off to get it in reverse. Although, once I had it parked on a hill, with the nose pointing up, it shifted into gear just fine with the engine running.

I can't seem to find any leaks anywhere, so I'm pretty sure I just didn't get it bled well enough.

My slave cylinder still has the metal line that makes a U kind of shape as it comes up out of the slave , makes a U-turn and heads down below the slave to a bracket, then connects to the SS line I replaced the rubber line with (I also removed the dampener).

Given that U shaped metal line path is in there, will that make it harder or impossible to fully bleed the clutch without disconnecting it from the tranny, or do I just need to keep working at it? Last time I went through two or three full clutch master cylinder containers worth of fluid. It seems to me that if I keep bleeding it the clutch fluid should push all that air down into the slave cylinder eventually.

I'd rather not be forced to remove the slave cylinder to get it lower to the ground, because I'm starting to think that will require removing the intercooler. Its not the end of the world, but since it is an AWIC it just takes some extra time and work.

Hmm...another thought, maybe I shouldn't have the front end of the car jacked up when I do the bleeding. Maybe it is allowing air to stay trapped in the slave cylinder because of the angle of the front of the car sitting on jack stands?

[vrg3]

Bleeding the clutch is a pain even without that metal line.

In my experience, you can get a lot of the air bubbles out of the clutch system by grabbing the top of clutch fork and firmly and quickly pushing it towards the back of the car, forcing the piston of the slave cylinder in. If you do this with the cap off the master cylinder, you'll probably see some bubbles rise up. Try seeing if that gets you anywhere.

[georryan]

So I just found this link. I'm going to try it out. It is similar to what Josh and Vrg3 have mentioned. I'll let you know how it works.

Bleeding Steps

The trick here is the last few steps, where instead of just pushing in the clutch fork, you then hold it. While holding it, you open the bleeder screw and have the person in the car slowly push the pedal to the floor. Once it is at the floor, close the bleeder screw, retract the pedal, top off the fluid and try it again a couple more times.

The process is there to purge the slave cylinder of any air, which is what I think my problem is.

[vrg3]

Sounds like a great idea to me. It's kind of like the standard bleeding procedure except you're pushing from both ends. Let us know if it does the trick!

[georryan]

So I'm pretty sure this method works real well. Make sure when you are doing the second part, where you hold the clutch fork in while a buddy pushes the pedal down with the bleeder screw open, that your buddy pushes the pedal slowly. There will be resistance on the clutch fork as well as fluid and air leaving through the bleeder screw.

My problem is this, and maybe it is a simple answer, but when the car was turned off, the pedal had lots of pressure and resistance and felt real good. Once I turned the car on and stepped on the pedal it seemed to loose some of its stiffness, and didn't seem to disengage the clutch all the way. Turn the car off and the pedal felt stiffer slightly. Turn the car on and it felt a little weaker. Does that sound like I just still have air in the lines? It would seem so to me, but maybe a bad slave cylinder? I don't see any leaks currently.

I lost my pedal monkey for the day so I'll have to bleed it some more later.

[Legacy777]

The clutch pedal feel should not really change at all depending on if the car is on or off. When you put the clutch MC back together did you adjust the cleavis pin so that there is not pressure on the clutch when the pedal is not depressed at all?

When I did my swap I had to play with things a little bit to get the feel and engagement just right, and so the MC was not leaving a little bit of residual pressure on the system.

[vrg3]

Yeah, nothing in the clutch hydraulic system really depends on the engine running or not running... When you say you noticed this change, was it immediately after starting or stopping the engine? Or was it after it warmed up and cooled down?

If it was the latter, there may still be air in the lines, yes -- it can expand with heat and change pedal feel. Or, like you guess, the seals in the slave cylinder may be failing when hot.

