Trouble Code P0303

It seems like the only times I ever post on here, it's about something I need rather than something I'm contributing. Most of that has been the fact that my Legacy has been in the body shop for ages on end. Hence the purchase that brings me to create this thread.

I recently purchased a 2005 Outback 2.5XT with a manual transmission so that I would have a Subaru daily-driver. It had high miles, but finding one of these in Texas with a manual transmission is nearly impossible. The price was VERY right, too. It's not without its issues, though. When I first got it, it was throwing 5 codes. Three were O2 sensor related (upstream and downstream) and the other two were for the starter input voltage and a neutral safety switch error. The car ran okay under heavy acceleration and even under partial throttle, but when I would decelerate quickly it would die. By this I mean nearly every time I came to a light or a stop sign, dropping the revs by pressing the clutch would kill the car. I'm against revving a car in neutral like an ass hat on principal, so I just tolerate it and restart it as often as I need. What's key in this "pre-repair" state is that the car never had a miss. None was ever felt and no code was ever registered with the scanner in 2 weeks of driving.

I bought new Denso OEM O2 sensors and proceeded to replace them. The install went smoothly and I fired it up. It ran okay in idle, but seemed like it was working a few quirks out every now and then. I forgot that I needed to reset the ECU, so I stopped it after about 2 minutes, disconnected the battery and reconnected it about 45 minutes later. I started it up and it seemed to idle the same again. I let it do that for about 30 minutes before I went to get in the car. As soon as I tapped the gas, I nearly killed it. This is in neutral, by the way. It recovered, but it let me know how taxed it was after that little effort of mine and took some time before settling on a normal rpm to idle at. Once I finally was able to coax it up enough to drive, I took it out around the neighborhood. The familiar check engine light was still blinking at me.

It runs very rough most of the time. Partial throttle acceleration will usually come with a little surprise as it seems to only make about 50-60% of its power and will then surge up out of nowhere. There is a vibration felt in the wheel, and the engine is very finicky about where it thinks a good spot to idle is. If nothing else, it's a bit embarrassing to drive between the shaking and stalling.

My boyfriend and I erased the codes and restarted the car to see if the CEL came back. It did. This time registering as a cylinder #3 misfire. Now let me tell you that I wouldn't be wasting the internet's precious time if I felt that there was some sensible logic behind what might be causing that code. In fact, I still may swap coils between cylinder #3 and #1 to see if the problem follows the coil. However, it never had a miss before I replaced the O2 sensors in 2 weeks of driving, nor was there a code present or stored. Obviously there is a miss present, but I guess what I'm hoping for is for someone to tell me something else that could be causing the coil to miss that could have also been an issue previously. A clogged fuel injector, perhaps, or something less obvious. I'm counting on the fact that this may be a common occurrence, as my search results would seem to indicate.

Thanks for the help! Pictures of the new car will follow when I'm able to get it cleaned! It's not much to look at, though. But it's MUCH faster than my Postal. When it's running well, that is.

You're on the right path. Swap the coils and see if it follows.

Your symptoms definitely point to misfire symptoms.

If the CEL is flashing then it is actively misfiring, if it is solid then it had a previous misfiring event.

Let us know what happens with the coil swap, that should give an idea of what course of action to take next.

Hopefully I'll have some time between class tomorrow when it's warmer out. It will give me a chance to check the condition of the spark plugs, too.

My confusion, though, has to do with the symptoms. It seems too coincidental for a coil pack to have suddenly gone bad once new O2 sensors were installed. But I can't come up with a string of logic that would explain why it would have suddenly gone bad, or why it could have been masked with poor O2s before. It's almost as if the car was run too long with faulty information from bad O2s and learned some habits that it can't quite shake now that it's being delivered accurate information.

Much of my apprehension comes from a previous "coil problem" on my '99 Postal Wagon that I posted about on here that turned out to be something completely unrelated. It turned out there was something that was plugged in under the dash that shouldn't have been, but for the life of me I can't remember what. I tested the factory coil six ways from Sunday to no avail, then ended up buying a new one for no reason at all. Just gun-shy, I guess. Thanks for the reply!

Well, it could be something to do with whether you’re running in closed-loop or open loop as well.

When you had bad O2 sensors you should've been operating in open loop. Fixing the O2 sensors then makes it so the ECU will run in closed loop again. At partial throttle the ECU will run in closed loop. It should kick to open loop when going to wide open throttle or over a certain RPM (3500ish IIRC).

If it doesn't follow the coil pack, then another easy thing to try is to unplug the MAF sensor thus forcing it into open loop ad see how that runs for a test.

Sounds good. And yes, WOT acceleration is substantially better that partial-throttle.

I just ordered a set of iridium plugs to aid in the diagnosis process, too. Someone on the subaruoutback.org forum linked another instance of someone having similar issues and it being caused by gunk build up in the throttle body. I suppose I can order a new gasket and check that, too. If that's dirty, then it's likely the intercooler is, too.

Reporting back in with an update now that I've had a chance to work on it a bit since I last posted. First, if you've never done spark plugs on these things, then get ready. I have tiny hands and they were completely necessary to get the coil packs and spark plugs out, as well as interchanging extensions and sockets in place. Subaru did us no favors by lining those up with the frame rails like that!

I now have 4 new ILFR6B spark plugs installed and we switched coil pack 3 and 1. The latter was fairly unnecessary as we discovered the issue with the coil fairly early on. We noticed a broken zip-tie around it when we got the air box off. It seems that someone who removed it before broke the connector on it, and secured the coil by zip-tying it. It's since broke and has no longer been getting a connection. Once we got new plugs in and the coil packs back on, we threw a new zip-tie on to secure it for the time being. We'll have to find a new one some other time and de-pin this one if it's possible. Not today, though.

My boyfriend helped me by removing the intercooler as subaruoutback.org said that the PCV clogged the intake path. We didn't have new intake manifold gaskets waiting, so I didn't bother pulling it to look at the throttle body as was also suggested.

After everything was plugged back in, we started the car and let it idle. I went to the backyard and played with dogs for a bit and came back to read the codes after erasing the old ones. I have five new codes.

P1518 - IACV
P0852 - Neutral Safety Switch
P1152 - Front O2 Low
P1153 - Front O2 High
P0137 - Rear O2 Low Voltage

Two things of note here. I replaced the O2 sensors with new Denso parts, so I'm doubting that a bad O2 is causing any issues. Also, the P0303 code that I started this thread for is gone! That tells me that the coil pack "repair" was successful. We've still got codes, though.

I did a little bit of searching before coming to ask for help again and found someone with a pretty similar situation. No conclusion here, but they did suggest to check grounds. Does anyone with a diagram know if these issues share a common ground?

http://forums.nasioc.com/forums/show....php?t=2332183