I have a 1997 Legacy outback, Auto, 236K on it, with a 2.2 swap, with an odd issue. If the car sits overnight, especially if it is wet and damp out, which, being that i live in the northeast, is common, after i start it and drive it, as it starts to warm up, the engine will shake, the CEL will flash, and then, the only way to stop the shaking is pull the car over, and let it sit for 1 to 5 minutes, then, the shaking will go away, but the CEL stays on, until i have stopped and restarted it 3 times, and then start all over the next day. I have heard everything from an intake leak, which, if so, wouldnt that affect both cylinders 2 and 4?, to an EGR problem, to a coolant problem, to stuck injectors, plug wires, coil, and everything in between. I have already replaced the following, with known good working or new parts: coil, plugs, wires, injectors, timing components (this issue was there before and after the timing component replacement), ignitor, as well as all the gaskets and seals, with the exception of head and intake gaskets. The engine that is in it came out of my old 1996 legacy l, and had that issue sporadically before the swap, but, the last 3 months or so, seems to have gotten worse. It doesnt lose oil, so i doubt a valve seal or anything internal is causing it. Two local shops, as well as the local dealer, have said they have no idea of what the exact issue could be, which is why i came here. Any advice or help will be greatly appreciated.

what brand plugs/wires?

what's the gas mileage - same as always or...??

 

need to read the codes and post exactly what they are.  you mention cylinder 2 and 4 misfire but only indirectly.

swap plugs and wires to the opposite side and see if the misfire moves.  i understand you've already replaced them, but I would want to verify.

 

check valve clearance. 

 

swap injectors to the other side

 

checking timing marks/tensioner.

Here's what I was told by a master Subaru tech about our 96 2.2.

 

The #4 cylinder's exhaust valve can stick even after you've warmed it up.

 

Ours would cough and wheeze etc as we'd get up onto the highway at our local entrance ramp. Same deal , pull over and let it correct then it would be fine.

 

Sorry to add another theory. Good luck with it.

 

Also , the coil pack could have a minor crack in it too. Seen those balk under load too.

MAF Sensor

My guess is the coil but you already changed that but I say it sounds like the coil was it changed with a new one or one off another engine

Coil was replaced, twice, once with a known good used one, and then when the issue kept coming back, with a new one. The code that comes up is strictly for a #2 misfire (P0302). Spark plugs are NGK, wires are the same performance wires i use on all my cars, including my wifes rally car, injectors are new bosch, with no change. I myself am leaning towards ignition components, or, at the very least, water getting on the ignition components, due to how it goes away faster or slower depending on how damp and cold it is outside, so, im wondering if it isnt somehow leaking through the factory hood scoop, even though i never see water under the hood. I have been chasing this issue since July of 2015, when the engine was in the old car, with no luck.

if it's correlated to one cylinder and humidity/water then i'd definitely be combing over the ignition components.

 

no other issues, check engine lights that might be pointing to something else?

 

if it happened in another car then is it safe to say the hood scoop isn't the issue? or did you also swap hoods?

 

my 1988 XT6 has no proper drainage for the hood scoop and no issues yet in the few months i've had it.

Edited January 26, 20179 yr by grossgary

The 1996 i swapped the engine from, and the 1997 it is in now, both had the same style hood, and the hood on the 96 had some very bad rust holes around the edges, which is why i didnt rule that out as a possibility. The 97 doesnt have any rust, anywhere, but that doesnt mean that the seal or whatever it is around the hood scoop isnt bad, or one of the hood seals isnt, even though they look good.

 

Every time it happens, the CEL only shows the P0302 code, and never any others (in fact, in almost 12 months of driving it, that is the only code that has ever shown up, which is why it stumped the dealer).

 

It isnt directly related to cold and humidity, as it will also do it at times, although not nearly as frequent, in the summer, but in the summer only seems to do it if i step on it a bit hard from a stop, or, if it is under load going up a steep (ish) incline, which i forgot to mention.

Edited January 26, 20179 yr by biggman100

Same as you , when we mashed the pedal or under load. In our case it was the same uphill on ramp and same spot it would start shaking etc.

 

Maybe I'm wrong on which number cylinder.

 

Lots of theories! Please post all progress.

Good luck.

So it's a 96 engine?

 

Should have hydraulic valve lash adjusters. I would guess one of the adjusters on #2 is clogged or sticky and it's hanging the valve open.

 

Or if it has solid adjusters, possible the clearance is too tight, but normally that would just end up as a burned valve.

can the #2 plugwire be swapped with another cyl? same for plug and injector.