[georryan]

How much space is there usually between the slave cylinder pin and the clutch fork? It seemed to me like the slave cylinder pin was recessed too far from the pin by default. It almost felt like the clutch pedal didn't engage till close to the floor and wouldn't fully disengage the clutch at the full pedal travel. I figured this was an air in the lines thing, but after pumping two full reservoir loads of fluid through the lines, I wasn't sure. I did manage to get more air out doing the secondary bleed technique mentioned in the above post, but after a couple times I was just seeing fluid.

So what I did was to preload the MC enough to have the slave cylinder pin just a fraction off the clutch fork. That made a big difference, but since my clutch fork spring broke (should be getting a new one in the mail today), there is some chatter at that point from clutch fork with the clutch "fully" disengaged. What I noticed after doing that, is that when it starts to chatter, I can put my foot under the clutch pedal and pull back on it slightly, the chatter disappears, but the next time I push the clutch in, it will have lost most of its pressure, and not get it back until I'm closer to the floor and it won't have fully disengaged the clutch. After a few tries of pushing the clutch in and out it will usually be better. I suspect air in the lines, but the odd part is once in a while when I push it in, I'll hear like a clink or something like a clink, and then the pedal will be real stiff and the engagement point seems to be right in the middle to further up on the pedal travel. Also, the chatter/vibration from the clutch fork disappears.

I want to see how it feels when I get the spring back in there. It almost seems like the MC isn't being compressed enough with the default clutch pedal travel without being pre-loaded a bit, or almost like the slave cylinder is too far away from the clutch fork and too much travel is taken up pushing the pin to the fork.

There isn't a lot points of failure in this system, so it can't be too many different things. You'd think if you had a faulty slave it would be leaking somewhere?

I'm going to spend some more time bleeding it, and if I can't seem to get it to feel better with the MC not preloaded I think I'll replace the slave cylinder.

Its really odd. I don't remember having this problem the last two or three times I've done clutch bleeding in either this car or my brother's car.

[georryan]

Also, what I mean by the engine running or not running...with the car off, the pedal felt good and fully pressurized. It was stiff and consistent all the way through the travel.

I'd put the clutch in, and turn the car on, and with it in neutral, I'd let it out, then put it back in, and it lost a lot of its firmness. I'd turn the car off, and it would stiffen back up again. It didn't really have to deal with any heat in the lines or anything, I think it was the first time I started the car up that day.

[Legacy777]

The clutch return spring keeps the clutch fork at the position so there is no extra clearance when depressing the clutch pedal. As you drive, the clutch fork may move around and cause that excess travel. I'd suggest getting the spring back in, and then drive/test to see if that helps at all.

[georryan]

Ok, so I bleed the clutch some more tonight. I never saw any air coming out of the SC. The pedal felt good and stiff. I also adjusted the clevis pin so that it lined up with the clutch pedal when they were disconnected and without any pressure to the clutch system (no preloading on the MC). Finally I installed the clutch fork return spring that arrived in the mail today. I started up the car with the clutch in and everything shifted fine. I released the clutch and there was still a bunch of chattering coming from the clutch fork. I went to put in the clutch and it lost alot of its firmness and sank almost to the floor. It did return, but not with the same firmness I would expect it to have. Not only that, but with the clutch all the way in, I could not shift gears.

Here is the catch. Without doing anymore bleeding, just playing with the clutch pedal by depressing it and releasing it, once in a while as I was pushing it in I'd hear a click kind of noise and the pedal would have a firmer feeling. At that point I would also be able to shift gears, and if I slowly let out the clutch the chatter I was hearing from the clutch fork would be pretty much gone. If I was careful and slow, I could push the clutch in and release it and maintain the same feel and result, but other times it would then slip back into its loose/chattery mode.

When the clutch fork was chattery, it would be vibrating and the return spring would also be vibrating. When the chattering was down to a minimum the clutch fork looked to be still and the spring wasn't moving, either, even though the car was running (the car being in neutral with the clutch out).

I'm starting to think it isn't my MC or my SC but maybe my clutch fork/throw-out bearing. I'm not relishing the thought of pulling the engine but does it sound like that is the right track to you guys?

