odonnell

I haven't gotten that far on the Subaru yet - a timing belt, water pump, rich condition, messed up VVT, and misfires in high vacuum have all popped up. It's a 1-owner 2010 XT with 116k and full dealership service history /rant But it does need routine brake service soon.


Going back to the 944, I forgot about that check valve on the booster. On bimmers they're integral to the venturi (or suction jet pump, as they call them). I remember getting a hiss when I unhook the vacuum line going to the "reservoir" by the battery tray a few hours after turning off the car - would this be a sufficient test of that valve? It's been a few months so I'll need to check again. To me, that would indicate that air can only flow out of the booster circuit.

porscharu

Did you do the timing belt on the subaru? I would set it to TDC and check your timing marks on all four cam gears. You will need a mirror for the lowers.

V2Rocket

try to cap the vac line coming out of the booster so that the only hose is the booster-to-venturi line...run the car at idle, how are the brakes?

if your booster is original you got your money's worth for that diaphragm.

odonnell

Yes I did, I made 100% sure everything is lined up. I used the crank cog marker (I believe it's lined up when all 4 pistons are at 50% stroke). Replaced all rollers and the water pump / thermostat. Made me realize how easy 944 belt jobs are The VVT angle and misfire issues were present before, on the original timing belt. I checked before I removed it, and it was also in time. I didn't rely on locking and reusing the positions of the camshafts when I did the job, I re-aligned them to make sure I was starting fresh. No CELs or anything. Thread 1, Thread 2, Thread 3

It really chaps my ***, because I can't even really drive the 944 while the Subaru gets figured out because of the brakes but, it'll all get sorted sooner or later.

porscharu

Not trying to crap on your thread here, but a lot of the ecu doctors have moved on from sf.org. I can reply over there.

Dougs951S

Maybe you should have bought an E46...it'd probably be more reliable 😂

thomasmryan

Your car detects a foreign object with the Volvo calipers....remove the infection))
Subaru stuff...

My opinion, but subaru uses a dash more phosphates in their branded oil to keep the oil control valves and sensors clean pertaining to the cam timing. Try subaru oil and filters.

Valve covers, the spark plug seals and the cover fasteners are vacuum leaks. At 100k, they're baked and probably pooling oil at the plugs. 1/4" stuff works fine but you may need to guess the torque on at least one fastener.

I assume a manual but some automatics have a 'lifetime' spin on ATF filter under the fender liner.

Subaru love but the piston clearance is super tight for the expansion characteristics of the assembly. Weepy head gasket coating allow the slow poisoning of the oil with coolant which fouls the bearing halves. The motor only has to get hot once, and it's over. Dropping a JDM mill is actually cheaper than doing the head gaskets in the long run.

V2Rocket

...Michael, if you go to do this, you will find that it is necessary (or at least easier) to remove the driver side engine mount nut and jack up that side of the engine 2 inches or so to get all the valve cover bolts off.

Had to do that on my EJ25 DOHC engine. Pulling the covers sure made swapping spark plugs cake, too.

odonnell

Ok gents, I did 3 tests on the booster.

1) Pump pedal with engine off. It gets stiffer every time, eventually only the pedal freeplay is left. Check.

2) With the pedal still pumped up from the first test, start car. Pedal should sink when engine catches. Check.

3) After idling with no brake manipulation, depress the pedal and kill the motor. Pedal shouldn't rise back up.

...mine pops up like nothing happened. Now I'm getting somewhere. My first thought is the check valve - but after the car was off for an hour or so, I pulled the vacuum line going between the check valve and the vacuum 'reservoir' by the battery tray. It hissed and was clearly under vacuum.

That indicates that the check valve between the booster and reservoir is working. But shouldn't there be another check valve between the booster and the rest of the vacuum system? Is it built into the venturi? Maybe the booster is just bad. I would hate to replace it when it could be a simple fitting though.




Subaru: I run Mobil 1 0W-40 European car formula oil and Idemitsu ATF-HP fluid, fwiw. All OEM filters. I got new iridium plugs to put in, and I also read about undoing the motor mounts and jacking it up. That's especially helpful on the turbo models apparently. I'll get there, but the 944 is my main squeeze and I have to get it stopping right first....

Van

Get or borrow a little vacuum pump, like this, and test it for real.

V2Rocket

there is nothing in the venturi aside from an air passageway and a ~1/8"
orifice drilled in a little aluminum thing clamped inside.

no check valve aside from whatever might be inside the little valve thing that plugs into the booster itself.

the vacuum can might be under vacuum just fine but the booster could still be leaking around the diaphragm. it's open to atmospheric pressure on the firewall side, on the engine side is the vacuum chamber and the differential times the diaphragm area makes your pedal assist.

odonnell

Thanks for the info. Planning on borrowing a mity-vac tomorrow. Clarification on the booster leaking around the diaphragm. When I do the "#1 test" above that involves pumping up the brakes with the car off - wouldn't the booster fail to build up pressure if it was leaking? The pedal being stiff except for free play after a few pumps seems to indicate that there's no leaking.

thomasmryan

More testing...

Can you pull the lines and cap off the feeds from the master to eliminate what's upstream and down stream?

Are your flex lines in decent shape? Internal swelling would explain the symptoms.


(My tribeca brakes have always been soft, for 100k miles, but it makes modulation easier on snow and ice.)

Dougs951S

I already spoke to Michael about this but I thought it was worth publicly noting this for future searchers...the fact that the pedal built up pressure with the car off simply indicates that the system has no air and no leaks in the hydraulics. That has nothing to do with the booster leaking or generating assist however. The pedal becoming firm after pumping with the car off just means you are bleeding off the residual vacuum and along with your other tests; it indicates that your vacuum canister and canister check valve are working properly. With all the vacuum bled off from pumping, the normal action of the brake is to become stiff, as now you are feeling only the pedal free play and the resistance felt is the imcompressibility of the fluid in the master cylinder. In short, that test is meaningless in determining if the diaphragm does indeed have a leak.

odonnell

Edit: I videotaped my firewall under hard pedal presses and it didn't move.

Got a MityVac and adapted it to the brake booster. I followed these tests:

http://ssbrakes.com/attachment/84298...0Diagnosis.pdf

In short, it DOES leak, but it seems to be very slight. When I put 15" of vacuum on it (20 would have taken all day with this tiny pump), it drops about 1" every 5 minutes. I repeated this at 5" vacuum and the leak rate was the same.

When I pump it back to 15" and press the brakes, I watched the gauge drop roughly 5" like it should. Pressing it again did the same thing. After the 3rd press, the MityVac read atmospheric.

Finally, I pumped it back up to 15" and held down the brake pedal for 30 seconds. It initially dropped to around 11-12" but then continued to slowly drop. I would estimate that for every 30 seconds I depressed the pedal, it fell by about 1".

So that means it's leaking. But the real question is, would this magnitude of leakage completely wreck power brake assist? It's obviously holding vacuum to some degree, it doesn't quickly drop. I'm having a hard time attributing the poor braking to a slight vacuum leak like this. At idle there should be around around 15-20" of vacuum acting on the booster or whatever is created by the venturi, if that vacuum is higher.


To get a reference, I also tested my 951's booster. It held zero vacuum it was in a fire, so I'm guessing it got destroyed.