Diagnosing misfire

I bought an ‘08 legacy, with a 2.5i naturally aspirated engine with 200,000 miles on it, the guy I bought it from said it was rebuilt around 70,000 miles ago and had pictures of it in the process apparently. A couple weeks after I got it, the CE light came on for a cylinder 2 misfire. Changed the spark plugs, light came back for cylinder 2 again. it already had new NGK wires, so I replaced coil pack next, light came back. And last I changed the injectors, and cylinder 2 misfire code came back. At this point I’ve read that I should do a compression test next, (was going to do it tomorrow). I’ve read it could be a ring or burnt valve or something, but I’m not clear on what the results of the compression test would indicate, nor do I know enough about the inner workings to try and diagnose the problem. With spark plugs, plug wires, coil pack, and injectors all new, but cylinder 2 still misfiring, how would you guys diagnose the problem from here? Also, if it were an O2 sensor or a MAF that was causing the misfire, wouldn’t the ecu throw codes for those? It’s my first Subaru and I love it. I’m prepared for some expensive work from this point on, but don’t want to throw away money needlessly. Much thanks

I read another post of yours just now, and this stuck out to me.

"After the coil pack the CE light went away for a couple weeks but just came back today. not really sure what the next step is"

 

I remember buying ngk wires back in like.. 2014, and returned them because they didn't click/snap into the coil pack the way oem wires did, and it felt like they wouldn't maintain a solid connection to the coil pack. I would push them on with my fingers, and the boots would just slide back off the coil pack a bit. I'm also not sure if aftermarket coil packs have the the ring/groove to snap into either.

 

So I wonder if that's what's going on. Would hate to see you go and replace those parts and have it not fix the problem (oem plugs and wires, or an ngk coil pack like me which is oem subaru/diamond with ngk painted on it).

 

So what I might do first in your shoes is maybe clear the code, firmly push the #2 plug wire onto the coil pack, and see what happens. Looking at pictures online, I think #2 is the top spark plug wire going to the coil on the driver side.

 

I don't think our non-turbos are prone to ringland failure or burnt valves

Edited January 2, 2021 by apexi

I read another post of yours just now, and this stuck out to me.

"After the coil pack the CE light went away for a couple weeks but just came back today. not really sure what the next step is"

 

I remember buying ngk wires back in like.. 2014, and returned them because they didn't click/snap into the coil pack the way oem wires did, and it felt like they wouldn't maintain a solid connection to the coil pack. I would push them on with my fingers, and the boots would just slide back off the coil pack a bit. I'm also not sure if aftermarket coil packs have the the ring/groove to snap into either.

 

So I wonder if that's what's going on. Would hate to see you go and replace those parts and have it not fix the problem (oem plugs and wires, or an ngk coil pack like me which is oem subaru/diamond with ngk painted on it).

 

So what I might do first in your shoes is maybe clear the code, firmly push the #2 plug wire onto the coil pack, and see what happens. Looking at pictures online, I think #2 is the top spark plug wire going to the coil on the driver side.

 

I don't think our non-turbos are prone to ringland failure or burnt valves

 

See the wires to me seemed like a hint that the previous owner had been trying to solve the misfire problem and sold it because he couldn’t figure it out, or because it was going to be expensive. I actually triple checked that both sides of the wires were pushed on all the way, so unless there’s something fundamentally wrong with them Idk. I was still thinking of changing them anyways but I’m doubtful it’ll help. I’m relieved to know the N/A legacy’s don’t have those problems as often, but suppose I get OEM wires and it still has the misfire, how would you narrow down the potential problems? I’m assuming since the misfire is exclusive to cylinder 2 it couldn’t be a general sensor or anything, but more like an internal, mechanical part. If I do a compression test or a leak down test (?), I don’t know how to use the test results to figure out what the probable cause is. I’m going to go buy a compression tester and check that, I’ll post the results when I do.

You've done quite a bit. I don't know enough about this stuff to really give any suggestions. In your shoes I'd probably try to find an independent shop with good reviews, and put 1 or maybe 2 hours of labor time into diagnosing the problem. What state are you in?

Before spending more money on parts or paying someone else, I'd verify the signal to the #2 fuel injector. Very easy to do. In a pinch, you can do it with a cheap pocket oscilloscope like this:

 

https://www.amazon.com/Welded-Assembled-DSO138-Digital-Oscilloscope/dp/B082PYBP94/ref=sr_1_7?dchild=1&keywords=oscilloscope&qid=1609623693&refinements=p_85%3A2470955011&rnid=2470954011&rps=1&sr=8-7

 

Start the car, let it idle, and pull the fuel pump fuse to let it draw down fuel pressure. Leaving the fuse out, disconnect the #2 injector connector and stick a pin in each lead inside of it. Connect the scope to the pins. Have a helper crank the engine over and watch the signal. Check pulse width, peak voltage, and signal clarity (shouldn't be super fuzzy).