[vrg3]

Yeah, you're probably right about the slave cylinder being okay. I don't know whether or not there are failure modes for the slave cylinder that don't involve external leaks, but I can't really visualize one.

Are you pretty confident in the rebuilt master cylinder?

Can you get someone to hold the clutch pedal down while you watch the clutch fork? I'm curious about whether the fork would slowly move back even while the pedal was pressed.

The chattering of the clutch fork could be a result of the throwout bearing not sitting right on the transmission snout. You didn't actually take your engine and transmission apart in this process though, right? Did you remove the damper? Maybe the chattering was happening before but the damper was concealing it. If there is a snout issue, a snout sleeve kit might be the ticket. It seems plausible that bad enough snout wear could make the clutch fail to disengage sometimes, except when it was all clicked into place.

[Legacy777]

Yeah I'd be suspicious of something with the clutch pressure plate, throw out bearing, or something inside there.

How many miles are on the clutch?

[georryan]

I'll get someone to play with the clutch while I watch it from the outside within a few days. I was bleeding it last night by myself with the old "long bar to hold the clutch in" trick.

The clutch is fairly new. Less than 50k miles, probably. I installed the clutch (exedy stage 1) and the lightweight flywheel at the same time a few years ago, but I haven't driven the car much the last couple years.

As far as the rebuilt master...there isn't much to it. I had a brand new spring/seal sitting in a box because my clutch had started to feel inconsistent a little over a year ago, and I had replaced some other parts without making it much better. So that was my next step. Its basically a spring with a seal on the end of it that you just slide inside the master. Then there is the rod with a plate and a snap ring. At this point I'm actually questioning if my master was ever bad to begin with, but that doesn't really matter at this point.

The part that gets me is that this only happens when the car is running and after the car has been full engaged in neutral. So its after the parts sync up and start rotating. If I turn the car on with the clutch in, and then turn it off I don't think it would make a difference. The pedal feels fine with the car off. Its goes bad after the clutch has been let out in neutral with the car running. I'm wondering if the throw out bearing isn't snapped into the pressure plate well enough...although I tried pushing it with my hand to see if I could make it snap in place or anything, but it didn't seem to help.

Vrg3, no I didn't pull the tranny or engine for this procedure. I rebuilt the engine and had it stroked to a 2.3 probably 2 or 3 years ago. The damper was removed a long time ago, so that probably isn't an issue, although I suppose it could have been hiding the problem, although I don't remember this chattering before. I think that is new since this clutch rebuild. Before doing that, the worst that would happen was the clutch just wouldn't fully disengage and that was it. It felt like it couldn't keep pressure in the line, but other times it would be fine enough.

Is it possible that the clutch fork is not sitting straight, but slightly tweaked and rotated to the side? I don't remember it being able to turn much or shift around much on the tranny snout, but maybe?

[georryan]

I forgot, the fork is held in by a pin...so it can't twist, only rotate along its allowed axis. So more than likely it is the throw out bearing. Maybe it isn't clipped in well to the pressure plate or no longer has a strong connection with it. Oddly enough, before I did all this, the pedal was having the same problems but I don't remember any chattering like I am getting now.

[vrg3]

Yeah, it does sound like maybe your master cylinder was okay to start with.

The reason I ask about the snout is that it's made of some pretty soft aluminum, and it's not uncommon for it to get grooved up over time.

It doesn't take much effort to snap the throwout bearing into the pressure plate. And the turbo clutch fork is a pretty solid piece of cast iron, and it sits on a spindle rather than on a ball, so it's hard for me to imagine that it's bent or not sitting right, but anything is possible.

[georryan]

Maybe the retaining clip in the pressure plate?

[vrg3]

I suppose anything's possible. I've never heard of that happening though. The clip isn't even a replaceable part, I don't think. I would expect that if the throwout bearing and pressure plate weren't clipped together right, then pressing on the pedal would just pop the bearing off of the plate.

