89Ru

Premature head gasket failure after 35k miles. The troubleshooting: OEM MAP didn't solve it. Saga continues. Car has been at the Subie dealer for over 5 months chasing the hard stall issue. Subie mechanic tells me he has swapped over everything electrical/sensor related in the engine bay and no happiness. The bad: So now he thinks it is a head gasket. Is losing a LOT of oil spilling onto the exhaust and smoking from the engine bay. Exhaust gases in coolant. He thinks the stalling is caused by severe loss of compression from a huge leak. Once upon a time the stalling only occurred during the summer and only after warm-up. Now it stalls in the winter too. Wants $4K to fix it. I did the head gaskets about 35k miles ago, pulled the engine, had the heads machined by a pro, used OEM turbo gaskets, torqued to spec. Re-used head bolts, meticulous on cleaning head surfaces and head bolt holes and threads. The sad: Subie mechanic looked up the head bolt torque specs I used and says I used the wrong specs. I got the procedure from the 2007 factory shop manual. I noted when I did my 2008 EJ253 head gaskets (same engine as the 2007) the torque procedure and specs were DIFFERENT one year to the next. Anyone else notice this? My 2008 is running awesome. No gasket leaks. Everything else was the same (machining, turbo gaskets, yada yada). Mech thinks the bottom end is fine. Compression is good, just bad head(s). So what to do? Have the dealer fix it for a pile of cash? Redo the heads myself and use the 2008 torque specs? Maybe throw a new set of head bolts in there. Spend $2500 for a JDM EJ253 and do an engine swap? Any help is awesome.

Update. Verified that the fuel injector covers are swappable left to right on the intake, however due to interference with the oil filler tube, the intake will NOT install with the injector covers reversed as previously thought. So the defect on the drivers side cover was preexistent. Used a cut off wheel to remove enough material (a sliver) for the intake to install flat to the block. Bolted it up. No codes. Will edit previous posts.

So you were able to get the axle pressed out of the hub. Good job to all. Getting all that irregular knuckle metal/mass balanced in a press can be a bear! At least you have access to a press, as your wheel bearing is press in/out. Getting a ball joint out without a special tool is no fun. Ball joint tools are not expensive. A bull pin is cheap and effective but has the risk of tearing the boot.

Always nice to find OEM axles. I have whacked things for hours in the past. Sometimes a better tool is the answer, such as a hub puller like OTC 6574, a beast. Probably didn't bother the transmission, it's mostly solid metal anyway and some of that force goes into the control arm.

Yes, remove the axle from the transmission so you aren't hammering against it. With the strut free you'll have some play to swing the knuckle away from the transmission and out pops the other end. Make sure you take out the axle pin on the transmission side first.

In essence, yes. In the first pic, the southwest corner is the "high side" of the gasket as it was not squashed at all. You can see flow patterns imprinted on the gasket leading toward the chamber.

Agree with what has been posted thus far. I realize you are past this point. However it cannot be overemphasized to avoid altering the backlash from factory setting. Messing with it opens a can of worms. However if you are reading this you may have (like me) blundered into rotating something you now regret. At some point you may need to go down this road if you purchase another transmission and someone has bungled it prior to your ownership. Going through this procedure will likely not fix your current situation as the damage has been done. I expect you have returned the ring to its former position and hope for the best. Proceed with caution! I did this once, years ago, and hope to not have to repeat it. As of this post the car is still on the road with that same transmission after 25000 miles. This procedure is expected to be done with the transmission completely out of the car and mounted on an engine stand for ease of maneuvering. In my opinion it is impossible to perform on car in its entirety. I had a failed 1996 Impreza Outback sport manual transmission (was popping out of 4th gear and 2nd gear was a mess) so I purchased a used transmission. Part of my prep getting ready for the swap was changing axle seals. I marked L side but did not realize the difficulty in setting the backlash and preload so I did not bother to mark the R side and lost its factory setting. This is how I managed to recover from my initial mistake. Setting front differential backlash/preload. basic idea is to return bearing preload to existing value. Do one side at a time. Do NOT remove both retainers at the same time. You may make a tool out of pvc pipe using a cutting wheel to carve out notches to fit into teeth of seal retainer for ease of rotation without breaking a tooth. Brush mineral spirits all around prior to removal of retainer. steps to set backlash and preload. basically following the FSM procedure plus other mods. mount transmission on engine stand Mark L seal retainer position with a punch or marking pen prior to removal remove L shaft seal retainer. Press out old seal and remove o-ring reinstall L seal retainer partially, w/o o-ring rotate transmission so L axle shaft side is down Mark R seal retainer position remove R seal retainer. remove seal and o-ring place 10 lb. weight on bearing ring using tail cone You will need to turn the transmission input shaft. I used a clutch spline mating ring and a c-clamp Turn L seal retainer in with PVC mating tool while turning input shaft to get the point where there is binding and back off 1/4 tooth. This point takes a while to get to feel the binding. Install tooth lock. Flipping tooth lock around will allow 1/2 tooth adjustments. I think the retainer was about 1/2 tooth from the original location. remove R bearing weight. Turn in R seal retainer w/o o-ring while turning input shaft until it begins to bind. It was about 4 teeth between slight binding and almost not being able to turn the seal retainer. install R seal retainer tooth lock. this is the zero point. If I remember right the procedure calls for loosening the L retainer by 1.5 teeth to set the backlash and tightening the R retainer by the same, plus 1 full tooth to set the bearing preload. Use a precision dial indicator to measure backlash through the oil drain plug access point Mark the pinion gear in four places on both sides with yellow gear marking paint. Rotate the input shaft until a clear pattern is seen on all the pinion gear teeth.