 

if the misfire doesn't move with any of those parts - compression and leak-down test next???

Edited January 27, 20179 yr by 1 Lucky Texan

Its a long shot, but try putting some dielectric grease on the plug wires, and coil connectors. Sounds like moisture is getting into the electrics somehow. Just for giggles, you could put the grease in the cam/crank sensor plugs too

That's what I'm talking about.

Do you hear a spark knock while it's running rough? When it does act up, try spraying the plug wires with a soapy water solution. This will help show if there is a crack in the wire or coil pack.

 

If it's truly a lifter issue, try adding some Marvel Mystery Oil to the engine during the next oil change. It may also be a good idea to run an engine flush through the engine before you change the oil.

I dont seem to hear any kind of spark knock, the car just starts to vibrate under any load once the CEL starts flashing, and then once the CEL becomes constant, the shaking goes away. To be honest, it could be oil related, since im not always diligent about maintenance, coupled with the fact the car i got the engine from, i bought from someone who almost never did any kind of repairs, as evidenced by the multitude of oil leaks i had to fix before i put it in the other car. As for engine flush, im leary of doing that, since it was proven that engine flush was directly responsible for blowing up two engines, one of mine, one of a friends.

With an engine flush product, you warm the engine up, drop the bottle of flush into the engine oil, run only at idle for 30 minutes and proceed to do an oil change. Once the oil is drained, substitute one quart of MMO for one quart of oil.

 

I do this constantly on Subarus all day. If done properly,you shouldn't have any issues.

The only time  I've had a po30? Code and not had it be the coil was oil in the spark plug tubes I changed the seals on the valve cover and it stoped check your plug wires is there oil on the end that goes in to the cylinders use a light mist of water under the hood while its running at night and look for spark leaks other than that dielectric grease all your electrical conectors and cross your fingers

 oil in the spark plug tubes I changed the seals on the valve cover and it stoped

 

his doesn't have spark plug tubes or seals, the original engine does, but the swapped EJ22 does not. 

Is the computer in the 96 Outback the original computer (the 2.5 ECU)?

 

If so, 96 was a special year for the Outback and required premium only fuel and was fairly aggressive in it's knock control. Since the 2.2 has lower compression than the 2.5, it might be picking up false knock and assuming it's a misfire. I'd unplug the knock sensor and see if it smooths things out. Yes it will pop a CEL, but that's ok during testing.

 

Regardless of the computer, I'd get a "cold" compression test in on that cylinder. If you have sticking valves or sticking rings, it will reveal itself as low or inconsistent compression.

Is the computer in the 96 Outback the original computer (the 2.5 ECU)?

 

If so, 96 was a special year for the Outback and required premium only fuel and was fairly aggressive in it's knock control. Since the 2.2 has lower compression than the 2.5, it might be picking up false knock and assuming it's a misfire. I'd unplug the knock sensor and see if it smooths things out. Yes it will pop a CEL, but that's ok during testing.

 

Regardless of the computer, I'd get a "cold" compression test in on that cylinder. If you have sticking valves or sticking rings, it will reveal itself as low or inconsistent compression.

Is the computer in the 96 Outback the original computer (the 2.5 ECU)?

 

If so, 96 was a special year for the Outback and required premium only fuel and was fairly aggressive in it's knock control. Since the 2.2 has lower compression than the 2.5, it might be picking up false knock and assuming it's a misfire. I'd unplug the knock sensor and see if it smooths things out. Yes it will pop a CEL, but that's ok during testing.

 

Regardless of the computer, I'd get a "cold" compression test in on that cylinder. If you have sticking valves or sticking rings, it will reveal itself as low or inconsistent compression.

The car is a 1997, that used to have a 2.5 in it, but now has the 2.2 from a 1996. I have done many of these swaps, and anytime it is a knock sensor issue, it shows a code for a knock sensor, as well as for a misfire. As for a compression test, the car is now sitting due to a completely destroyed transmission, so i am just going to replace the lifters while its apart in a couple weeks.

An update on this. I didn't do anything with the lifters, at least not yet, since it has been in the 50's and 60's here. and not once in the last 3 days has it acted up, until today, and it only happened later in the day, after it started raining out, so this issue seems to be related to ignition components and moisture, so my next step is to replace the plug wires, and if that doesn't solve it, i am going to replace the coil with an MSD street coil. I use them on the street drag and ice racing cars, and in 10 years, have never had one act up, even in snow, ice, and freezing temps.

Subaru or NGK wires, but I'm sure you know that.