BAC5.2

And if your wife is walking alone in a perfectly safe part of town, a sketchy dangerous criminal is following her. She unlocks the car... and both can get in. This has actually happened to a friend of mine. Had his keyfob unlocked all the doors, someone would have been able to hop into the car with him.

$108 is a pretty raw dog price: http://www.subaruparts.com/cart/?pn=60319FA201 - $85 shipped.

I'll snag one when I get home. The final weight was around 565g, so ~1.25lbs. It was awesome. The only downside was cutting a 1" relief in the top for a gel-coat sticker that never ended up getting made. If it was full round, I would have used it for longer. It was also only 2" in diameter. I might re-make one with the cue-ball diameter and full-round, because it's the perfect size. I'd just want to get it powdercoated in white, with a SMOOTH finish rather than the traditionally rough texture.

The one I made is steel, powdercoated black. I haven't used it in a while, since I made the cue-ball shifter. Maybe I'll go back to using it for a bit. It's tough, because the cue ball is the best shifter I've ever had.

Make it out of tungsten. I already have the CAD drawing done, I just haven't had the time to cut it... and I haven't wanted to dump a couple hundred bucks on a chunk of tungsten that size.

I wish I could get away with cutting random parts on the mills at my work.

You probably shouldn't have gotten a Spec clutch.

That first pic looks fan-freakin-tastic. I want a set of silver BBS's bad. Oh, and I would remove those wheels immediately. That is NOT safe. Not even in the most remote sense of the word. Not at all. Those wheels are now garbage.

Is it locked closed? Fold down the seats and crawl into the car. There is a lock access panel in the middle of the hatch. It's oblong shaped, and pops off with a panel tool or screwdriver. From there, you can use a screwdriver to pop open the latch and open the hatch. Then, pull the trim off and visually inspect the system. You should be able to disconnect the harness from here, and use a multimeter to probe and find out if voltage is going to the connector when unlock is pressed. If so, then it's a mechanism fault. If not, then it is almost definitely a BIU problem. You can remove the rear trim piece off of the outside of the car, there are like 18 bolts to remove, you need a deep-well 8mm socket to get them out. Once that is unbolted, you need to unplug the license plate lights, the reverse lights, and pull the wiring harness out of the door (a rubber plug). That'll give you access to the handle from the outside. Then, unbolt the handle and replace the whole latch unit. Looks like the PN is 63032AG00B for the latch and actuator, and should cost a little less than $50. If you need to replace the handle as well, that PN is 63160AG00A which costs about $20 or so.

Pull off the hatch trim and check for interference while pushing the unlock button. Better that it get stuck locked, than unlocked. But a terminally locked hatch, turns your versatile wagon into the most inconvenient sedan in the world.

My solution doesn't involve anything with the knuckles. Just machine and heat-treat a new hub to fit the stock bearings. Voila. 5x114 without any drama. If the STi bearings are, indeed, much beefier than the stock Legacy ones. Then machine a new carrier cup with the proper bolt pattern to bolt to the stock knuckles. Those would be my approaches.

What about the front?

I don't know how cut and dry it will be. Especially swapping stubs on the axles. There are three approaches I'm thinking about. 1) SpecB axles make running an R180 in our cars very easy. The axles aren't cheap, but they are a bolt-in solution. No hub changes necessary, unless we want to run less expensive wheels (5x114 is easy to find). For this scenario, I'm half contemplating designing an OE-splined hub that fits the Legacy bearings, and has a 5x114 bolt pattern. For this, you would press your hub out, and press the new one in. This would work with the OE axles, as well as Spec-B axles. Easy as pie, no muss or fuss. In the front, this would be the same approach. Press out the old, press in the new. 2) STi rear axles are dirt cheap, because they only fit STi's. So I'd follow the same approach as above, designing a new hub that presses into stock bearings. But instead of having the OE R160 spline, I'd have an R180 spline. This would let you use cheaper STi axles, but completely excludes people who just want to go 5x114 and keep their stock rear axles. Spec-B guys are shafted, because they can't run 5x114 without buying new axles. This is offset, because their axles are worth more to people who don't want 5x114. 3) The 5x100 wheel bearings are questionable in quality and durability. The 5x114 wheel bearings aren't thus effected. Since I think drilling out the trailing arm is a little worrisome, I'd rather design a bearing cup that fits the STi's larger bearings but bolts directly to the trailing arm. This would use OE STi bearings, hubs, and axles. The fronts would be similar, but you can retain the stock Legacy axles, which you'd be doing no matter what. I'm not sure if this is actually possible, because I haven't seen a trailing link removed from a vehicle, so some machine work might need to be done. If that is the case, then it's not worthwhile, and solutions 1 and 2 are better.