Replace the other front axle.

Heartless you nailed it. No clamping force on that portion of the gasket leading to its eventual complete failure. Here's why. The metal covers for the fuel injectors are reversible! It appears I reinstalled them swapped left to right. (Edit: covers CAN be swapped but CANNOT install in this configuration). As a result there is a 1/16" portion of metal interference of the cover vs. the head leaving a huge gap between the intake and the head, as you can infer on the straight edge at the 2.5" mark. Hard to believe it ran like this at all. The vacuum must have been able to keep up with the leak for a while! Definitely pilot error on this. Not the fault of the gasket. Thankfully the intake flats do not appear to be warped!

Cooling fan speed and duty cycling depend on cooling demand. If your car is overheating both fans will run max speed. These fans have a number of ways they can die. Locked rotor, bearings, interference with the shroud, etc. Aftermarket fans tend to run larger footprint and fitment can be tricky but work well otherwise.

So possible scenario since both gaskets look bad. Airbox and hose leaks induce a lean condition, runs a bit hotter all around, further damages intake gaskets to the point of a full breach, flags code. All this in 5000 miles. Never overheated. Should I go with a different brand of gasket?

Did you swap the '03 harness to the newer short block? The only time I have seen a 340 code is from a salvage car with a broken timing belt, which in your running condition is obviously not the case.

Parts dealer is replacing the gaskets (parts only) w/o charge. I had local dealer do the diagnostic as my parts supplier is across the state. This is how the head looked after pulling the intake.

2008 Outback wagon, auto, 2.4L base model. 178K miles. Rebuilt heads at 173k after previous owner had the misfortune of a TB belt break. Used subie gaskets for intake on rebuild and turbo gaskets for the heads. Running great for 5000 miles then flagged a CEL P0171, running lean. Found a crack in the airbox where it clamps to the throttle body. Put in a new subie airbox. Didn't fix the code. Finally took it to subaru dealer. They smoke tested it, found a cracked hose from the DS valve cover to the airbox (my bad, missed that). Also found a leak at the DS intake manifold gasket. Price to fix it was too steep for me so I just pulled the intake off. Both intake gaskets look rough. The DS intake manifold gasket is cracked fully through on cylinder 4 and has hairline cracks in other places. The PS gasket is also failing with hairline cracks. Will try to post some pics. Worried about using new gaskets and this thing failing in another 5k miles. Maybe the gaskets were damaged before the install, like handled rough? They look like some kind of metal sandwich. Also I have been advised about installing them with the marking up by the dealer. (Edit, gaskets actually have no markings and cannot be installed 'upside down').

Well the car was running great for nearly three months until the past week, when it started the rough idle and stalling again at stoplights when warm. It dies with gusto. Gives an ominous THUNK when it stalls. My wife was driving it when it first happened right in the middle of getting her first Covid vaccine. The lady giving the shot said "that didn't sound good." Starts right back up although it likes to stall again at idle. Seems to drive fine until another stop then it stalls. Stalled in the driveway twice. Its still running a relatively new aftermarket MAP sensor. Disconnected the little three-wire connector from the MAP sensor on the intake and it runs fine with smooth idle, although the power seems a bit weak, with the CEL on. Drove it two days this week without any issues. I live about 500' above sea level. I don't recommend running without a MAP sensor for a monster hill climb to 12,000 feet. Going to bite the big one and get an OEM MAP from the dealer and see if that helps. Oh by the way I fixed the oven, had a bunch of cold solder joints on the control board connector for the temperature sensor and the lower element was burned out. Last pizza night was mostly a success until the pizza stone cracked, thankfully on the very last pizza.

