Cylinder misfire AGAIN in different cylinder after changing plugs and wires

[cascadeclimbn]

My 96 subaru impreza outback would lose acceleration power on inclines (and sometimes flat road) at highway speeds. I had 3 check engine lights: a knock sensor, map sensor, and cylinder misfire detected in cylinder 1. I replaced the knock sensor, hooked the map sensor back up properly and replaced plugs and wires with NGK plugs and Duralast wires. Reset the check engine light and haven't had any problems in the last 2 days after driving 300 some miles on a road trip through all types of terrain and speeds. Till today...was driving home from school on the freeway and the car lost accelerating power the check engine light started flashing and I had to pull over. After five minutes and starting the car twice I drove a few miles to autozone where they pulled the code and there's a misfire in a different cylinder than before!! Cylinder 3. Why would these be happening after new plugs, what could be causing it, and how do I go about diagnosing and fixing it?

[Fairtax4me]

Did you get standard copper plugs?

Have you made sure the wires are firmly clicked onto the plugs? How about into the coil pack?

[Rooster2]

I would bet the farm that the problem is the "Duralast Wires." Years back, I bought Duralast wires from Autozone. They started misfiring shortly after installation, and I concluded that the wires were of poor quality. Recommend you return the Duralast wires for a refund, then apply the refund to the most expensive set of wires that Autozone sells. I bet your problem will be solved.

[cascadeclimbn]

I would bet the farm that the problem is the "Duralast Wires." Years back, I bought Duralast wires from Autozone. They started misfiring shortly after installation, and I concluded that the wires were of poor quality. Recommend you return the Duralast wires for a refund, then apply the refund to the most expensive set of wires that Autozone sells. I bet your problem will be solved.

I was kind of thinking it was the wires as well. Or at least that should be the first place to start. Would Bosch wires work? They are like the mid level ones that Autozone sells...the NGK set is $100 which I can't afford right now. If it's not the wires what else could it be?

[cascadeclimbn]

Did you get standard copper plugs?

Have you made sure the wires are firmly clicked onto the plugs? How about into the coil pack?

I'll double check my work but I'm pretty positive I did everything solid. The plugs are bought aren't copper I don't think...they are NGK V-Power...the best ones autozone had.

[Olnick]

For plugs all you need is the most basic copper--NGK BKR6E-11, I believe.  Maybe $2 each.

 

For wires, get Subaru OEM.  Check prices at a dealer or online.  $100 for a set sounds ridiculous to me--it's been a few years, but the last set I bought was around $35.

[tbolt1003]

My daughter's 99 Forester was running poorly. CEL was on, showing a misfire on cylinder #4. I did new NGK plugs and high quality wires. Still a misfire on #4. Turned out the coil pack wasn't giving spark to #4, but 1-3 were fine.

[lmdew]

As listed above NGK Copper core Plugs and Subaru or NGK wires.  

 

First step.  If it turns out you need a coil, I have several Subaru coils.

[Rooster2]

I was kind of thinking it was the wires as well. Or at least that should be the first place to start. Would Bosch wires work? They are like the mid level ones that Autozone sells...the NGK set is $100 which I can't afford right now. If it's not the wires what else could it be?

Bosch wires should work okay. I am thinking that is what I am running on one of my Subies. If doesn't fix the problem, you can always return the wires. Autozone is really good about accepting returns, providing you have a recent receipt, and product is clean, and can be resold in it's original box. Since the misfire moves around, when you move the plug wires around, I still think it is your wire set. I don't think your coil pack is the problem.

 

My experience with Autozone, is that their house brand products, and all other lowest priced auto parts they sell should be avoided. Quality is generally poor, with short life span. Quality is much better with name brands, and higher priced, high end priced parts. I learned my lesson when I bought their Kelly-Springfield fan belts. They wouldn't break, but would stretch. Every 6 months it seemed, I was tightening the belts. I finally wised up, and bought their best belts, either Gates, or Good Year Gator Backs, and got out of the belt tightening routine.

