'13 Legacy 3.6R Cyl 3 Misfire

[SchwarzeEwigkt]

So, for the last week or so, my car has felt a little lumpy at idle from time to time. My mechanic's spider-sense indicated "miss," but I didn't have any lights and it would drive fine. I thought maybe it was a bad tank of gas or something, so I grabbed a can of fuel injector cleaner to put in the next tank. I don't drive it much and wanted to run the tank down before I put it in, so it's still in the cupholder...

Well, today, my wife borrowed my car because she needed to get to work while her car was in the shop and on the way home, the dashboard lit up like a Christmas tree. You know, like Subarus do when something's wrong. Pulled the codes and all that's there is a P0303 - Misfire cyl 3. Fired the car up and it's definitely misfiring; it's lumpy as hell. Cleared the code and let it idle for a minute to see if it'd reset the code, trying to prove it was still cyl 3. Code came right back.

I haven't torn it apart yet — it's only been like half an hour, so it's still hot — but I'm brainstorming. It looks fine under there; no wire harness damage I can see and no oil leaks in that area. Given that the car has about 110k on it now and has the original coils, I figure it's probably a bad ignition coil. I had new spark plugs put in about 10k and a year and a half ago, so I doubt it's that. I should have had the mechanic I paid to do the plugs peek at all the coils, but I forgot. Might've been able to see this coming.

Because it's a bit of a pain in the ass to take that side of the engine compartment apart, do you guys know of any test procedures for the coils? Resistance checks, etc? I'm reading about coils corroding and delaminating, so I'll check for that too.

I'm trying to avoid taking it apart and putting it back together multiple times. If there's nothing obviously wrong with the coil, I guess I'll switch the #3 coil to #5 and see if the misfire follows it, particularly since a coil's about a week out if I order one. I'm not enabling the local stealership by paying their prices for parts.

Other thoughts would be appreciated too, should you have any.

[Scubaboo]

There are tests you can make with a meter, at least for BMW coils but I assume Subarus are the same in that regard. 

However mostly it was swap and see, so I didnt ever do them personally.  But the coils were easier to reach in those.

The meter tests and other diagnostic procedures you can do may be in the vacation pics somewhere here.  I forget where they are, sorry.

Hopefully someone with more experience will chime in soon.

EDIT: 

I also remember experiencing misfires due to leaking valve cover gasket spark plug grommets leaking oil into the plug area.  Misfires were multiple though. (BMW, not Subaru)

Thinking about it, In your situation I would swap out the coils so I can check for oil in the area and spark plug condition.  It would suck equally to test the coil, order it, then find out you need to do more stuff then you thought.  

But...it's probably just the coil IMO.

Edited June 29, 2023 by Scubaboo

[silverton]

If it were me I'd swap the coils with another cylinder and see if it follows, it looks more intimidating than it is.  Plugs and coils are far easier on a 3.6 than a turbo EJ. if it doesn't follow it's either the spark plug or the injector but I've also seen a bad ECM cause this issue.

[SchwarzeEwigkt]

3 hours ago, silverton said:

If it were me I'd swap the coils with another cylinder and see if it follows, it looks more intimidating than it is.  Plugs and coils are far easier on a 3.6 than a turbo EJ. if it doesn't follow it's either the spark plug or the injector but I've also seen a bad ECM cause this issue.

Yeah, I’ve got a whole week before my new coil shows up. Sounds like this weekend I should swap the coils and see where the misfire goes to. A spark plug I’d mind less than the other things. Replacing an injector would be a pain in my ass.

I really hope it’s not an ECM. That ain’t in the budget. Although, you’d think that a failing ECM would do other weird stuff too and it isn’t, so there’s hope there.

[SchwarzeEwigkt]

5 hours ago, Scubaboo said:

There are tests you can make with a meter, at least for BMW coils but I assume Subarus are the same in that regard. 

However mostly it was swap and see, so I didnt ever do them personally.  But the coils were easier to reach in those.

The meter tests and other diagnostic procedures you can do may be in the vacation pics somewhere here.  I forget where they are, sorry.

