I had a bad EGR valve in my '98 that gave me a persistent P0400, but I never noticed a driveability symptom.

 

*) Can EGR go bad? -- YES
*) Can EGR fail without throwing code? -- YES
*) Does EGR fail on Subaru? -- YES (I am aware of several late 90s outbacks failing)

From reading your post, you have not actually PROVEN that the EGR is the problem. You have made a reasonable assumption... but nothing more than that.

Can you try PLUGGING the EGR pipe so the exhaust gas cannot get into the intake manifold? If this makes the stuttering problem go away, then you would have a better idea that you are on the right track.

 

I re-installed the valve body, but left the solenoid off, but plugged in. Started the engine, no codes. I raised the throttle while depressing the plunger, and got the same symptom. Is the valve OK, and solenoid bad? Or is the valve and solenoid OK, and the computer telling it to open, possibly too wide, at inappropriate times? The struggle is real....

 

Is the valve OK, and solenoid bad?

If you can 'manually' open/close the valve and detect a difference the way the engine runs.... I would assume the valve is working.

Or is the valve and solenoid OK, and the computer telling it to open, possibly too wide, at inappropriate times?

To answer this, you may need an OBDII reader. Using such a device, you can monitor the signal that the COMPUTER is sending to the EGR valve.

Another way I might look at this would be to put a voltmeter across the EGR solenoid terminals and monitor the voltage being sent to it as the throttle is opened.

It is also possible that the INSIDE of your intake-plenum has a heavy layer of carbon where the EGR feeds into it. I know that on some VWs, the intake-plenum gets so carbonized that it has to be removed and a torch used to burn-off the layer of carbon.


Another idea: What if your exhaust is somehow partially plugged and when the EGR opens, too much pressure is fed into the intake-plenum?

 

Is this a 4 or 6?

When the EGR operates, the ECM opens it at cruising speeds, it allows exhaust back in to the intake stream to cool the combustion. At idle and deceleration, the EGR closes. If the EGR is faulty by sticking open, the car will idle poor. If the EGR doesn't open at all, you wouldn't notice that its not operating but the computer will because it judges EGR operation on pressure feedback.

Unless the valve is partially open at idle, your problem lies elsewhere.

Could be air metering from the MAF, dirty air filter, fuel filter, TPS adjustment, IAC valve sticking on the low side(does it idle low?), AF sensor(s), battery, grounding, or PCV valve. These are just a few things. I doubt its the EGR.

 

Sorry for the omission, it's a 2.5. I have the EGR valve installed, with the solenoid plugged in, but off the valve. So, essentially, there's no EGR. I drove it briefly this morning, and right about where and when the problem would start, it behaved beautifully. So, I think I've narrowed it down to a faulty solenoid (doubtful), or the computer operating it erroneously. Cardoc, I will pursue the ideas you've suggested, maybe some sensor offering an erroneous input, causing the computer to send an ill-timed command. Stay tuned, and thanks....

 

My '05 forester has had to have the electric EGR valve freed up a few times, symptom was bad idle, ok elsewhere.

Freed right up after removing and disassembling.

 

Did it randomly idle bad or all the time ?

 

OK, some research. Drove about twenty miles. Felt great, as I expected, air/fuel charge not diluted with exhaust gas. Not tapping/pinging like I thought it would, given that the intake charge was hotter, not being given the benefit of lower combustion temps due to the diluted intake charge. BUT, it threw the P0400 code, which I was expecting, and it immediately cleared when I used the Ultragauge to reset it. Got home, pulled into the garage, popped the hood, figuring I'd remount the solenoid. It was screaming hot, to the point of not being able to handle it. It was hanging on its cable, resting on the vacuum line to the brake booster, away from the engine heat. I would expect an energized solenoid to be warm, but this hot? Anyway, I'm going to remount it, to prevent any damage, and continue to dig, using suggestions from the contributors to this thread. I'm thinking, maybe grab a couple at the junkyard, do some swapping, possibly isolate the problem. Thanks guys.

 

Time for an update. Haven't been driving the car much, but got back to it yesterday. Out of desperation, looked on Youtube for these symptoms, and all the results pointed to the MAF sensor (hats off to cardoc!). Got a can of cleaner, gave the MAF the recommended shots of fluid, and magically, it ran great! I put the EGR back together, and it seems to be behaving! Looking at the prices for a new sensor, the nine bucks for a can of cleaner is a bargain. Thank you to those who've contributed to this thread.

 

Update my update. Apparently, my '04 OBW 2.5 has no MAF sensor. It relies in a MAP sensor for operation. Can this be cleaned, or just replaced?

 

My '84 Corvette uses a MAP sensor instead of a MAF sensor. Not much to clean out. Personally if acting up I would just bite the bullet and replace it.

