ScottFW

Clutch hydraulic hose had been seeping for a while so I installed a new one along with a new slave cylinder. Changed the transmission oil. Rear diff oil will get done later this week. New front brake rotors, re-greased the sliding bits and bled the system. New rear shocks, with 3/8" saggy butt spacers. Car rides a thousand times better, no longer squats like I'm hauling a trunk full of dope, and it brakes from highway speeds without shaking violently, which is nice. Front struts will have to wait a week or two before I have time to do those, ideally on a Thurs/Fri followed by a Saturday alignment.

I've got Auto Zone remans on both sides up front. Passenger side was replaced first of course, but by now they both have at least a couple years and 50K+ miles and they're still fine. I could not swear that they were rebuilt OEM axles though.

Where the air goes is the mode actuator. Recirc versus fresh air is a separate actuator. Rarely fails because it's the least used. Passenger and driver side each have their own temperature (air mix) actuator. The mode, recirc, and passenger side temp actuators are accessible by removing center trim, dropping the glove box, etc like you would to change the cabin air filter. Sounds like you only need to do the mode actuator. As long as the passenger side temp control works, and the fresh/recirc functions properly, you can leave those alone. It's not a monumental PITA to dig in there again if one of those fails in the future. The driver's side temp actuator requires removing the whole dash to access it. Huge PITA, did it back in January.

Yes, mine is a 2005. 9 hours sounds kinda in the ballpark for a DIY job. Probably took me slightly longer in total, but a significant % that time was spent looking at the manual on my laptop to figure out what I had to do next, since I had never previously torn into it beyond what's necessary to change the cabin air filter (no easy access panel for that on early 4th gens). Gut feeling is it should take an experienced dealership mechanic quite a bit less than 9 hours. The actuator itself cost me $62 + s/h. There's a lot of stuff that has to come out, and I don't know how 5th gens differ on this, but none of it was particularly difficult. Find a free day and dig in.

I replaced that actuator a few months ago. It died stuck in the full cold position, in winter, naturally. You don't have to remove the heater core, but yes, the dash has to come out to access that actuator. Center vent/cubby assembly, center console & trim, stereo head unit & HVAC controls, drop the glove box & trim, drivers side lower & upper dash trim, instrument cluster, A pillar trim, all of that, then you can pull the dash. I forget the exact order, but the shop manual is actually somewhat helpful provided you have the patience to follow the trail of info for each panel. Just think of it like a "choose your own adventure" where you have to navigate from one page to the next! It pretty well points out the screws that need to be removed, and they are not too hard to find when you're in there working on it. I had never removed the vent cubby or instrument cluster before, let alone the entire dash, and I did the actuator replacement in one day (several hours). Still need to do the mode select actuator (noisy but still works) but that one is child's play compared to the drivers' side air mix one. Honestly, I should have at least called the stealership to get a quote on this job. It would have been worth a few hundred to have somebody else #@&% with it, but I was feeling masochistic and had no other plans for my New Year's Day this year. It would also be my strong opinion that you should install a brand new part, rather than try to regrease or re-bend the contacts or do some other well-intentioned but cheap-ass solution to reuse the existing one, given the PITA of the surgery. Couple of pics looking down through the windshield. Actuator is the part with the green plug going into it.

As long as it says Dexron III you're good. G versus H and previous iterations are just minor changes in the additives and such. I believe H is the most recent and it has been around for a good while, since before 4th gen LGTs started being produced.

Mine won't use any T6 for about 1500 miles after a change, then uses about 1.5 quarts by the 5000-5500 mile mark when I do a change. I check the level about every 500 miles and top off whenever it needs 1/4 to 1/3 qt. I do a lot of highway driving at steady >3000 rpm. Higher rpm is going to shear the oil down marginally faster than putting around town at 2000-2500 rpm just from being pumped through the engine at higher pressure. There are also plenty of situations when I'm off throttle in 5th and decelerating, which pulls a nice vacuum to help oil get past aging rings and valve stem seals. Last Blackstone UOA I ran on T6 (80K miles ago?) the viscosity and everything else still looked very good at 5500 miles and they said to "try for 7000 miles next time." LOL no thanks. @Max FWIW on the turbo seals, mine used oil at basically the same rate from 96K miles when I bought the car until a little over 182K miles, when I got the first inkling that it seemed to be using a little more (more than the normal increased use over an oil change interval). Watching it like a hawk over the next 1-2 weeks confirmed the increase, then one day it used 1 quart in the 90 mile round trip to work and back. That freaked me out enough to leave it parked until I got the new CHRA installed. So my experience was that when the VF40 died it was not a subtle maybe/maybe not kinda deal. You get a slight warning followed shortly by "OMFG it's toast."