Well I tried various combinations of the ignition key's ON-START time such as instant on (no wait, just turn key and fire in the hole), to several seconds wait, its still groundhog day. It did sputter one time but no start. With the ignition key in the ON position, I can hear engine compartment relay click and under passenger dash relay click, and also hearing the fuel pump whirr up for a second or so. Tried putting the jumper cable ground in different spots from body to intake without a huge loop without any difference. Still getting P0335 right away after clearing codes. So looking like I need to dive into the harness. Funny how I'm just now remembering that I had this same problem on a 98 OBW where the harness pigtail connector that mates to the crank sensor had a broken wire, with the exact symptoms, left me stranded that time, crank no fire, same P0335 code too.

That's interesting. I am spending a fair amount of wait time between ON and START, mostly to set up the OBD reader communications and also since 20 seconds is the wait time for the Idle Relearn procedure. This is my 16 year old daughter's car, that I broke borrowing parts to fix my car. I'm gonna have some splaining to do if I can't fix these unstable moods. She thinks my car's bad juju infected hers.

Suspecting a grounding problem due to large voltage swings seen by the OBD reader while cranking. Cleaned up the engine compartment frame rail grounds which really weren't too bad since I had just replaced two that were rotted. Attached some heavy jumper cables from the body to the intake. Reset codes. Tried to start, just cranking. Got the same P0335 code again. On the 2nd attempt it STARTED right away and sounded normal. Cut out after about a minute. Further attempts to start, nothing, but it sounds back to normal like the spark is firing during cranking which is a change. Took the jumpers off wondering if the giant ground loops were picking up interference, no change.

Back together after finding nothing wrong behind the timing cover. Attempt to start, sounded ok, just cranking without firing, then after a few cycles the engine sounds changed, like the wheezing sound it gets when you pull the plug wires like doing a compression test. Checked the ECU and it had a P0335 code, which is crank sensor, which I saw before. So I swapped the crank sensor from a running 2007 outback wagon manual. No change. Still cranks without firing. Moving the 2008 crank sensor to the 2007 doesn't break the '07, it runs fine. So the sensor appears ok. During cranking the OBD reader momentarily reads 8.5V, then stabilizes at 10. Reset codes and repeat, get a new P0335 code again with flashing CEL while cranking. Recall I got the P0335 code before diving into the tbelt cover. So a previously running car now won't start after running with a known bad coil. Some possibilities. 1. The bad coil did some damage...to what. Wires/plugs/ECU/crank sensor circuit. 2. Belt slipped a tooth on startup...can check with compression test. 3. Battery is failing? On the charger again.

Oh definitely. Nothing fixed the preignition until a new coil was dropped in. The only thing I didn't do to confirm was spray water on the coil, my electrical engineering self just couldn't justify wetting a transformer... The idle was crap until I disconnected the MAP connector then it was perfect, all else unchanged.

Not really but I'm hopeful to move to the porch by spring. And now wife is touchy because the oven went lukewarm and pizza night was almost a bust. Teenage daughter moaned her sweet potato fries were too crunchy. Time for the oven catapult.

Finally got down there. Nothing appears obviously wrong after pulling off the belt covers. Crank pully no wobble, crank sprocket is ok, woodruff key on snout is ok. Removed all the timing belt pulleys, all ok. All timing components were new OEM at last replacement 6 months ago, including the tensioner. The tensioner bracket is good, threads aren't stripped. All bolts were tight. Gonna take me a while to get it back together, work is so backed up. Probably going to seafoam the intake for good measure after it starts running semi normal, and do the idle relearn procedure and hopefully it "learns" how to run again lol. I'll check for magnetic bits on the crank sensor, thanks Texan.

Running awesome now for the past week. Put a new NGK coil in and a cheap MAP sensor because three ben franklins is too big of a hit for an OEM MAP imho. OEM from a parts yard would do, subies are rare in my neck of the woods though. I'm the biggest skeptic but honestly it hasn't running this well in...ever. No more pre ignition or wandering idle. New wires and plug swap didn't help this problem. So bottom line is: 0. disassemble throttle body and clean throttle plate, so nasty 1. clean intake with seafoam spray 2. can of seafoam in gas 3. idle relearn procedure 3.5 new coil 4. new MAP Kudos to all who got me through this.

If the crank pulley was wonky that could make a clunk, but the crank sensor is timed off the crank sprocket and I can't figure on that backing off the snout. Still worth a look, will pull off the timing cover and check it out.