avk

Jim: looks like used O-rings simply take shape of the cavity they sit in. Those behind cam supports are pushed by the flange into a chamfered corner of the camshaft bore in the head. The remover tool did tear the seal, but there was an extra one, bought because I was afraid to botch the installation, although not in the way it happened.

Yesterday finished working on a timing belt. A first replacement, both for me and for the 1995 2.2 l engine, at 80000 miles. Here's just some random notes in no particular order. Crankshaft pulley removal: The bolt came out without much fighting but after 11 years, the pulley wouldn't part with the crankshaft. Left it with PB-Blaster overnight, still didn't budge. Had to use a puller. It was in fact assembled from three different 3-jaw pullers, one from Sears and two others, big and small, rented from PepBoys. Only on the Craftsman puller the arbor screw was short enough to fit between the pulley and AC condenser (the radiator was removed which I think is a very good practice), but the hooks on the jaws weren't the right shape. I used the jaws with flat, square-angled hooks from a smaller set, and the connecting links from the third one (which were longer than on the Craftsman). Also had to use a big hose clamp around the jaws to keep them on the pulley, because of the pulley's width and of the angle the jaws were at. Finally, because the crankshaft is drilled for the bolt, I had to insert a suitable object (a 1/4" hex driver extension) for the driving screw to rest against. The removal took a whole afternoon and the next day's morning. The rubber layer on the old pulley already showed some radial cracks even before removal, so I put a new pulley on, as well as the bolt and Woodruff key (thanks SubaruBucks): 12305AA242, 800213030 and 804505060. The rust from crankshaft was cleaned with Naval Jelly, and polishing compound. A new key was inserted with ChannelLock type pliers, with some cardboard to protect the crank. Cam support O-rings: There are O-rings between cylinder heads and bolted-on camshaft thrust supports at the front left and rear right corners of the engine. The support on the left side also serves as a spacer between the head and the timing cover, and holds the camshaft seal and the camshaft sensor. Both of those O-rings, same p/n 806946030, had been leaking (slowly but surely) and were replaced. It may be a good idea to replace the one on the left side just in case when doing the t-belt because of how hard it is to get to. When replacing that O-ring, note that the camshaft journal going through the support has a sharp machined edge on the back (not chamfered like on the seal side), and care shoud be taken that the bore in the support does not get caught on it during removal. Also the dipstick tube will need to be removed, or at least its bracket unscrewed from the cam support. If taking the dipstick tube out of the way, one might want to have new O-rings sealing it to the oil pan. There are two, p/n is 806910170. Timing cover seals: I used all new seals for the timing covers (can update the post with numbers if requested). Shaft seals: From the factory, camshaft seals were not driven to the bottom of their bores but were installed with front sitting at the bottom of chamfered edge. There might be a good reason for this because at the bottom of bore on either side there's a pair of oil return holes, approximately at 5 and 7 o'clock positions. There's no danger of the holes being blocked off because the seal's edge is thin enough, but it's probably best to install the new seals to the original depth. Here's how this observation came about. A sharp-eyed neighbor of mine stopped by and noted that the just installed cam seal on the right was sitting deeper than the old one. I thought about it and decided to redo the seal, since I had an extra one on hand. Then I remembered that to seat the seal square and to the correct depth, an old seal can be sandwiched between the new one and the driving tool. You can feel it when the old seal reaches the end of the chamfer, and then stop. The crankshaft seal does sit at the bottom of its bore, but its face is also even with at the bottom of the chamfer. To remove old seals, I used a Craftsman seal remover that looks like a bent screwdriver with a notch on the side by the tip. Works very well. Oil pump and water pump: The bead of sealant on the oil pump sealing surface should be very thin. It only fills tiny voids between precision milled surfaces, and most of it will squeeze out anyway. One can see this when examining the pump upon removal. If using Permatex Ultra Grey 599 in a tube, recommended by SOA as a substitute for ThreeBond 1215, cut the nozzle at the very first step (1/16") after the tip. With water pump, I had some difficulty keeping the new gasket in position. First I put a pair of bolts through the pump and hang the gasket on them but before I could aim bolts into the holes in the block, they would back out and the gasket would fall off. It's made of very thin, rubber-coated metal and can bend easiy, or get scratched on sharp machined edge on the inward side of the sealing flange. I ended up inserting three wooden dowels into the holes in the block and hanging the gasket and pump on them, then threading bolts into three remaining holes, removing the dowels and threading the other three. One can just hold the gasket in place temporarily with some grease but I'm not sure if grease belongs there. Rememebr the FSM specifies a tightening sequence for the water pump, clockwise beginning at the rightmost (when facing the pump) bolt, in two stages to 7-10 lb-ft. When cleaning the sealing surfaces on the block for both oil and water pumps, beware of the sharp edges mentioned above. They did give me a few cuts and then I had to clean some blood from there, too, but nothing major. The water pump had no noticeable play in the bearing, even compared to the new one. On the bypass hose connected to the pump, the tabs should point towards the rear of the car. If pointing down, they'll interfere with the radiator hose just below (on non-turbo models). If they point elsewhere, the clamp will not be accessible w/o removing the radiator hose. Timing belt idler pulleys: I replaced the tensioner pulley and three idler pulleys (two smooth, same p/n, and one cogged at water pump). There was some axial play felt in the old ones while the new ones had none. Many thanks to everyone who shared their t-belt tips before, here and elsewhere, as well as to Jamie and subarugenuineparts for supplying everything needed for the project.