Only reason I wouldn't do it exactly that way is because undoing the strut bolts from the spindle means you've undone the alignment. Are a couple of scribe marks sufficient to get it reinstalled perfectly? Maybe, but I'm neurotic enough that I'd feel compelled to verify in some semi-accurate way. I've got a camber/caster gauge and toe plates, but checking all that is no more or less of a PITA than pulling the entire CV shaft from the car. Then you could re-boot it on your workbench and know your alignment is still good after reinstalling.

Not sure what brand of 55w came in the car when I bought it but IIRC they were both dead in 3-4 weeks with the DRLs. Put in ordinary bulbs from my FLAPS, probably garden variety Sylvanias (definitely NOT Silver Stars as they had a reputation for short life). I promptly yanked the DRLs and that set of bulbs still died way too soon, thus leading me to find the Osrams. I agree 14 months is not as long as I'd like the clearing job to last, but my car is doing 70-80 mph about half the time that the engine is running. Getting blasted with road grit/gravel, rain, or whatever else takes a toll. Wifey's lives outside in the elements full time and the coating lasts much longer, presumably because it's not getting constantly highway blasted. With the regular 3M polishing kit, and no real coating/sealing layer of significant thickness, the lenses started looking crappy again after 6-8 months, so 14-18 months is significantly better. Not sure what would last longer, other than $buying new housings$. Those LEDs do look pretty insanely bright. Will be ordering a set soon. EDIT: It appears they may be temporarily out of stock on the H7s as I don't see that option listed. Were you able to get the rear housing covers back on after installing the LEDs, or do you run without them?

The biggest difference is that the Osrams really filled out the pattern of the stock projectors, more even light without the dead spots, plus there's just more of it. Also, even with DRLs yanked, I got poor life from stock 55w bulbs. IIRC one set gave me less than 6 months and 10K miles, versus 2-3 years and 40-50K miles from the Osrams. I'm due to re-clear my headlights pretty soon. The method of using Varathane cut with mineral spirits lasted me well enough (first signs of flaking after ~14 months, now looking pretty scraggly at 18 months) that I'll probably just do that again. That same coating is still going strong on wifey's MDX, but she doesn't drive anywhere near the mileage I do. Thanks for the pics. The dark road low-beam only shot, is that perhaps appearing brighter on screen than your eyes see IRL, or is it pretty close? I ask because the road signs and a portion of the double-yellow line appear more blown-out (exposure wise) than I've seen when driving HID/LED lit cars.

Thanks for guinea pigging this. I've been on the baller Osram 65W since not long after buying this car in 2012. With DRLs yanked I've gotten well over 2 years per set (might actually be closer to 3 years than 2) with me driving about 22K per year. They seem to be nearly unobtanium now, and I see that the vendor who used to be the main source for them is trying real hard with the marketing language to convince people that a 55W Vosla bulb might be a remotely suitable replacement. Shyeah right. I've been happy enough with them to have never felt the need to mess with HID conversions in the intervening years, but still, my beam pattern gets swamped by the more modern vehicles as I pass them on the highway. At $37 I might give these LEDs a go in springtime, assuming yours are still holding up. Got a pic of these projecting down a dark road?

I put in a reman a couple of years ago after my re-lubed OEM starter died completely (contacts/solenoid). Needed to get to work the next day so I went with "anything available locally that night." Managed to locate a reman at Auto Zone (Carquest brand maybe?) and it has been fine for 2-3 years.

4+ years after this mod, still running incandescent bulbs, no issues. Good to know how to repair and/or beef up the BIU though.

The destiny of just about every Legacy that's had a windshield replaced in its lifetime. Nobody pays proper attention to priming and painting the giant gouges down to the metal that they create when removing the old windshield, just slap a new one in there and hope the sealant and trim take care of it. I had rust much worse than yours, but I recently put in a new windshield and I had the body shop fix the rust while they were in there. Between the heated windshield and the rust repair, it wasn't cheap. Mine was previously leaking water too, but only when I was driving on the highway during rain, not when hosing off the car during a wash or sitting parked in a downpour. At highway speed the water was running up the windshield and getting under the trim that way. A little bit of silicone sealant worked in along the front edge of the trim completely stopped the water from dripping on my leg. It won't do anything to slow the cancer, but it bought me another 18 drip-free months until the glass got bad enough to need replacing anyway.

