DAlgie

The tapered roller does have parallel contact faces, it's just that they have tapered surfaces (and rollers) to make up for the circumference difference so there is no scuffing. Take one apart and study it, you'll see how it all makes sense.

A tapered roller is a higher capacity bearing than an angular contact ball. if you think about it, the tapered roller has a lot more contact area, and the roller runs in a perfect circumference, with no scuffing at all. The angular contact ball has small contact area, and the edges of the contact area are at a different circumference to the middle, some differences in contact speeds are seen, hence the slight scuffing.

Yeah, I would have thought that both cars are the same. See if you can find out what the '01 Legacy uses, wondering if it also is a ball, meaning that they went to ball, away from tapered roller, which is a shame. The dimensions you list there, the Legacy numbers are probably converted back from metric, and are actually the same. Bearings are always full millimeter sizes, and are almost always metric. Anyone have any info on the tapered rollers failing?

This had no Loctite on it at all, you can always tell if there's any there. I would not use Loctite, this will only increase the bearing fit, the absolute last thing you want in this case. Just assemble the new bearing in it's bore with gearbox oil to lubricate it going in. You can regrease them, but the work involved is a real pain in the rear, for about $110 per wheel, I'll put new bearings in, especially if you have 100k miles now. This old bearing went through the hard chrome, yours might be getting there now, and you would never know it. New grease only might delay it, but not by that much.

Yes, we run just one race with the bearings, don't always replace them though, just a clean, inspect and regrease. I think these right now run maybe 1200 miles before we pitch them. A bearing of this size (The Impreza) pressed into cast steel, it's quite a wide bearing, I would give it maybe .0007" to .001" interference fit, I didn't measure this one, it's back together again, but I would guess it had .0025" to .003" interference, way too much in my book. You can get away with more with a heavy wall outer race bearing, it will not shrink down from the housing it's pressed into, but this has a thin outer race and the housing is thick, so it will command the OD of the bearing to whatever the fit is. I'm hoping that me pressing the old one out and pressing a new one in has burnished the bore slightly, and now has a little less interference. It did go in easier than the old one coming out for sure. A twin tapered roller will withstand more preload, but that's not a fix for too much in the first place really.

I guess realising how the ECU determines that there is a misfire is in order. It looks at a notched wheel on the crank through a hall effect (or similar) sensor and measures the time between each notch, comparing this to a formula with rpm. It knows how much less time there is between all of these notches after a cylinder fires. In other words, the crank is accelerating and deaccelerating with each firing impulse and cycle in between, if the time between notches drops out of the table it's comparing to, this is seen as a misfire on that cylinder. No thought is given to what caused that misfire, just that it didn't fire. Older systems (OBD1?) didn't have this ability, therefore if you had a problem and no check engine light then first order was to look at the ignition, there was no system to check this normally.

Indy cars use different bearing packages from year to year, common was a combo twin angular contact like this, with a plain needle roller as well, older cars had twin tapered rollers, problems with those is they are very touchy with preload, much the same as the angular contact package bearing, but those, like this Sube bearing, have preset preload.

No, there is a retaining ring that holds it in there. It presses up to a shoulder, and then you fit the ring. Really, I have a lot of experience with bearing fits, this one is too tight. Quite possible that this one is on the low end of the tolerance, but that also means that the range is too large, or all are low in the range. From the problems they have, I would bet the range is on the low side to begin with. This car is a 2001, 68k miles, everything else on it is in really nice shape, I have owned it since new and it's never been abused or hit a curb. No water was present in the bearing, and the seals weren't leaking grease out.

Pretty sure that a misfire code isn't just that there is no spark. Also, you need to ground the spark plug to the block, not hold it away, otherwise you are forcing the ignition to jump TWO gaps, which is hard to do normally. The misfire code is telling you that those two cylinders didn't fire, but that might not be ignition. Possibly that bank has bad leakage, blown head gasket even. An engine needs compression, reasonable quantity of fuel, and spark at the right time to run.