If you can't find a problem in the hydraulics, I guess you'll have to pop the engine and transmission apart and take a look.

You can probably do this without removing either the engine or transmission. Pull the radiator, unbolt the motor mounts and transmission mount, lift the motor and transmission using a trolley jack under the oil pan and another jack under the transmission, undo the bellhousing bolts, and roll the engine forward a few inches. I've done clutch jobs in this way before.

[georryan]

Hm...Thanks for the input guys. When I can get a good look at analyzing the hydraulics (hopefully tonight), I'll give an update.

[georryan]

Ok, so after watching my clutch monkey work the clutch with the car off and on, it appears to me that the slave cylinder is working fine. It doesn't seem to be pushing the fork any less or more in either scenario. The other thing I noticed was a distinctive "ting" sound when the clutch was released from inside the bell housing. I actually heard it both with the car running and not running. I wasn't able to get the the whole assembly working correctly with the car running this time, so I couldn't really compare or listen for anything in that situation.

I couldn't pull the rubber off the clutch fork peek hole, because my intercooler sits above it and I wanted to view it while the car was running. Either way, I think I'm going to have to pull the sucker apart to figure out what is going on.

My best guess is that the throw out bearing is not being held into the pressure plate well, and when the car is running it slips out easier. With everything stopped and no rotation going on, it manages to hang on a little better and I get the resistance back in the pedal.

[vrg3]

Blergh. Good luck.

'Ting' sound? Weird. It kind of jives with my worn snout theory though. You might consider spending the money on a snout sleeve kit even if you don't know if you absolutely need it. They really do make the clutch pedal a lot smoother than stock, even if the snout's not in terrible shape.

[georryan]

Snout sleeve, huh? I'll do that. Might as well, right? You think a bad snout would cause the clutch to not fully disengage, or do you think it is just one of a couple problems that are going on?

[vrg3]

I think if it's grooved badly enough it could cause the throwout bearing to bind. You'd feel that at the pedal too. I don't feel much certainty that your problem is the snout -- I'm just kind of grasping at straws here. And, like I said, I feel like the snout kit's a great idea -- why not replace a soft aluminum bearing surface with hardened steel?

[georryan]

True. Plus, if the throw out bearing is the problem..maybe a new one will fix it, and the new snout comes with a new throw out bearing. I'm going to pull my intercooler and see about getting a look down that observation hole the fork comes out of.

The cheapest price I could find a snout from came from subaruwrxparts.com. For reference in case anyone else is looking for a good price on the product. If you can ship it to your office as well, you'll save 2 bucks on shipping over a residence.

I found a post where Josh was talking about the problems he was having with his, but that was due to an amalgamation of subaru gear. The site I linked states the part is for a 92-94. I don't know if it doesn't fit a 91 or not, but I believe my current tranny is out of a 94, so I should be in good shape.

[Legacy777]

Yeah, it sounds like you'll need to split the engine and trans to figure out what the issue is. I guess the thing I'll note is that the main shaft on the trans always rotates, even if the trans is in neutral. The only time it does not rotate is when you depress the clutch in. When you split the engine & trans I would suggest checking the main shaft for any play, as well as checking the trans snout.

From my experience with messing around with the throw out bearing, the clip that holds it to the pressure plate allows the throw out bearing to come out about 1/8" - 1/4" or so before it starts disengaging pressure on the clutch. As is discussed above, that clip may be binding or bad.

Regarding the clutch sleeve, after thinking about it, I believe I got a used WRX clutch fork,
http://surrealmirage.com/subaru/images/ ... chfork.jpg

If the pivot ratio on the clutch fork is slightly different than the original turbo legacy, then that might be why I had issues with the sleeve. I'm going to do a little bit of homework on this, now that I have a plausible theory....

When you take apart the clutch, can you take a measurement of the length on the trans snout. I'd be curious to know how long it is.

Here are some pictures of the sleeve installed on my trans.
http://www.main.experiencetherave.com/s ... s/ej22t/29