 

The MAP can be cleaned but you have to be careful doing it. Carbon will get inside the orifice and create a blockage that inhibits the baffle movement and the MAP sends erratic voltage back to the ECM. Subaru only MAP if you decide to replace it as the aftermarket MAPs do not scale at the proper voltage range.

 

A bit of an update. I got a MAP sensor from the junkyard, cleaned it with the MAF Sensor cleaner I erroneously bought, and it seems to be behaving. Not perfect, but so much better.
Thanks to all for your responses.

 

heyhar:
how did you test your EGR valve?
I have an 08 2.5L NA; and I'm not sure if it's a stepper motor or solenoid version.

 

As far as testing it, I tried leaving it installed and unplugging it, sets a code.
Take the screws out, to separate the 'motor' from the valve body. Takes a longer while, to set a code.
Unplug the one on the engine, plug in another. Computer figures it out, sets a code.

I was convinced that my problem was the fault of the EGR valve, but through all the combinations of testing/isolating the EGR 'solenoid' and its accompanying valve, the one thing that remained constant was the skipping/stuttering/shaking of the engine. I sort-of figured that the EGR was in-fact operating normally, as without it operating, the engine would ping/valve-tap, because the intake charge is calibrated for a little 'spent' EGR exhaust gas, and when it was missing the cooling effect of the spent gasses, the combustion temp tends to rise, causing a pinging episode.

So, I set my sights on the MAP sensor, and cleaning and swapping among the few I've gotten from the junkyard has, at least to my mind, convinced me that this little critter is what's at fault. Warmer weather and less overtime may make it easier to prove my hunch.

 

It was stuck hard so bad all the time.
Something like that could be sticky sometimes only.

 

Well, a bit of an update. The fluttering, hesitating, general undrivability got a lot worse. I'd dropped the exhaust after the cats, seemed to run better, I wad due for a new exhaust anyway, so it got replaced. It ran until it didn't. Took it to a mechanic to have it scanned, and the ECU wouldn't display some functions. He did find a wonky injector, so they were replaced. New crank sensor, plugs, wires, too. Still crappy. I took it to another mechanic, he plugs in his Snap On diagnostic tool, immediately finds pressure, not vacuum, in the intake. I guess my new MAP sensor is working!. I came home, removed the EGR, and it was jammed open! Made a temporary block-off plate, the engine behaved, but with a slight misfire. I went back, and he said to try Techron, the black bottle. I followed the directions, and within a tankful, the misfire was gone. So, now I'm back to a running car, with the same annoying stutter.

Long story short, I looked on YouTube for videos on bad knock sensor. I unplugged it, and the problem was gone!
Seems as though it was the problem all along. So, I ordered a new one, should be here this week.

It never threw a code, other than when I unplugged it. I'd like to know how they operate. Disconnected, no jerking/stuttering, but it did feel a little down on power. Does the sensor allow the timing to advance until it detects whatever it's listening for, and then retard the timing until the noise is past? And without it connected, does the timing default down to a safe level? I'd really like to understand this process a little better.

Thanks to all for your responses, and putting up with my graphic novel!

 

Time to update the update. Yes, unplugging the knock sensor seemed to mitigate the problem, as the timing and power were reduced, lessening the feeling of the stutter. I finally hooked up with a hard to find fellow, who'd been recommended to me. He's a Subaru dealership mechanic by day, and a performance builder by night. I described the problem, and what I'd done, and that I've failed. He insisted that it was a driveline vibe, and instructed me on how to track it down. Separate the rear drive system from the front, see what's still happening. I installed a fuse in the FWD slot in the fuse block, effectively removing the rear end from the equation. Still doing it. How old are the front axles? They've both been replaced. OK, start shaking stuff. I get the car up on a lift, and, damn if the right front axle wasn't really loose! Better rubber technology means that the boots outlive the bits inside, so it looked just fine. I replaced the axle, and the problem is solved! Two years chasing a problem that's gotten worse over time, but my focus was drawn by other problems that are typical of a fifteen year-old car.
It's almost never the first (or second, or third) thing you look at!
Thanks to all that have followed this tortured thread, and chimed in to offer ideas....

 

. . . if the right front axle wasn't really loose! Better rubber technology means that the boots outlive the bits inside, so it looked just fine.

So, a badly worn, or broken, CV joint. That's a really different outcome than what was expected. (Was the joint disassembled to see where/how it failed?)

Side benefit, perhaps, but "can it go bad . . . " had me wondering how the EGR valve works, and led to https://www.subaruoutback.org/threa...eads/subaru-electronic-egr-valve-how-its-made-works-and-might-be-tested.505085/.