The vast majority of my miles are on the DC beltway and I-270. Probably 20% stop & go and 80% cruising, and the road surfaces are pretty smooth. I measure tread depth every time I rotate them (about every 7-8K miles). Inside, middle, and outside on all 4 tires to make sure they're wearing evenly across the tread, since alignment issues can sometimes be detected that way before they're felt by the driver. Anecdotally I can say the rate of wear has gotten progressively slower throughout the life of the tire. The first 2 or 3/32" went away probably within 3-4 months, and now towards the latter part of their life I doubt they've lost 3/32" in the past 12 months. Road noise has gone up as a function of decreasing tread depth, but dry & wet performance have been consistently good. Tire Rack prices for stock size 215/45-17: Conti DWS $95 on closeout. Maybe a consideration for budget-conscious people in snowier areas farther north, but the new DWS 06 is far better. Conti DWS 06 $128. This tire gets trounced by the Michelin Pilot AS3+ and the Pirelli P Zero All Season Plus in all areas except snow performance, which they haven't formally tested for all 3 yet. It's presumed the DWS 06 will win handily there but I'm not willing to live with its inferior dry and rain performance on the 355 days of the year when we don't have snow. Michelin Pilot AS3 (old version) $125 on closeout. No reason to buy at this price since I can get the newer and better AS3+ from Costco for about the same price. If they were more like $100 ea (perhaps if Michelin threw in a rebate?) they might get some buyers. Pirelli P Zero All Season Plus (new compound) $109 after rebate. I probably just need to buy these now before the rebate ends in 4 days. Then there are a bunch that are out of the running for me: Hankook Ventus S1 Noble2 $92 General G-Max AS-03 $92 BFG Comp-2 A/S $99 Dunlop Signature HP $102 These are all priced close enough to the Pirellis but get edged out in most tests and in particular have worse rain performance. Most of the above are 500 treadwear, same as your Coopers, so hard to say if they'd last any longer for you. The Conti DWS and 06 are 540 & 560 treadwear, but have the same treadlife warranty as the rest of the lot, which is 45K or 50K miles. Problem is those warranties don't kick in until they're on the wear bars, and any tire is going to be much more prone to hydroplaning in the rain and will be useless in any amount of snow by the time the tread gets that low.

You can try what the tire shop guy recommended to me when I got a bolt through the tread that was too close to the shoulder for him to be comfortable patching it himself... slather the plug with RTV before you insert it. Keep the tire at 10-15 psi until the RTV cures, then inflate to your normal pressure and see if it holds that pressure overnight. I did this to one of my Michelin Pilot AS3s and it’s been trouble-free with well over 20,000 miles on it since then. Saved me having to buy a whole new set or having to dick with getting one tire shaved down. My AS3s have over 50,000 on them and are at ~5/32” give or take. I don’t drive slow, but OTOH I’m not exceeding their limits and sliding them around with great frequency. IMO you are getting shitty life from a 500 treadwear tire with only 24K miles on them, provided you have a good alignment, keep them properly inflated, rotate them regularly, and aren’t autocrossing or doing HPDEs on them. Coopers seem generally well regarded but I don’t know how much of that is just because they’re made in the USA, or if they’re “good for the price” versus good [period]. I have a set of their AT3s on my 4Runner. I barely drive that thing so I can’t speak to their longevity, but I’m generally happy with them. In ultra high performance all seasons maybe look at Michelins or Pirellis. They cost 50% more than Coopers but you might get double the life. I say “might” because I can’t divine exactly why yours are dying young. Since I bought mine, Michelin has updated the tread compound on the AS3 (now called AS3+) and Pirelli has quietly updated the compound on the P Zero All Season Plus but not given it a new model designation. Tire Rack’s tests show the new compound Pirellis outperforming the AS3+ in noise & comfort and in the wet (and my old version AS3s were already pretty impressive in the rain). AS3+ still had the edge in dry performance but rides harsher and makes more noise (a bit of a complaint I have with my AS3s). Pirellis with rebate plus shipping from Tire Rack would be about $50 cheaper for the set than the AS3+ from Costco with sales tax. Haven’t decided yet but will probably go with one of those two options pretty soon here.

