Buckethead

Well, CRAP. Car crapped out finally, a few days ago, while wife was driving it. Would start (I confirmed) but instantly lose idle and die. Had to have it towed to shop -- yay for roadside assistance on my insurance policy. After much diagnosis they replaced the MAF sensor (which I replaced last fall, BTW). They charged me $345 to diagnose and replace. AAAAAHHHH. They said they had a devil of a time finding anything wrong (which I totally believe, after what I've been through so far) and finally narrowed it down to the MAF. And after all that? Car still idles pretty much like crap. It doesn't stall, and doesn't throw a code, but that's the state I've been in for months now. Yeah maybe the MAF kicked the bucket, but I still think something else is going on and that still is not the root cause. Oh, and yes, it was a Bosch 02 sensor that I installed a few weeks ago. I bought it online though, and not from a dealership. I have since done some reading that Bosch somehow sells the same part, with the same part number, but it is somehow "different" from the OEM version? I don't know what to believe now. Ranting, ranting. I am about to give up and take it to a Subaru specialist in another town, to have them take a look.

UPDATE Last weekend I replaced the upstream 02 sensor only. It looked worse for the wear, very charred (black) at the base and "white" at the tip. A far cry from the appearance of the new one (shiny metal). The car ran better, and I took it on a short roadtrip 45 minutes away. It ran fine, hardly any sign of idle roughness. Yesterday it threw a P0170 Fuel Trim (Bank 1) code. Is this code generated by the O2 sensor? Did I screw up by putting in a sensor from the auto parts store instead of the dealer?

Thanks for the reply. All the plugs looked pretty similar, but plug #1 might have been a little worse than the others. It had the worst oil leak around the valve cover, so I guess that doesn't surprise me. (ha, I said but plug) . . . I might replace the ground wires completely. Although the issue is intermittent, it is "regular enough" I guess, that it doesn't strike me as a loose connection or anything. It's almost as if it just loses its marbles periodically while idling . . . I'm hoping the O2 sensor helps.

OK, the upstream O2 Sensor is ordered, along with the ECTS. I'll swap those two out -- probably have to wait til the weekend -- and report back. Thanks for the suggestions!

Well, here's what I found over the weekend. 1) Replaced valve cover gasket with a Felpro kit from Amazon. The #1 and #4 spark plug gaskets were the problem there, they were stiff and almost "D"-shaped instead of round. So yeah, they needed it. 2) Replaced spark plug wires. 3) Replaced coil pack - it was cheap, and I figured what the hell. 4) I looked all over for vacuum leaks, I went over 100% of the engine bay with my handy propane torch. Absolute zilch. Result? Almost no difference. Runs great down the highway and at any throttle level. Idles fine 90% of the time, then drops down into the gutter and shakes. AAAAH. This is so infuriating to me now, even though the car is running better overall than it has for some time. I guess the upstream O2 sensor is my next stop? Sigh. Oh, I forgot to mention that I found this list on the web about rough idling: 2. Rough idling 1) Manifold absolute pressure sensor 2) Mass air flow and intake air temperature sensor 3) Engine coolant temperature sensor 4) Ignition parts 5) Air intake system 6) Fuel injection parts 7) Electronic throttle control 8) Crankshaft position sensor 9) Camshaft position sensor 10) Oxygen sensor 11) Fuel pump and fuel pump relay This ranks O2 sensor as 10th on the list . . . I've already changed the MAF and done "ignition parts" . . . nothing to note from air intake, except cleaning the intake and IAC valve... Not sure the temperature sensors would matter, it doesn't seem to change behavior much when cold or hot...?

OK, I did a couple of things today . . . 1) Replaced spark plugs. The old ones were definitely worn, but relatively clean -- except for one plug, which had oil in the spark plug well. Looks like I get to do a valve cover gasket, yaaay. 2) I decided to check the ground wire connections. The battery-to-chassis connection was sort of crusty looking. I cleaned it up with sandpaper and a wire brush, and re-attached with some dielectric grease. The battery-to-transmission ground seemed OK, but I gave it the same clean and grease treatment. The result . . . ANOTHER slight improvement. Car idles fine in the driveway, and in the neighborhood. I went up the highway a bit to get the engine hot, and cruised back into town. On the way back at a stoplight I had ONE repeat of the old "drop to 400 RPMs and near-stall" . . . but it never actually stalled. This RPM-drop now seems to happen only if the car idles at a stoplight for more than 30 seconds. Eventually the engine will drop in RPM to about 350-400RPM, then it will either stall, or correct itself. If the light is short, however, the idle routine never reaches that "drop-down" point. So in a nutshell, I couldn't make it die. Another slight improvement is that rapidly pumping the brake doesn't seem to have as big of an effect on reducing RPMs as it did prior to today. Next steps: (ideas?) I plan to just replace the ground wires with new ones. I'll use some higher-gauge wires, and add a couple more grounding points. I definitely saw green corrosion on the ends of one of the cables, in the braided wire. I've heard the grounding is only just "adequate" in these cars, and somewhat prone to corrosion over the years. (true? I dunno, but it's a cheap fix) I need to do the valve-cover-gasket replacement, while I'm at all of this. I will also investigate the front 02 sensor, thanks for the suggestion mikaleda. Any other thoughts or suggestions welcome, thanks Sobaroo for pointing me toward the plugs, don't know why I didn't think of that!