I had a thread going on here erlier about rear cv removal. The thing ended up having a dead rear wheel bearing, BIG time noise, so loud at freeway speeds that it was deafening. ANYWAY, I finally got the bearing out of the rear hub carrier/ upright. I have done a lot of race car wheel bearing diagnosis, we work our Indy car wheel bearings very hard on an oval. Subes tend to have whel bearing problems, yet nobody can figure out what the real problem is. My diagnosis is that the bearing is sized properly for the loads it sees, the package seems to be well suited for the load. it is a double ball angular contact bearing that I think is very nice, well engineered. When I pulled it apart my first thoughts were that it either had poor quality grease, or too much preload. I doubt that the problem is just unsuitable grease. Therefore it has too much preload. HOWEVER, when I went to press the outer race out of the carrier, it took many tons to move, far more than normally needed. This means that the bearing fit into the carrier is too tight, this will shrink a thin walled race down in size, and force the preload on an angular contact bearing to be too high. Subaru, if you read this, please look into what interference fit the bearing manufacturer reccomends for this bearing package. I believe you are out of their spec on this.

Well, chances are that the oil lubricated parts don't wear that fast. I was hesitatant to look at the diff originally, but everything pointed to that. I would, next time, look at the wheel bearings first. You need to remove some part of the suspension joints to let the inner CV clear the diff, and actually get it out of the diff totally. The upper two big strut bolts come off easily, remove these, but the best bolt to remove is the lower big long one that connects both lower arms. Soak it in some kind of penetrating oil for a day maybe, spray it inside where it comes through the lower part of the upright casting. Before removing this, take the caliper off from the two rear main bolts, not just the small slider bolts, then remove the disc as well. All this will let you rotate the hub by hand without any sort of interference, you can then easily tell if it's that bearing pair or not. Amazing how loud of a noise it will make in the car, yet the bearing not feel too bad by hand. Forget trying to feel for bad bearings from wheel play, mine had none.

Well, Next, to find this bearing noise, I dropped the center driveshaft hanger bearing down and felt it for roughness. Nope. Put it all back, then removed the rear calipers and discs. Bingo! right rear wheel bearing, who would have ever thought?? Amazing thing to me is that there is no easy way to diagnose this problem (loud bearing noise under the rear of the car somewhere) without going to the trouble of 1. Removing the axles from the diff, which requires that you pull some suspension component to get the axle to clear out from the diff itself, 2. removing the calipers and discs. Then you can freely turn the rear hubs and find the offending bearing. Believe me, I went to the trouble of raising the whole car up, starting it and letting it idle in fourth gear while I crawled around UNDER it, looking for the source of the noise. Not exactly the safest thing I ever did. AWD is nice, but makes diagnosis of driveline noises very time consuming, you can't just spin a wheel and listen, and this one never had any wheel play ever.

Well, I couldn't get the lower control arms off, like you said, the big lower bolt in the upright is probably rusted to the rubber bush liners. I did, though, remove both strut legs, the two lower strut bolts on both sides. This gave me JUST enough to pull one axle out, then pop the other. Ok, so I cleaned the diff up, removed both side bearing retainers, and no bearing problem. Darn! So, dead end there. Maybe will pull the driveshaft out and check it's bearings next. This noise is really loud, I can't believe I can't find the offending bearing there somewhere.

The rest is still there, complete. I did use more force, and, yes, they popped out as described, so thanks frag for the info. Main problem now is I can't get them apart enough to get one of the axles out of the diff, the CVs won't travel enough to let one come out of it's spline. I tried to remove the rear toe link off the upright outboard but my impact wouldn't budge the large bolt, and a three foot breaker bar didn't do it either. Damm, so close.....

I did put a prybar in there and do what you say earlier, but didn't get any movement at all, so I didn't try anymore. I'll go back and try more force then.