 

Then, repeat on one of the other injector connectors, and compare the results. You'll have to get acquainted with the basics of oscilloscope use if you aren't already, but it's pretty straightforward stuff.

 

You can also use the scope to check the signal to the coil, though trying to measure the voltage to the plug will probably fry the scope. You could try a spark tester at the plug. On a related note - how did the plug that came out of cylinder 2 visually look compared to the others?

Edited January 2, 2021 by awfulwaffle

Figured I’d update. Today the CE light just randomly turned off on its own. Not sure if it was the ecu taking that long to clear the code after the new fuel injector I changed a couple days ago, or if it was because I cleaned the connection points on the wire a couple hours earlier, but so far it hasn’t come back. Fingers crossed that it stays that way.

You can also use the scope to check the signal to the coil, though trying to measure the voltage to the plug will probably fry the scope. You could try a spark tester at the plug. On a related note - how did the plug that came out of cylinder 2 visually look compared to the others?

 

All of the plugs were fairly dark, but cylinder 2 was straight black, I assumed bc of a rich mixture, but who knows. CE light turned off today on its own, so Idk if it was the new injector or me cleaning wire connections. Hopefully it stays gone

You're making me wonder about the maf sensor not reading the amount of air coming in properly, and running rich, too rich. Wouldn't hurt to inspect and clean the resistors in the maf sensor. When you pull the maf sensor the air intake temp sensor is most visible (Pic 1), and the wires/resistors that meter the air are down inside the body of the sensor (pic 2).

 

Well yesterday while I was idling my car waiting for it to warm up the CEL came back on, cylinder 2 misfire again. I got some new plug wires to try out even though the pretty new NGK ones on it looked totally fine. If it comes back then cylinder compression test time.

You're making me wonder about the maf sensor not reading the amount of air coming in properly, and running rich, too rich. Wouldn't hurt to inspect and clean the resistors in the maf sensor. When you pull the maf sensor the air intake temp sensor is most visible (Pic 1), and the wires/resistors that meter the air are down inside the body of the sensor (pic 2).

 

 

See I thought the same thing at first bc of it being dark, however, if the MAF was bad, wouldn’t I get misfires in other cylinders occasionally? And wouldn’t it throw a code for something besides just a misfire?

I saw you reply in another thread that you're using plugs gapped at 0.044". Parts stores will sell you these because their system says so but they aren't the correct spec. On N/A models, the fsm calls for a gap of 0.039-0.043" (1.0-1.1mm). You want NGK FR5AP-10. These are gapped correctly out of the box.

You want NGK FR5AP-10. These are gapped correctly out of the box.

 

Yeah those oem ones are what I run but just minor correction, AP-11

Yeah those oem ones are what I run but just minor correction, AP-11

I was talking to OP sorry. But, the AP-11 is the part that is wrong. It specifies a 1.1mm gap out of the box which is at the large end of the tolerance set in the fsm.

 

The alternate part number is NGK 6371.

I was talking to OP sorry. But, the AP-11 is the part that is wrong. It specifies a 1.1mm gap out of the box which is at the large end of the tolerance set in the fsm.

 

The alternate part number is NGK 6371.

 

You've recommended the wrong plug in multiple threads. NGK FR5AP-11, 5463.

They come pregapped at around .041" every time I've checked the new ones I've bought.

You've recommended the wrong plug in multiple threads. NGK FR5AP-11, 5463.

They come pregapped at around .041" every time I've checked the new ones I've bought.

 

I understand what part number the fsm has but it's not actually the right gap that the fsm calls for. The AP-11 plug does work but the gap out of the box is at the end of spec. From the NGK specs, FR5AP-11 is gapped at 1.1mm (0.043") out of the box which is the end of the fsm spec. AP-10 is gapped at 1.0mm (0.039") out of the box which is at the beginning of the box.

 

I'm just offering a potential reason for misfires.

 

See the thread by TIC on nasioc here for more detail.

 

https://forums.nasioc.com/forums/showthread.php?t=1889917

I totally get your logic, especially when the gap will widen over time. I think both plugs would work fine.

I totally get your logic, especially when the gap will widen over time. I think both plugs would work fine.

I agree but again wanted to give a potential culprit (with weak ignition coil at high mileage) even though it's most likely something else.