Thank you for the comments, I'll do another inspection of the intake and look for vacuum leaks. The plugs should only be a few years old, but I'll check the records and maybe just swap them anyway just for grins. Maybe my memory is bad and they're older than I think. I hadn't really thought about the plugs because at higher RPMs the car runs pretty well. I appreciate the interest and the feedback! I will report back on what I find.

For starters, thanks for this great site and forum. I've read many threads here over the years and finally wanted to ask a question, so I stopped being a lurker. I have read all the "idling problem" threads and did some troubleshooting myself (see below) but now I don't know which direction to go... My '99 Forester S has 230K miles on it -- It's engine was blown and rebuilt back at around 80K miles (and 8 years ago!) . . . and it has been a great and reliable engine since then. I've had a rough idle and intermittent stalling issue for about a year now. Since I buckled down to try to fix it once and for all, the issue has been driving me batty. Each fix I do seems to help "a little" but the problem doesn't go away. The behaviors are familiar suspects, as follows: Idle seeks around between 700-1200 RPMs, sometimes drops to 500 RPM when at a stoplight. Pumping the brake rapidly can cause RPMs to drop further *sometimes*. Sometimes the rapid-fire braking has no effect. Hesitates at acceleration, again, only sometimes, and only at lower RPMs. No CEL or codes (yes the bulbs work and everything illuminates correctly at startup) Timeline starting around last fall: -- FALL 2015 -- 1) Got a P0104 code, so I replaced the MAF sensor. Super easy fix, right? CEL went away, and it made *a small improvement* to how the car ran. Meaning, a little rough at idle, but otherwise OK. 2) Cleaned the air filter while I was at it. It was dirty. Cleaned throttle body (didn't remove it from the car). It was also quite dirty. This seemed to help *a little*. Still noticed low-idle from time to time, but no biggie. 3) Had an oil-change and winter service checkup at my local garage. I told them about the rough idle I was seeing. They couldn't reproduce it and said the car was pretty tip-top, so I was satisfied. -- Winter 2015 -- Wife drove the car a couple times a week, she sometimes complained of stalling. It never stalled for me when I drove it... go figure. Was cold weather helping or hurting? No CEL. -- SPRING 2016 -- In April the car started stalling more often and running rougher, idle speed dropping to 500 RPM at stoplights quite often. More noticeable stalling. Warmer weather? Something getting worse? I decided to look closer. 4) I noticed some oil drips and noticed some peculiar oil consumption. The oil pressure sensor had given out and was oozing oil. Replaced oil pressure sensor. Replaced PCV valve. While doing these things, my engine finally threw a BRAKE/BATTERY warning light. I thought, Aha! Finally! . . . but, it was just a blown alternator fuse. I put in a new fuse and that did the trick. 5) After blowing the fuse the car ran SUPER CRAPPY. More research and many tries at the ECU reset procedure. I did this ECU reset several times trying to get it right, it would definitely affect the idle speed whenever I did it. I finally got it right, and returned the idle speed to back to where it was before. 6) I cleaned the throttle body AGAIN (again, still on the car). It didn't seem that dirty, but I had just cleaned it a few months prior . . . 7) I pulled the IAC Valve. It was VERY dirty. I reassembled, and the car ran SUPER CRAPPY yet again, but another ECU reset procedure got it back to running "like normal" again. After all this the car runs *a little better* but the behavior is not gone. I drove it a lot yesterday and could never get the car to stall -- I couldn't even *force* a stall by mashing the brakes, but the symptoms are still there: Idle sometimes drops to 500 RPMs and recovers, but takes several seconds. Braking sometimes makes idle drop *a little*. Hesitation sometimes on acceleration from low RPM. So, what to do next? I haven't investigated the brake system yet. I'm not convinced a vacuum issue is the real smoking gun though, but maybe that's where I'm going wrong? Should I do more cleaning? Should I just replace the IAC valve? Could other sensors be failing or dirty, without throwing a code? I just ran across another thread somewhere that suggested it could be a failing MAP Sensor; someone suggested these can fail without throwing codes. I am trying not to just throw parts at this issue and want to pinpoint where its coming from, I don't think I've found the smoking gun. Thanks in advance, and sorry if you all are getting sick of people and their rough idle topics.