Hey all,

I'm at the point where the hydraulic clutch setup needs to be mapped out for my Subie conversion. The donor trans is from a 2006 WRX with a 13/16 slave cylinder. Initial reaction is to buy a 13/16 generic (wildwood or CNC) master cylinder so they match and give a “factory” clutch feel but after some internet searches it seems that pedal travel is also a factor and might mean a different size would be better. I’ve seen some GREAT hydraulic clutch threads but could not find one for the Subie trans setup.

I’d assume the pedal throw is probably similar but could be wrong. BIGKAT is using a 3/4 and mentioned he might rather have a 5/8. His trans likely has a ¾ slave.

Any insight or thoughts on this? You don't have to be right! Brainstorming is great too.

Matching the existing size is the rule of thumb.

If you had a 3/4 slave and push it with a 5/8, it would take longer throw of the pedal for full motion of the slave.

If you have a 5/8 slave and push it with 3/4, it would take less throw of the pedal.

This doesn't take into account of pedal force (foot pressure) that it take to push the larger master.

The good news is that Wilwood masters are cheap. Around $50. So you could experiment without breaking the bank.

Thanks Mike! Think I will start with a 13/16 and see how it goes... hope your project is going well.

Be sure to check what the stock Subaru clutch master cylinder sizing is, and then match it. I know my Suby master cylinder inner diameter was different than the Suby slave cylinder inner diameter. I ended up using a Honda master cylinder that has the same ID as the stock Suby one (because the Honda slave worked better for my design -- essentially because the Honda Civic clutch master has a remote reservoir) and used the stock Suby slave cylinder on the Suby trans. YMMV...

Thank you. I'll check that for sure. I've owned a boat load of Civics (2 CRX, 4 hatch)...would be nice to have a piece of one in the 914...

Turns out the clutch MC is mismatched...I'm seeing mostly 11/16 as the bore....hmm....problem is there isn't anything available in that bore from wilwood or CNC. When I search ebay NOTHING comes up for cars, so the next best thing is probably a 3/4. Maybe I'll try that and if it catches too high it can be changed out pretty cheaply as Mike said. I'll post the results when I get the parts in....it's getting closer now.

This is a timely question as I am in the same place. I have the Subie slave already. What Honda MC did you use?

There is more to it that just matching the cylinders. If you are doing a direct cable to hydraulic swap on a 914 system (not Subaru) and rotate the push pin on the pedal all things remain the same and 1:1 sizing is what you should use. However, that is not the case when switching from another car. The Subaru pedals are longer/shorter, the pivot point to push rod is different and overall leverage is different. Bottom line, it's not that simple and some calculations are in order.

You can achieve the same goals by changing hydraulic cylinder size as well as mechanics leverage. If it were me, I would measure the full stroke of the slave, calculate the volume based on diameter\stroke

Next I would measure the full stroke DISTANCE of the 914 clutch pedal PUSH ROD rod and using that stroke distance, calculate what diameter equates to the volume needed to fully actuate the slave. This takes into account both mechanical and hydraulic motion and leverage.

Hope I worded this simple enough....

Whoa... I think I GET IT. And if I get it...everyone else will probably get it... (I'm slow ) Great work...

First, thank you Paul for the very thoughtful response...should help everyone that is in the same boat. So the question I have is do we want the "stroke"of the pedal (volume of fluid pushed) to equal the stroke of the slave cylinder ideally so they both move at exactly the same rate and the pedal bottoms at the same time the slave cylinder does or does the master just need to push AT LEAST enough to get the slave fully depressed? I'm thinking if the master was too strong the clutch would engage at an awkward time...too low maybe.

For years I had a clutch problem on my Subaru conversion, the car would grind the
gears when shifting. The problem turn out to be that the clutch master cylinder did
not have enough travel when the clutch pedal was press. The problem turn out to be I had to move the fulcrum point on the clutch pedal to the cylinder and add a 6 degree shim under the pedal to get full travel. Now it works great. Make sure that you measure the travel at the clutch fork with the clutch installed you will be surprise that under pressure things change!

If it were me, I would probably crawl under a working Subaru and measure the amount of actual movement of the slave at full pedal stroke and try to duplicate that movement on the 914....it's the safe way and I hate to do things twice....I'm willing to bet that the 914 pedal (at your foot) moves much more than the Subaru pedal so keep that in mind so the actuation is not painstakingly slow.

All this said, I would try to stick to a generic brand M/C like a CNC so that if you decide later to change sizes it's a simple and cheap change (they look identical on the outside) and you don't have to remake all the brackets.

Good luck

https://www.youtube.com/watch?v=bvrjuLo5MFE


Here's what it looks like...not mine unfortunately...found it on youtube.

Knowing the Subie slave travel (in order to calculate the volume) is generally a safe approach, though what you really need to know is how much travel is required by the slave to dis-engage the clutch. Also, I suspect the subie slave is forward sprung, meaning that the slave has an internal spring that keeps the piston in its most forward position at all times and keeps the TOB in constant contact with the PP fingers. With this type of slave arrangement, the volume of fluid required to dis-engage the clutch is kept to a minimum. So my advice would be to determine if the Subie slave is internally sprung forward first, followed by how much travel of the TOB/PP fingers is required to dis-engage the clutch in order to get the correct fluid volume required. Some of the travel to dis-engage is dependent on the clutch configuration (I'm not familiar with Subie stuff), but the travel requirements for a clutch disc that has a Marcel spring design will be different than one that is a puck design or bronze/ceramic design. My appologies if I've done nothing but added to the confusion :-}

When I was configuring my clutch system designed around QuarterMaster comonents (not for a Subie), they were able to provide me with all the tech info I could have ever hoped for.

Andys

Thanks andys and 904 SVO for weighing in....I'm not sure I'm experienced enough to tackle this....wonder if I should just back date to a push clutch cable...

Eh what the heck I'm going for it. Getting a .75 wilwood master. Should I get the master with the clevis fork or no fork? I think no fork but that's just a guess based on looking at PRS914-6's thread. And where's the best place to get the hydraulic line?

Let us know how it works.