BRAAP

Update. Performed the above mod on daughters freinds '97 Legacy GT. Bought myself '03 LL-Bean H6 Outback wagon, cuddled up with wiring diagrams, DMM and performed same mod here as well, even easier ont eh '03, just unplug and install a jumper so the lighting systems sees permenant "OFF" signal, no inadvertent dead batteries after resetting trip meter etc.

Will do, thanks again.

Heads back back on, engine back in the car, cooling system burped, fired right up, shake down run was good, handed keys back to daughter, she's been running all over town today, so far so good. Not hold holding my breath for long term. Thanks a bunch guys, greatly apprecaited. Hope this thread ends up being a help to others.

Thank you for the heads up, definitely a learning curve in adapting to the Sube nuances.

Agreed, something is going on to cause the gasket to pull in like that.

Water pump and Timing belt are new, praying it wont let go again, but not holding my breath.... (read on).... Bought new OE SUBARU head bolts this AM, installed the heads, clean and dry deck surfaces, thin coat of motor oil on head bolt threads, thin coat oil under washer and head bolt thrust face. Here are the head bolts, old were what was in the car when I originally tore it down, (forging around the head appears identical to the new OE), the grade 10.9 are the aftermarket I purchased, and the breand new head bolts from Subara. Hindsight being 20/20.... When I originally tore the heads head off the first time I recall noting the open deck design and rolling my eyes, (as a performance engine builder, not a fan of open decks, it can and does work fine but is susceptible to cylinder walk where closed deck designs dont), also noted how the sheetmetal rungs in the head gaskets on the exhaust side had gathered/bunched up/pinched, odd but carried on not giving it a second thought. After installing the heads this last time I stumbled across those old head gaskets again, light bulb.. the cylinders are walking, pulling and bunching the gasket up. This happaned in the same exact location that the head gaskets I installed let go, fire rings also show signs of movement/rubbing from the wandering cylinder decks..... Even though this time I am using OE gaskets, new OE head bolts, paying even clsoer attention tothe cleneliness of hte deck surfaces, I dont expect it to last past 10,000-15,000 miles.

Heads decked and true. The headbolts I bought and installed were aftermarket. Still have the old head bolts, not sure if they are OE, is there any way to tell if they are? Being as the heads were warped it was clearly overheated in its past, I assume that is why previous owner had head gaskets done, no idea how bad the overheat was other than the amount of warp the ehads received Cylinder walls show no sign of skirt/piston seizure, (not sure at what point Suby skirts or thick meat around the pin boses swells to the point of seize as seen on severaly overheated inline/V engines). Any how, heads cleaned up at .006" and .007" for a clear no shadow deck when I decked 8k ago, they are still perfectly flat and true. Being cam-in-head I am a tad concerned about the cam line not being true now, though at only .006'-.007" warp cam bearing/housing wear should be minimal. Case deck warpage/cup is just under .002", both sides.

Thanks for the heads up, just looked at the new OEM gasket, it is a Graphite gasket. For some reason I thought it was a paper gasket which would lend some credence to the application of tack etc I had read elsewhere. Being Graphite it has been my experience, as you suggested, never NEVER use sealants on the Graphite face, not a good practice. Fire ring on the OEM gasket appears to be a thicker/more robust wire within the fire ring vs the aftermarket gasket I originally used, should help keep the combustion pressures from escaping the chamber. Is retorquing the head gaskets after so many engine run/heat cycles recommended? Paul

Short version, ’98 Legacy L Sedan, EJ22, daughters car she recently purchased. Previous owner had done head gaskets in the past, (within 10’000 miles or so). Leaking oil all over including head gaskets dripping oil from the oil drain back, removed engine, bought a budget gasket set with the silver graphite head gaskets, new head bolts, replaced all seals, crank, cam, pulled heads, etc. Heads were warped, (wasn’t dealt with during previous work done) so I decked the heads, put all back together, torqued the hoky-pokey torque sequence spelled in Haynes, had factory Subaru T-stat. Ran great, no issues till about 8,000 miles, then after long down hill no-load condition, (1+ mile down hill), at about the 1 mile point temp would literally rocket up, gauge would swing almost as fast a tach could swing, coolant blowing out of the overflow all over engine bay. Replaced T-stat with another new Subaru brand T-stat thinking T-stat stuck. All well till next trip down same long hill, same rapid temp rise issue. Long drive from Portland to Seattle and back, no problem, but down that hill it would rapidly overheat again. Fought this for couple weeks and hand full of cycles, Opened radiator cap, distinct catalyst exhaust smell, yet no water in oil. Pulled engine again, removed heads, evidence of exhaust gasses getting past fire ring into the coolant chamber towards the center on bottom/exhaust side. heads still perfectly flat, case decks will allow .0015” feeler but not a .002”, decks are acceptable. I’m at a loss for how/why these gaskets let go, dont want another premature failure. As professional engine builder/machinist since the ‘90’s, (literally thousands of engine builds from stock rebuilds to very high-performance builds including world record setting 1146HP TT BB endurance race engines), all my work has been traditional inline and V-engines, lots of Datsun L-series, domestic iron and modern GM LSx, no Suby engines, this is the first. Applying the techniques I know to work with typical graphite head gaskets, clean, dry, oil free deck surfaces, I assumed it would be fine, turns out I was wrong as these gaskets didn’t last past 8000 miles, with the tough and stable EJ22 at that. Obviously my fault some how, right? Not wanting to repeat the failure, little digging revealed current factory head gaskets are favored amongst those that have been tinkering on these over the years along with application of Permatex Aviation Form-a-gasket on both sides of gasket. Just want to be sure I am covering all bases before reassembling. Recap before reassembly; Head decks are perfectly flat/true, smooth flycut surface, deburred on all bolt holes and exterior edges and Combustion chambers, left raw-as cut in oil and coolant transfer passages. Case decks are flat, (less than .002” from true). Factory Suby Head gaskets, thin film of Permatex Aviation Form-A-Gasket on both sides of each. Both case and head decks cleaned with brake clean, oil free. All head bolt holes blown out to remove any fluid/debris. Light coat of motor oil on head bolt threads and under thrust surface of the head bolts. (Seems reusing head bolts is acceptable?) Follow the torque, release and retorque sequence in Haynes, (recently calibrated Snap-On torque wrench and Snap-On dial-a torque degree wheel). Am I missing anything, got anything crossed or messed up,off my rocker?..... Will also be cleaning radiator and back flushing heater core and verifying adequate flow through both, along with tacking water pump impeller to shaft, new Suby T-stat.