Hopefully someone with more experience will chime in soon.

EDIT: 

I also remember experiencing misfires due to leaking valve cover gasket spark plug grommets leaking oil into the plug area.  Misfires were multiple though. (BMW, not Subaru)

Thinking about it, In your situation I would swap out the coils so I can check for oil in the area and spark plug condition.  It would suck equally to test the coil, order it, then find out you need to do more stuff then you thought.  

But...it's probably just the coil IMO.

I had a coil pack up on me in an old Audi I had before this Subaru. That was super easy to diagnose because Audi provides a ton of instrumentation; you can easily log the misfire count per cylinder. Plus, the coils are right on top, given it was an I4. Coils were cheap, too…like $15. Good thing, though; you do a lot of diagnostics on a car put together like that…

I did check my copy of the vacation pics and didn’t see anything about direct coil diagnostics, just stuff about checking the voltage to it and the wiring harness.

I doubt it’s oil contamination given that the whole area around it is dry as a bone. Maybe, I guess; I’ll know when I pull it out.

Well, whatever. Sounds like I know what I’m doing Saturday morning.

[silverton]

1 minute ago, SchwarzeEwigkt said:

you can easily log the misfire count per cylinder.

Subaru also provides a 'simple roughness monitor'.  But I don't know if a generic scanner can read the PIDs. Injector's aren't too bad, but everything is intimidating if you haven't done it before.

Good luck!

[SchwarzeEwigkt]

19 minutes ago, silverton said:

Subaru also provides a 'simple roughness monitor'.  But I don't know if a generic scanner can read the PIDs. Injector's aren't too bad, but everything is intimidating if you haven't done it before.

Good luck!

I’ve got a USB to OBD-II cable and a copy of FreeSSM kicking around. I suppose I could fire all that up and see what it’s got. My generic code reader is a Bluetooth adapter that talks to an app on my phone that reads lots of PID’s, but it didn’t seem to list anything about misfire count. Thing is, the misfire is bad enough that the ECM detects it and starts about setting a code pretty quick. It went from cleared to pending in about 30s. I’m unlikely to need any fancy diagnostics in this state.

You’re probably right about an injector. I’ve never done it and certainly not on this engine, but I’ve seen plenty of videos and it doesn’t look too hard. I guess I’ll cross that bridge when I get to it.

[Scubaboo]

5 hours ago, SchwarzeEwigkt said:

I had a coil pack up on me in an old Audi I had before this Subaru. That was super easy to diagnose because Audi provides a ton of instrumentation; you can easily log the misfire count per cylinder. Plus, the coils are right on top, given it was an I4. Coils were cheap, too…like $15. Good thing, though; you do a lot of diagnostics on a car put together like that…

I did check my copy of the vacation pics and didn’t see anything about direct coil diagnostics, just stuff about checking the voltage to it and the wiring harness.

I doubt it’s oil contamination given that the whole area around it is dry as a bone. Maybe, I guess; I’ll know when I pull it out.

Well, whatever. Sounds like I know what I’m doing Saturday morning.

Yup as mentioned by silverton you can get misfire data on our cars too.  In my case I do it with Romraider and a Tactrix cable.  And yes I do a lot less diagnosing on the Subaru than a BMW lol!

Thanks for letting us know what the vacay pics say about this, I was wondering too.

I dont think oil contamination is it either as it is consistently on one cylinder.  But I can tell you when I had misfires due to oil contamination the head and valve cover was dry and plug area was wet.  So since then I check for contamination even if there are no outward signs of leaks.

Let us know how it goes!

 

[SchwarzeEwigkt]

2 hours ago, Scubaboo said:

Yup as mentioned by silverton you can get misfire data on our cars too.  In my case I do it with Romraider and a Tactrix cable.  And yes I do a lot less diagnosing on the Subaru than a BMW lol!

Thanks for letting us know what the vacay pics say about this, I was wondering too.

I dont think oil contamination is it either as it is consistently on one cylinder.  But I can tell you when I had misfires due to oil contamination the head and valve cover was dry and plug area was wet.  So since then I check for contamination even if there are no outward signs of leaks.