[tirod]

If you are chasing a CEL, that is all it is, a code. Since it's moving, then something up the chain in the ignition could be at fault. If there is a misfire with a brand new set of plug wires, I leave that to the quality of the product. You get what you pay for. It does seem odd they would fail in just a few days. 

 

Were the plugs bright plated? If so, antiseize added to the threads can contribute to a loss of conductivity and the system will see them misfire. It's a point of controversy with countermen and corporate programs to sell more add on items to the ticket. Plated plugs don't need antiseize, but corporate sales mongers insist we push it. The customer then uses something that contributes to more problems - like those who smear "brake grease" on the pads because it's not real clear on the package where it goes. It just shows a brake pad. It's caliper pin grease, and marketing is at fault. 

 

Subaru has some programming inherent in their system that takes a code and makes things more difficult. I see it as a JDM response to things, force the repair early on because of a much lower tolerance to things being out of order. Misfire codes in other vehicles don't force it into limp mode nearly as quickly. 

 

As for belts, they do not stretch to the extent many believe. It's contrary to their construction to have the fibers continually relaxing and making it longer and longer. The number I have seen over the counter, and in conversations with industry reps, leads me to conclude they aren't stretching as much as getting thinner. In the older V belts, they lose width and drop down into the groove deeper, which is fixed by moving the pulley further out, and they drop down more, until they finally are bottomed out. They lose the frictional resistance of driving on the angled sides of the pulley when they are seated on the bottom of it. Same for serpentine belts, just the opposite direction, the belt finally bottoms on the top of the ribs and the multirib V's lose grip at that point. In either case, they aren't jammed into the V providing friction - they are sitting on a part of the pulley and lose grip. 

 

That doesn't mean that belts can't be made to stretch, the newer cars now use them to eliminate having an adjuster in the drive. You have to force them over the pulley to install them. It's a distinctly different kind of construction that does allow some stretch, as opposed to the 80 year trend to prevent it. 

Edited April 9, 2014 by tirod

[heartless]

standard NGK plugs as already noted

 

Subaru or NGK wires

NGK part # FX41/stock # 8004 - do a little shopping around, they can be had for between 35 & 45 bucks.

[Fairtax4me]

V-powers are copper and work fine.

 

The old 2.2s are not dreadfully dependent on having the highest quality plug wires like some of the newer engines. I've found these work just fine with a decent quality set of wires from the parts store. I've run Carquest house brand, Napa house brand, and Federated Auto house brand plug wires on these (the cheapest from each of those, all around $35-40 at the time) and had no problems with misfires. Granted, the cheapest wires from those stores are, in general, still better quality than the mid-range wires at the big chain stores like advance or autozone.

[cascadeclimbn]

So I will definitely get new wires and hopefully take care of the misfire for good but I'm still concerned that wires aren't my real problem. The issue I've been having is the car losing acceleration. When I first changed the wires it still did this even though I hadn't had a misfire code pull back up. Then I changed the knock sensor and the issue went away for 300 miles till it happened again the other day and the CEL came on for a misfire after the loss of acceleration happened last time. So my question is can a misfire cause the loss of acceleration I've been describing? And also, why would it seem to have resolved itself after replacing the knock sensor but then happen again and the CEL be for a misfire? Hope I'm stating all this coherent enough

[lmdew]

Did you get the knock sensor properly positioned?  It has to be clocked so the plug end of the sensor is in the cutout and it sits flat on the engine.  If you didn't do that you may have damaged the new knock sensor.

 

Yes, mis fires can cause loss of power, as will the retarded timing due to sensed knocks.

[tirod]

Because of the dynamics of the combustion process, when the knock sensor "hears" the threshold vibration of knock, it signals the computer. The computer has to dial back the advance a whole lot more than we think - not two degrees, more like 15-18. That is the loss of power you sense. The computer goes back to it's routine of advancing it a few at a time until it finds a happy medium. It checks every now or then and if it bumps knock again it dials back, then starts advancing again. 