I had a gunky injector causing roughness counts in one cylinder that could be heard & felt at idle, but got no roughness counts or knock and it ran smoothly while driving. I pulled the injector from that cylinder and did a DIY cleaning by letting it sit in a shot glass with some Sea Foam and then pulsing it with a 12V battery. That cured the roughness, but I went ahead and had my injectors professionally cleaned by Deatschwerks anyway, and I haven't gotten any roughness in the ~50,000 miles since then. I should say that my injectors were not too crummy looking even before cleaning. Just a slight brown film/varnish on the nozzle end, nothing particularly alarming, but that's no indication of how dirty they may have been inside. Point being, unless your used OEM injector/s came with a receipt for a cleaning service, they could be just as bad. Conceptually it made sense to me that a non-optimal or dribbly spray pattern is going to have a higher likelihood of manifesting itself at short pulse widths (idle) versus longer PWs under higher rpm & load conditions. I don't think a coil or plug would only cause misfires at idle and not under load, actually I think the opposite is more likely. And your wiring will be jiggling plenty when driving down the road, so that would not be my prime suspect for idle-specific issues.

Curb Weight as the EPA defines it is supposed to include all fluids including oil, coolant, and a full tank of gas, and all normal extra equipment like spare tire and jack. Per my VA registration mine has the same 3,365 curb weight, but it's also just listed as a Subaru Legacy. No differentiation of it as a GT, or of trim level, so that might be the curb weight of a 2.5i non-limited. Turbo + intercooler + supporting bits would probably account for that difference. Never put my LGT on scales but being a Limited it also has heavier power seats and whatnot.

Those pulses in vac lines make it very difficult to tune ECUs that use MAP-based fueling (such as the Megasquirt in my Miata) without a line filter. Small lawn mower fuel filters do a pretty good job if you don't have a spare stock part lying around. http://static.hardwarestore.com/media/product/283275_front500.jpg The pulses of course are coming from the intake valves opening and closing, so sourcing the line closer to those valves or only in one cylinder's runner is not ideal. The filter should smooth these out to the point of being negligible for a boost gauge.

Yours is right around the same odometer reading mine was at when I had the same injector problem. Once I pinpointed the guilty injector my crude DIY cleaning worked well enough to get rid of the roughness. Basically soaked it for a while in a shot glass with Seafoam, rigged up a 12V battery to a momentary push button switch and pulsed the injector a bunch while it was immersed in the solvent. Of course, that method doesn't really flow solvent well enough to thoroughly clean & flush out the innards. I also didn't want to be running around long-term in a turbo car with injectors that were likely not delivering fuel evenly to all cylinders, so I sent them in to Deatschwerks for cleaning & flow testing. Looks like you have another DD you can use, so I'd recommend that course of action if it won't kill you to have the LGT down for a week. My car has about 20K on it since the injector cleaning and all is well.

Subbed, need to do this. I cleaned & lubed just the front part when I had the starter out for my clutch replacement 2 years ago. It helped for a while, but I didn't disassemble and lube the rest of it, and it's squealing intermittently now.

High A learning and normal-ish B, C, D is most likely a vac leak. The reason the B, C, D values are closer to zero is not because the leak seals up under boost; it's because with higher volumes of air coming into the engine the size of the leak is a smaller percentage of the total. In range A you have relatively less air coming in, so a small leak makes a bigger % difference to the total. At 9-10% it is probably a small leak. Mine was at 13-14% and the leak was just a loose TMIC to throttle body coupler, on the throttle body side. A smoke test would find it, but I used a homemade pressure tester without smoke. Hooked it up right after the air box, put about 10 psi into the intake (determined using a mechanical boost gauge), and listened real close for the hissing. It can still be difficult to pinpoint even if your ear gets you in the ballpark, especially if the leak is on the inner/under side of that coupler where you can't get your hand to feel the air stream.

Ah, gotcha. A M12x1.25 bung plug is about $5-6 on ebay. I'd be more inclined to go that route than bother with cutting down a bolt, but that will work too.

It's not the sensor that's the worrisome turbo-damaging failure part; it's the cat in the OEM uppipe. We're replacing the UP with one that doesn't have a cat, and no EGT sensor either, hence no bung that requires plugging. It doesn't make a whole lot of sense to remove the EGT sensor but keep the stock UP.

Yes, have kept the incandescents since I did the mod ~2 years ago, no issues.