Let us know how it goes!

 

Hmm. I’ll make sure to get my mirror out and look down there.

[SchwarzeEwigkt]

Seems like it was just the coil. Put a new NGK coil in and it’s back to purring like it normally does. Well, except for the noisy idler pulley bearing and the heat shield recently came loose, but ya know.

Seems like this coil was doing the fairly typical rust induced delamination thing. You can’t see it in the picture, but the one side where the transformer core goes into the part of the coil potted with epoxy looked especially bad and expanded. I expect that it probably cracked the windings inside or something. I’ll measure the resistance of the windings later just for fun. I’ll bet it’s probably somewhere between really high and infinite.

Thanks for the confidence boost for tearing into that part of the motor. I feel like a bit of a goof having paid a guy to replace my plugs after seeing how you get in there. At least it was just $120 and a 12pk of beer.

All that said, while I know this sort of failure is common, what I haven’t been able to get a read on is if they all tend to start failing together. Like, since I had one coil pack up on me, should I expect others to start to go on me soon?

Edited July 3, 2023 by SchwarzeEwigkt
Spelling.

[Scubaboo]

Thanks for getting back to us with what happened.

Since the usual failure mode is visible maybe you can do a visual check on the rest.

 

[SchwarzeEwigkt]

5 minutes ago, Scubaboo said:

Thanks for getting back to us with what happened.

Since the usual failure mode is visible maybe you can do a visual check on the rest.

 

Probably not without taking them out. They’re in recesses in the valve cover and the cables run around them.

[Scubaboo]

Yeah, kind of a hassle to do that.

IIRC in most of the stuff I read about people who've had to do coils, they get away with doing the broken ones.  I dont remember a lot of replacing entire sets.

 

[motorbreath]

I had a pretty bad misfire issue with #4 when I first bought my 2 .5GT, exact same rust/delamination issue. After changing the spark plugs and the #4 coil pack, all was good.

I did end up changing the other coil packs within a year because they looked to be in rough shape and I was getting a tune.

One interesting thing I learned is that the ECM doesn't directly detect misfires, apparently it infers them from crank and cam position sensors. So if you get occasional misfire events, especially at idle or off throttle, it's considered normal.

[creep_nu]

yeah i ended up replacing my entire set...with most things, once one goes the others follow. they're cheap enough that it was good preventative maintenance, and i keep one of the known good ones in the trunk as a spare...might as well.

[motorbreath]

Better safe than sorry, especially with Rock Auto's prices!

[SchwarzeEwigkt]

10 minutes ago, motorbreath said:

Better safe than sorry, especially with Rock Auto's prices!

Maybe, but coils are still $70 apiece. Keeping $420 worth of coils around just in case doesn’t really work for the budget.

Besides, my car has since COVID more or less become the village bicycle; other people in my family who need wheels because theirs are down for whatever reason put more miles on my car than I do at this point. Factor in that the coils are stocked in a warehouse only about 100mi away from me and show up in a couple days, I think it makes the most sense for me to just order one when it fails.

It seems likely that I’ll be ordering a couple in the medium term, but I’m treating that as a bridge-crossed-as-it’s-encountered kind of situation.

Edited July 5, 2023 by SchwarzeEwigkt
Numbers!

[motorbreath]

I guess my situation was different, I was getting a knock event at WOT in addition to the misfires, so new plugs and coils seemed prudent to safeguard the engine.

[SchwarzeEwigkt]

41 minutes ago, motorbreath said:

I guess my situation was different, I was getting a knock event at WOT in addition to the misfires, so new plugs and coils seemed prudent to safeguard the engine.

Yeah, that’s a situation where I agree it’s prudent. Your situation is a recipe for melted piston crowns and burnt valves. Fresh ignition components is a safe thing to have, even if it just checks off a box on the list of stuff that could be going wrong. In my case, I risk catalyst damage from unburnt fuel from one cylinder, but only if I continue to drive it for a while. Also, my dash lights up like the Vegas strip and half the car’s functions are disabled.