 

Knock goes to octane rating. The less octane, the less timing, the less power. I refuse to buy 87 octane fuel for my Subaru, and I can't afford 50c a gallon more, but I do select a slightly better grade. I'm getting 24mpg, which is so much better than the previous car's 16 that a dollar or two more a fillup isn't a major economic hardship. I'm already saving a tankful a month. 

 

When I do see it not running as well as I like, I bump up another grade, which usually restores the power and adds more - because higher octane allows more timing. More timing, more power. More power, less throttle if you drive the same. Funny, but a better grade of gas can actually give more mileage and helps pay its way. My experience with my Subaru is that it's more sensitive to gas octane than my previous car, which didn't seem to care whatsoever. Again, 16mpg. It was a hog. 

 

If you can find it, fill up with no alcohol premium and see if the problems disappear for a tank. Likely most of it will, and what is left will be engine related issues. 

[RallyKeith]

I know everyone here is going to say I'm totally off, and for your sake I hope I am, but it's possibly an issue with dropped exhaust valve guides. Somewhat common problem on the 96-9 single port heads. We had it happen on my father's 96. Shows up as a misfire code and it will come and go.

 

http://allwheeldriveauto.com/subaru-service-seattle-dropped-exhaust-valve-guides-explained/

[Fairtax4me]

I have never heard of that on the phase 1 heads. Regardless, a dropped exhaust valve guide will not produce a moving misfire. It will only affect the cylinder with the dropped guide.

 

Misfires produce power loss because the cylinder that misfires is not producing power when the misfire happens.

 

It takes several successive misfires in order to set a code, so the engine can misfire for some time before the light comes on if the misfires are spaced out enough. To illuminate the CEL there must be approximately 10 misfires within 1,000 revolutions of the engine.

10 misfires within 200 revolutions of the engine will cause the CEL to blink rapidly, indicating the misfire has the potential to damage the catalytic converters by pumping unburned fuel through them in the exhaust.

[cascadeclimbn]

I have never heard of that on the phase 1 heads. Regardless, a dropped exhaust valve guide will not produce a moving misfire. It will only affect the cylinder with the dropped guide.

 

Misfires produce power loss because the cylinder that misfires is not producing power when the misfire happens.

 

It takes several successive misfires in order to set a code, so the engine can misfire for some time before the light comes on if the misfires are spaced out enough. To illuminate the CEL there must be approximately 10 misfires within 1,000 revolutions of the engine.

10 misfires within 200 revolutions of the engine will cause the CEL to blink rapidly, indicating the misfire has the potential to damage the catalytic converters by pumping unburned fuel through them in the exhaust.

Thanks for explaining to me exactly what happens during a misfire. In response to another post about situating the knock sensor...yes I did place the sensor at 45 degree angle between the "notch". It's not the knock sensor that the car seems to think is causing the problem, the CEL is for a misfire.

[Rooster2]

Thanks for explaining to me exactly what happens during a misfire. In response to another post about situating the knock sensor...yes I did place the sensor at 45 degree angle between the "notch". It's not the knock sensor that the car seems to think is causing the problem, the CEL is for a misfire.

Yes, you are correct, the CEL code for knock sensor would be different from misfire code.

[RallyKeith]

I have never heard of that on the phase 1 heads. Regardless, a dropped exhaust valve guide will not produce a moving misfire. It will only affect the cylinder with the dropped guide.

 

 

I wouldn't call those phase 1 heads. You might be able to persuade me to call them phase 1.5, but they went from dual port exhaust to single port and the loss of flow created some problems. It's not super common but it is something to keep in mind. It is very hard to diagnose which is why I brought it up. We spent months chasing a misfire that in the end was a dropped valve guide. It was only after talking to a friend who is a Subaru Master Mechanic that he brought this up because it is something they see. I agree that since it moved from one cylinder to another it's not highly probable, but it is as I said a possibility on that engine.