I saw you reply in another thread that you're using plugs gapped at 0.044". Parts stores will sell you these because their system says so but they aren't the correct spec. On N/A models, the fsm calls for a gap of 0.039-0.043" (1.0-1.1mm). You want NGK FR5AP-10. These are gapped correctly out of the box.

 

I don’t know where everybody finds gap size but I checked subaru’s website and that’s what it said for the 08 legacy 2.5 non turbo. and it has a brand new ignition coil for a 4 cylinders, and if it was a gap problem I would get misfires on cylinders other than just #2. If you can send me a link to the correct gap info though I would appreciate it

I saw you reply in another thread that you're using plugs gapped at 0.044". Parts stores will sell you these because their system says so but they aren't the correct spec. On N/A models, the fsm calls for a gap of 0.039-0.043" (1.0-1.1mm). You want NGK FR5AP-10. These are gapped correctly out of the box.

 

Straight out of my owners manual: FR5AP-11 (NGK) non turbo models. gap is 0.044

Straight out of my owners manual: FR5AP-11 (NGK) non turbo models. gap is 0.044

I understand the owner's manual has the FR5AP-11 as the part but the factory service manual calls for a 1.0-1.1mm (0.039-0.043") gap is all I'm saying.

I don’t know where everybody finds gap size but I checked subaru’s website and that’s what it said for the 08 legacy 2.5 non turbo. and it has a brand new ignition coil for a 4 cylinders, and if it was a gap problem I would get misfires on cylinders other than just #2. If you can send me a link to the correct gap info though I would appreciate it

In the factory service manual, go to "Engine", choose the "H4SO" engine, then choose "Ignition" and finally, "General Description".

 

In the factory service manual, go to "Engine", choose the "H4SO" engine, then choose "Ignition" and finally, "General Description".

 

 

Idk man, you totally may be right. From my perspective though, every auto parts store, along with the manual for my exact car says to use a plug that comes gapped at 0.044, vs a factory service manual that lists a plug that’s gapped at 0.044, but then says 0.039-0.043. Both manuals say to use the same plug, but the FSM lists a different gap from both the owners manual and the plug they say to use. You even recommended using a plug different from their recommendation to correct them, so I find it hard to put my faith in with they say on there. Either way, I doubt a 0.001 gap difference will have any affect. I do appreciate trying to correct me before something potentially gets messed up though

I will admit you’ve made me doubt it enough that I plan on calling a subaru dealership tomorrow to ask them and clarify once and for all, LOL

Edited January 7, 2021 by Stphnhrrng

So, in diagnosing the misfire, have you tried swapping parts to other cylinders to see if the misfire code moves to another cylinder? That's usually my first step in diagnosing a misfire. I know you changed a lot of these parts, but it still might be worth it to try. At least that would rule out if one of the plugs/wires/injectors isn't working properly. Next I would do a smoke test. I understand that it's NA, but vac leaks affect NA cars too. I'd also run a can of seafoam through the intake to see if that changes anything. 200K could have a bunch of carbon buildup on the valves. Just my thoughts on cheap or free things you could try before firing the parts cannon at it.

So I got fed up trying to chase down said misfire, took it to the Subaru dealership. They couldn’t find anything wrong with it. Did swapped all plugs, wires, injectors ect, nothing. They did a compression test, Cyl #1: 180psi, Cyl #2: 197psi, Cyl #3: 175psi, Cyl #4: 205psi. They noticed the compression varied a bit so tey did leakdown, no psi jump wet. Checked valve clearances, all good. Pulled timing cover, all good. Said at this point the only things they can think of are something wrong in the cylinder, or a ECM problem. Any of you have any bright ideas? Or any idea how to check if the ECM could be messing up?

Side note: Since I got it back, whenever I press in the clutch to come to a stop the rpm’s drop to about 500 and it even died once. It’s happened a couple times before I took it in but it’s a lot worse now. Does anybody know what might cause this? Or if not, could I just buy a tuner and set the idle control to 900 and would that help? Also thinking about getting a tuner just to get rid of the CEL light if I can’t find a way to fix the problem.

Side note: Since I got it back, whenever I press in the clutch to come to a stop the rpm’s drop to about 500 and it even died once. It’s happened a couple times before I took it in but it’s a lot worse now. Does anybody know what might cause this? Or if not, could I just buy a tuner and set the idle control to 900 and would that help? Also thinking about getting a tuner just to get rid of the CEL light if I can’t find a way to fix the problem.

 

That's normal on our non-turbos after the battery has been disconnected/ecu reset, it'll go away in a few days or so, depends how much driving you do.