I care about it. I want to see the hub disassembled before I jump to any conclusions.

So you want to steal his idea, produce a "kit" and then sell them for profit? That's a pretty dick move.

I have one for sale!

I thought about keeping an Excel sheet of everything, but just got bored with keeping track. I do track fuel and mileage, but that's really it. If I paid someone else to do the labor, then I'd keep track. But I don't, so it's not even worth thinking about.

http://i267.photobucket.com/albums/ii287/bac52/dont_feed_the_trolls.jpg

When I swapped to Bilsteins and Spec-B Wagon springs, the car quieted down quite a bit. It was remarkable. My seat squeaks, and it drives me mad. I'm not willing to put up with the dealer about it, though, so I'll just try and find some time to look into fix it myself.

Thanks! Since I have to run wires through there soon, ill sheath the wires to protect them.

Probably not. I was just pointing out. I definitely wouldn't "ream to fit" though. But I also don't have a 5EAT.

There should be directionality to any kind of ridge or surface condition, though. You can't just have a chopped up surface. My transmission reference is first gear on the 5MT. It uses a polished surface on first gear, and a polished, but directionally cut mating surface on the output shaft.. You can't just have a randomly scored surface. I don't know what kind of surface finish you get with reaming, but I'd think polishing to somewhere in the 10 micron range would be sufficient, and easy. It's nothing more than a cloth-wrapped dowel with some micron-range polishing compound spun at some reasonably moderate speed. Less than a minute of work, and you'll have taken down any big scores. 10 microns is about the diameter of an average hair. But then again, my knowledge of the 5EAT is... none really.

I'm just talking about good practice. I haven't really been following the thread much more than a casual glance. If relative rotation is low, then it'll probably be fine. But polishing doesn't take very much time with something as soft as brass. Typically, reaming of soft-metal bushings is to get a "rough fit" and then the bushing will wear-fit and deposit a bunch of shavings into the fluid. I figured it'd be best to avoid this if possible.

Frankster is right. Also, a ten-thousandths interference doesn't require cooling and heating to press. A ten-thousandths will probably slide together by hand. If you want to do a proper shrink fit, you need an interference of several thousandths. For a gear-drive system I did a few years ago, I did a 0.007 interference fit that required heating of the gear and cooling of the shaft (accomplished by spraying the shaft with an air-duster turned upside down). Even then, a high level of precision is required when pressing. It's not just something you can throw together by hand. For this application, a shrink-fit isn't necessary. A simple press-fit would be sufficient and would decrease unnecessary stress in the bushing. As for "reaming" the bushing out, that's not typically good practice for a bearing surface. You want a micro-polished and true surface. Ideally, you'd order precision ID stock, turn the OD on a high-speed lathe, then polish the OD and press fit the bushing. Finish it off by trimming the excess and finishing the ends on a lathe.

Almost all adhesives are poor for extremely high heat. The optimal solution would be to use a non-adhesive backed blocking material affixed to a pipe with some type of constant tension clamp (like one used for a CV joint boot). IMO, I wouldn't use any adhesive backed heat protection on any exhaust components. I would use them on intake and adjacent components (firewall, under side of hood, etc).