It is a better material: [/url]http://www.iisrp.com/WebPolymers/06ModFluoroelastomersv1-4.pdf

The brown-colored material is Viton (a DuPont trademark). They made it brown to look diferent from the old type (either nitrile or neoprene rubber).

I have just removed a belt from a 1995 2.2l engine with 80000 miles and visibly, it was in a perfect shape too. But the crankshaft and camshaft seals, which were the old black-colored type, hardened up and were due for replacement, as were O-rings that seal camshaft supports to the heads. I'd say if there are no oil leaks, there's no reason to repalce the belt before 100000 miles, especially if it's true that currently supplied timing belts are all made for 105000 mile replacement interval.

Runnin' rich!

Now this sounds frustrating. If something has to be destroyed, maybe it should be the dipstick handle, as long as the debris does not fall into the tube. Also, the whole dipstick tube assembly can be removed. It's held to the transmission case with a couple of bolts. Maybe someone lost the O-ring and put some silicon there instead?

Has the transmission pan ever been removed? It it has, the dipstick might have been caught between the pan and the case upon reassembly, if it was left in the tube. But I never heard of it happening with this transmission.

The difference between 13/16" and 21 mm is less than 0.4 mm, so it's indeed an exact fit within any reasonable tolerances. Remember you need a sealing washer for that plug. Correction made, 0.4 mm instead of previous 0.04.

I follow the idea of opening the bleed ports before retracting pistons, to push the old fluid out and not back into the system.

With factory pads, the hardware is uncluded, if you mean shims and spring clips. At Midas I'm sure it's a separate charge, if not the biggest one.

If you're intererested in OE parts, both rotors and remanufactured calipers are available from Subaru (if you're sure calipers are needed). I changed rear brake pads last weekend. I had on hand a caliper overhaul kit, which I bought mostly because it has replacement boots for the sliding pins, but it was not needed. The pad and caliper kits were purchased from subarugenuineparts (Jamie a.k.a. subiegal). If you want to go aftermarket, make sure you get a caliper specified for wagon. On Impreza, the pads are the same between wagon and sedan but a wagon caliper has a bigger piston.

I installed the kit on my '95, so the number is correct. It may be discontinued by the manufacturer but probably still in inventory. The dealer should even be able to tell you who has it, if not them. But try the discount places first, in any case. The kit for 1997-2001 will not fit. The cutouts and the brackets are both different.

They probably still sell the kit, p/n H4510FA100.

The pin bore is offset from the center, probably to reduce the shear load. So the chamfered ends should coincide lest crookedness ensues.

Just want to include an example of what one might see when charging with actual two-gauge manifold. Last year, I did it with a minivan. Initial pressure readings were 180 high and 35-40 low. The low-side figure was actually OK for that vehicle (dual evaporators) but since high side was on the low end of the range, I thought cooling can be improved. When adding refrigerant, the pressures changed to 200 and 25-30. This should be typical at least when a system is slightly undercharged: the suction pressure in fact goes down. As to Interdynamics, I would only buy pure refrigerant under their brand (but then you can buy DuPont). They don't write on their cans exactly what's put in. For instance, in kits with oil, the oil is esther, while factory fill in the system is PAG. The only way to find was from MSDS on their site.

The instructions that come with the kit do not tell you that lower temperature means low pressure. A good system can operate at as low as 20-30 psi on the suction side. On the colored gauge, that would be near the bottom of the scale. They also don't tell you to purge air from the hose, although some kits come with a pre-attached hose already fileld with refrigerant.

A new product recently hit the market:

Threebond is a Fuji-branded sealant and there's a part number for it which I think I have, but it's not worth the expense.

It is possible, for example, if the shaft is broken or the hub bolt is loose. This may sound strange but so is the problem itself. So are you positive the compressor is pumping? Even if the expansion valve is stuck open, inlet and outlet pressures should be different.

If the compressor is engaged there has to be higher pressure on the discharge side, unless the valves inside failed completely. Even then there would be variations as nipper says. Can you turn the hub by hand (with engine off, no need to remove the belt) and see if the compressor shaft spins with it?

If the outlet temperature is 20F below ambient, it may not be enough to be comfortable, even according to your own standard. Add: see the post right after yours to see what the outlet temperatures should be, 30-40 below ambient is about right.

The temperatures at the outlet grilles (as in the original question) and in the interior are two different things.

Use a gauge set and read the pressures. It may still be undercharged. Pressures at both ports are important. Also do not rely on those color-band gauges from "recharge" kits.

Using cable ties is fast, cheap, and good.