keeweechris

Thanks man, much appreciated.

Thanks everyone for your help. Turned out to be #3 coil. Running smooth as silk now.

This is promising... last fkn cylinder (#4), swapped coil and plug, idling mostly normal... no shudders and dropping rpm. Will take for a test drive tomorrow. All cylinders 150psi compression, which is good. Is there any harm not having the fuel filter core inserted for a short blat?

Doesn't seem to run any differently, which is weird. Also, if I try to start it without MAF plugged it, it fires up for a second or 2, then stalls.

Thanks man, yeah, it seems like a bit of mahi, but hopefully I crack it. Got cylinder 2&3 done (compression 150psi, no change when swapping out coils and plugs), then got bored and tried the fuel pump, haha. Will get onto 1&4 tonight. Tried removing alt belt, but it still ran bad without the alt. Thanks for the idea though. Thinking about it this morn, it seems like something not right with closed loop system. It gets going well on boost and above 2000rpm. So I've got to understand whats involved in closed loop running... O2 + TPS + MAF I guess. But I've swapped all those... scratching head a bit.

Swapped fuel pump and removed fuel filter... no change. Jeees. Gotta be coils, plugs, or injectors! Or headgasket or valves.

Just a regular stock GTB twin turbo engine bay... stock BOV, stock airbox (K&N panel), everything clean a tidy, replaced intake pipe 3 months ago, and all vacuum lines. Was running very nicely after that.... smooth and 11/15psi boost. Yeah, all subaru's I've owned have had that slightly irregular idle, where it shakes ever so slightly every so often. But this is not that. It shakes constantly, changes exhaust tone, and misfires/stumbles on partial throttle anywhere between idle and 0psi. Then under boost it takes off and has no noticeable problems. I'm wondering now if it has some headgasket or valve issue. Leakdown will reveal that I'm hoping. Doesn't seem to be knocking. Data for "Knock Correction Advance (degrees)" and "Knocking Signal #1 (On/Off)" was all zero's for the time I logged it on cold idle the other day. Gotta find time to swap out the coils and plugs and injectors systematically, to rule them out. And fuel pump. Hoping it's one of those!

rpm fluctuates (like a misfires every second or so), and hesitations when accelerating or coasting from idle up to 0psi. On boost it appears to be ok.

Nah, stock BOV. Yeah, that's why I logged data from the ECU. I seemed to be getting realistic data from the MAF sensor, but keen to hear your opinion. Here is MAF data on cold idle: And here is the O2 sensor & RPM. It's all over the place, and shows the idle speed flapping around too. I logged heaps of data, but it's hard to show all on one graph. Let me know if you want to see any other data. Cheers Chris

Cambelt all good still... Well, all cams line up. Can't see crank, but seems unlikely that it skipped a tooth, being wrapped in cambelt 270degs. Will focus now on coils, plugs, injectors, comp/leakdown, then fuel pump.

I haven't tried starting it without MAF though, but was curious how to force it into Open Loop. Are there map values for lower revs? I'll try this anyway, because if the bad idling (misfiring) sticks around, I guess it proves it's mechanical, fuel, or sparks, not sensor-related..?

Haha, yeah, bit of a gamble. Came off a BH5B (same engine) at 255,000km. No, comp and leakdown is on the list. No smoke/smell from oil filler, and only a moderate puff puff puff of air, so rings are prob ok I'm guessing? I'm pretty new to this. I'll have to get set up for a leakdown test.

Well, I did open one cam belt cover to check the belt wasn't damaged, but didn't check if it had skipped. Good idea though. Can you check without talking the centre cover and pulley off? I kinda figured because there was a very clear point at 0psi where it started running well, and it boosted nicely, that the cambelt/timing was all good. Never experienced a belt with 1 tooth off before. And the problem is getting slowly progressively worse. But hey, no harm in checking, thanks! I've got a cheap SCA boost gauge, and that looks to match the readings the ECU is seeing. What do you mean by "borderline"? Is -60kpa/18inHg not quite right? What's it a sign off? Smoke tested again yesterday, and couldn't find any leaks. Yeah, no change after swapping to pick-a-part MAF. The MAF's (mine and pick-a-part) look very clean though. So I figure the problem is somewhere else. Gotta try the easy stuff first though, right.

Cheers, yeah, I did that. Each injector I unplugged caused the engine to change the way it ran, but surprisingly it did keep running, just. Yeah, I was hoping removing 1 injector wouldn't make any difference. I got a coil and injector from pick-a-part yesterday, so will spend the week swapping each out in turn. Fun job huh.

Changed MAF, crank & can sensors, no difference. On to spark plugs and coils now.