Been bit by this one as well, 98 Legacy L wagon, (our Plan "B" transportation) which our oldest daughter has been driving for past 3 years, (she put over 40,000 miles on it), we gave her till end of Aug to get her own car, this past weekend she bought a sweetheart '98 Legacy L sedan. In giving it the once over, replacing the bad knoock sensor, new battery, tune up, etc, thought it a good time to figure out a way to properly disable the Virgin switch "feature" for her car and ours. Being as we have a '95 Legacy L sedan parts car on hand I set out to do just that today. Tearing into the parts car discovered that switch is 3 wire SPDT switch, COMMON, OFF, and ON. In wanting to disable this feature in a way that does not interfrere with the normal lighting funtionality and alleviate the inadvertant dead battery, (were not so lazy that we cant start the engine to turn on our parking lights if we desire them for any reason), I just spliced the OFF wire to the COMMON on the switch side of the connector under the steering column, functionally the car sees the switch in the OFF position permantly now, switch is no longer operational. the '95 and both our '98's have the same connector and wire color coding. Disclaimer. If attemtping this yourself I strongly encourage verifing whcih wire is which using a OHM meter before cutting and splicing as wire color coding/pinouts may be model specific, i.e. not matching what I'm showing here. Be sure to "safe off" the unsed wires, just good practice. ;-) Works perfectly, lights function normally, switch on top of the steering column disabled.

As I mentioned, thank goodness for my wifes sense of humor and willingness to help. Normally for stuff like this, she would said NO WAY! Regular cheapo mechanics stethoscope. The hood removes in a matter of minutes, then... well... lets just say… On another forum, the rule is, No pictures, then it didn't happen..

Problem FOUND and FIXED!!! WOOO HOOO….. Driver side front wheel bearing. First off, a huge thank you to my lovely wife of 20 years come this November, for her help and having a sense of humor during this diagnoses. After much time hanging out the windows listening, (didn’t hear the growl at all just listening out the windows), with the stethoscope in my ears all over the car as she drove, Transmission, front suspension, suspension pickup points on the car, seat belt mounting points, rear hatch area, etc, the noise showed itself on the drivers side with stethoscope on the lower control arm, T/C rod, and also on top of the drivers strut, while driving. In the car, it still sounded like passenger side. My wife and best friend after driving and riding in the car, were all dead seat on the issue coming from the passenger front, as I was, not the driver side. Any how, removed the old bearings tonight, they were nasty, catchy when trying to free spin them by hand etc. Installed the new bearings, test drive, NO noise or growl whatsoever! Thank you everyone for all your input and to those who kept saying not to overlook other portions of the car because the issue could be somewhere other than where the noise sounds like it is coming from, which was the case here.

I think the rear was mentioned earlier in this thread as well. Later today I’ll round up a co-driver and investigate the possibility of it coming from the rear. Thank you, Paul

I’m leaning towards the possibility of the wrong half shaft in the passenger side causing this growling noise, though I’m not 100% sure I should trust the table posted above regarding which half shafts I have vs. which ones I am supposed to have in this car. Hoping one of the Suby guru’s savvy with half-shafts will chime in. If it is the wrong half shaft and it is too short/long, that might explain the noise I’m experiencing?!?!?

The micro switch you are referring to is the Hi/Lo indicator switch and it is fine. It is the other switch that indicates when 4WD is engaged that I want to adjust. I’m starting to get the impression that the switch I want to adjust is not really adjustable like I hope it would being in/on the transaxle and actuated internally, i.e. no amount of shift linkage adjustment will do anything other than affect the shift lever position, not alter "when" the switch illuminates the 4WD light. If this hunch is correct, these are the options I see available in wanting to retain the 4WD and Hi/Lo indicator lights. (feel free to add more options...) 1) Just live with it the way it is. 2) Remove the bulb or disconnect the switch all together and go without 4WD dash indication. 3) Bungy the shift lever down until I need to use 4WD. 4) Rewire the 4WD indicator light to the Hi/Lo micro switch and then reposition the micro switch so that it indicates 4WD. This option eliminates Hi/Lo indication, which is really pretty self evident when engaged anyhow. 5) Add another switch at some point along the 4WD linkage and rewire the 4WD indicator to that switch. If the OE switch is not adjustable, I'm thinking #5.