The two back by the starter spring to mind first but it sounds like you found those. I think they're on the intake manifold not the actual head but I'm going off the GD and older cars so I could be mistaken. I think there is at least one on one of the bellhousing bolts so pulling the IC would be a good move to be extra thorough.

 

I missed the two off the back of the LH TGV previously, they were well camouflaged by oxidization. I've had the IC off already, but it's tough to see through all the wiring, plumbing, and pieces parts around the back of the RH head.

nothing wrong with adding a few grounds I'd guess - but I also have seen them on the IM, not on the head. I have looked around and refreshed a lot of grounds, but perhaps not as diligently as other people.

does your car have wires or coil-on-plug? either way, swap the coil/wire to another cylinder.

Not looking to add any grounds, though I may end up making a grounding harness depending on where I end up after trying to refurb the factory wiring. The additional grounds I added to my 98 Legacy GT made a surprisingly big difference in smoothing out throttle response and snappiness, but this XT has quite a few more grounds (and a fair bit more control devices/sensors wired up in the engine bay).

Coil packs on this car. Swapping them around did not move the misfire. Fixing a couple vacuum leaks definitely smoothed out the idle, and reduced fuel trim at idle. I'll swap around injectors this weekend. Also need to test cam and crank sensors. MAF voltages look a little high, but I'd expect a bad MAF sensor (which I did clean a few weeks ago) would effect all cylinders. Might just be the way the ECU determines what's a "misfire".

Starter ground.

Intake/TGV ground

Both sides of the heads...ground.

Small neg battery ground.

Ask @plain OM for the rest....

Got all those, thanks. TGV area is LH side only, or do you know if there's one on the RH too?

 

nothing wrong with adding a few grounds I'd guess - but I also have seen them on the IM, not on the head. I have looked around and refreshed a lot of grounds, but perhaps not as diligently as other people.

does your car have wires or coil-on-plug? either way, swap the coil/wire to another cylinder.

 

Starter ground.

Intake/TGV ground

Both sides of the heads...ground.

Small neg battery ground.

Ask @plain OM for the rest....

 

Just responding to the "mention" . . . I'm not sure I can help much here with the grounds. I have only the same FSMs to work from, and they can well have inconsistencies, as was noted.

But I was wondering . . . other than the misfire being reported by BtSSM (http://www.subaruoutback.org/forums...109-gen-3-2005-2009/377385-can-t-figure-out-cylinder-1-misfire.html#post4360250), are you experiencing any actual signs of malfunctioning? (Wondering if the discussion might be broadened to other factors, such as traildogck's suggestion regarding fuel pressure, or as in the linked thread.)

 

wonder if the engine would run even that well if the crank sprocket were off a tooth? might explain universally low compression?

maybe double check timing if no luck with the current investigation?

 

I can't say how it would run being off a tooth on the crank, but most likely not well at all. Compression is compression, regardless of anything else. It's just how well each individual cylinder builds pressure.

Timing at idle looks good, hovers around 12°.

 

sorry, cam/crank timing - not ignition

exhaust is popping

if valves are very slightly open at TDC, you'd get lower pressures, and maybe the odd exhaust sounds you mention? - but , as you say, perhaps worse-running engine than you seem to have.

 

Got it, thanks for clarifying. I have not verified that timing belt placement is perfect but for what it's worth, the car does pull pretty hard, as much as I want to push it while I know it's not running 100%. My perspective is handicapped since I never drove this car new, and I've never driven another XT to have a comparison.

Might end up being a visit to an expert if I can't sort it out myself, as much as that feels like defeat.

 

how long since the timing belt system was serviced?

checking the timing is not too difficult - you can usually find pics, maybe even videos, that would help if a search for instructions isn't fruitful. But, on a cold engine, your numbers are really not that low plus, the difference is only 9 from low to high. I THINK ideally 7psi is what folks like to see. Just proceed with the leakdown - maybe you have a burned valve?

other outlier issues to consider; 'dropped' valve guide, failing crank position sensor?

 

PO had a Gates kit installed at 108K, I'm at 131K now. Belt condition looks great.

I will definitely check timing this weekend, easy enough to cross it off the list.

Engine was warm when tested, not hot to the touch like when I was pulling coils and plugs, expecting my buddy to arrive any minute. I don't know enough to say if that temp difference would matter or not, but I'll find out when I re-test. Agree that my comp numbers are decent and fairly consistent. Yes, 7 psi range is the spec. I was expecting way lower in cyl 1, so I took that as a little bit of a win.

Bad valve or similar is definitely on my radar, in the "thar be dragons" area of the map. Crank sensor gets tested this weekend.

 

Stationary and idling or stationary and revved up to 1500/2000 rpm, you can hear it and feel it misfiring. Worst is when stationary and at around 2000, the car shakes and exhaust is popping,

With BtSSM (or other SSM-based scanners such as Romraider) you should have access to a wide range of data from the ECM. If there's noticeable symptoms at idle, recorded logs might reveal an anomaly. The FSM has a table of the SSM data that should be available. It's in the Engine (DOHC) Diagnostics section, in the chapter "Subaru Select Monitor", under "Normal Mode". The preset list of "fields" in BtSSM has some of these, but all can be manually added to the Fields list.

 

It sounds like a lot of rabbit holes are popping up. Logging would be good, fuel pressure check would be good. Clean fuel filter verification would be good.

I would revisit the vacuum/boost leak area again. Use the attached as a reference.

 

Lots of rabbit holes, for sure. Idle fuel trim is around +9% so I wouldn't be surprised if there are more leaks to be found. I smoke tested with the engine cold, I'd better try with it warmed up to see if leaks appear as vac lines soften. I used that same sheet when testing -- thanks for posting it again.