Trent Bates Posted April 7, 2012 Share Posted April 7, 2012 Grrrr! Hi All, I've been fighting a very noticeable light-throttle retarded-timing issue the last few weeks. No CEL's are set. For the last year, I have noticed the same light-throttle speed variations that many others have mentioned. They were minor and easily ignored. Now the problem is considerably worse! I can step on the gas lightly and the car will chug along for a 10-15 seconds without gaining speed. Once it finally gets everything "right", it takes off like the car it used to be without changing throttle position! Also, you can sometimes override this chugging behavior by stepping on the throttle all the way. It takes a few seconds, then it starts to accelerate roughly but not nearly as well as it should. I've been through many ideas and diagnosed many items thinking it might be spark plugs, fuel injectors, bad gas, sticking/failing turbo, various valves and sensors, etc. I have finally narrowed down everything to the knock sensor. That is, If I unhook it, I get reduced performance from the computer going into "safer mode" and a CEL, but the light-throttle-acceleration is almost normal. If I move the knock sensor to a different location, I get almost normal performance back and no CEL. I have replaced the knock sensor and with the new sensor in the factory location the performance is almost as bad as it was with the original sensor. The sensors look identical. :icon_exclThis leads me to think that there really might be some noise issues going on. I can't hear any pinging with or without the knock sensor hooked up. I am using 93 octane fuel. I keep the car well maintained. For what it's worth: I have 64,000 miles on the car. I'm the original owner. I've always used Mobil 1 and changed it every 5,000 miles. Fluid levels are all where they should be. The spark plugs look pretty good. No oil, not too lean, etc. The turbo spins freely and there's no excessive play in either direction. The one fuel injector I looked at was clean and opened nicely with 12V. The oil valves near the injectors both look good and move freely with 12V. I have hunted for vacuum leaks. I've zip-tied the blue T-connector. Both oxygen sensors seem to react quickly and be within range. I don't yet have any good data logging apparatus. I bought Autotap Express DIY because I used a much more complete Autotap back in 1998-1999 but I'm not at all happy with Express DIY. It doesn't even show knock events and seems very limited all around. Anyone with thoughts? I'm willing to move the new knock sensor to a less troublesome location, but if there really is something going on, I'd rather fix it and not ignore it! I'm about to run Seafoam through the intake but haven't done it yet. Link to comment Share on other sites More sharing options...
Max Capacity Posted April 7, 2012 Share Posted April 7, 2012 Have you cleaned the MAF sensor ? 305,600miles 5/2012 ej257 short block, 8/2011 installed VF52 turbo, @20.8psi, 280whp, 300ftlbs. (SOLD). CHECK your oil, these cars use it. Engine Build - Click Here Link to comment Share on other sites More sharing options...
Spun Muffler Bearings Posted April 7, 2012 Share Posted April 7, 2012 page 3 of this pdf has some useful knock sensor test procedures. Not everyone has access to an o-scope, but hopefully this will help a bit. Link to comment Share on other sites More sharing options...
Trent Bates Posted April 7, 2012 Author Share Posted April 7, 2012 Have you cleaned the MAF sensor ? Yes, I did do that as well. Sorry to have left that out. I have cleaned it twice. I think, but now can't remember that I watched some numbers on the MAF through Autotap. They seemed reasonable. I guess it could be running rich but the plugs don't seem to show that. Now I wonder... Link to comment Share on other sites More sharing options...
Trent Bates Posted April 7, 2012 Author Share Posted April 7, 2012 page 3 of this pdf has some useful knock sensor test procedures. Not everyone has access to an o-scope, but hopefully this will help a bit. Okay, I don't have a scope, but I have hooked up the knock sensor to a Digital Multi-Meter and was seeing .02mV to about .70mV from idle to WOT. (carefully applied) The increase seemed linear. This was with the sensor completely isolated from the PCM so the reference voltage wasn't present. I didn't see any big jumps in voltage, but it's possible that the sampling rate is too low with my inexpensive DMM. Also, I can generate a signal by tapping on the knock sensor. Both the old and new sensors seem to be operating within the same ranges. It's too bad the sensor is buried so deeply under the intake. I'm getting tired of taking the intercooler off! Link to comment Share on other sites More sharing options...
Max Capacity Posted April 7, 2012 Share Posted April 7, 2012 This stuff is way beyond me, have you looked in the service manual for a trouble shooting flow chart ? Do you need a copy of the manual ? I can help there. in fact I'll send you one just in case. 305,600miles 5/2012 ej257 short block, 8/2011 installed VF52 turbo, @20.8psi, 280whp, 300ftlbs. (SOLD). CHECK your oil, these cars use it. Engine Build - Click Here Link to comment Share on other sites More sharing options...
Trent Bates Posted April 8, 2012 Author Share Posted April 8, 2012 This stuff is way beyond me, have you looked in the service manual for a trouble shooting flow chart ? Do you need a copy of the manual ? I can help there. in fact I'll send you one just in case. I do have possibly the original version of "vacation pics" from 2005. Thanks! I need to see if there is any good troubleshooting chart in there. Link to comment Share on other sites More sharing options...
Spun Muffler Bearings Posted April 8, 2012 Share Posted April 8, 2012 Okay, I don't have a scope, but I have hooked up the knock sensor to a Digital Multi-Meter and was seeing .02mV to about .70mV from idle to WOT. (carefully applied) The increase seemed linear. This was with the sensor completely isolated from the PCM so the reference voltage wasn't present. I didn't see any big jumps in voltage, but it's possible that the sampling rate is too low with my inexpensive DMM. Also, I can generate a signal by tapping on the knock sensor. Both the old and new sensors seem to be operating within the same ranges. It's too bad the sensor is buried so deeply under the intake. I'm getting tired of taking the intercooler off! Sounds like about the most useful test you can do without a fast meter or scope... Given the relatively weak signal strength, have you investigated noisy grounds? possible the ECU is losing it to other noise, or getting false positive indications? Any chance something's hitting it accidentally? worn motor mount or a hose not routed normally allowing something to vibrate a lil harder and get an occasional hit near it? Knock sensors technically aren't very different from some phonograph cartridges, and the signal is in roughly the same range. I have wondered about using a portable pre-amp to bring it up to line level and present it to the car's own stereo (or a cheap digital audio recorder) as a backyard diagnostic... but these days it's probably easier to score a cheap scope on ebay. Link to comment Share on other sites More sharing options...
DariusD Posted August 27, 2014 Share Posted August 27, 2014 Any updates on this issue? Link to comment Share on other sites More sharing options...
Trent Bates Posted August 27, 2014 Author Share Posted August 27, 2014 Hi All, Yes. IN MY CASE (Don't use my experience to diagnose your own symptoms ) this turned out to be a bad head gasket that let exhaust bubbles through the water jacket under the knock sensor. That's apparently noisy. I was able to stay on top of the coolant loss for nearly 2 years (Had too many other things going on to pull the motor and fix) but eventually it bit me in the butt and now I have to replace my turbo AND pull the motor. I would suggest checking for a bad head gasket if the symptoms I had described above occur. But I don't want to say that a head gasket is definitely the problem. I hope that helps! Link to comment Share on other sites More sharing options...
xt2005bonbon Posted August 27, 2014 Share Posted August 27, 2014 very interesting. First time I hear that bad HG can lead to knock events. wow. Link to comment Share on other sites More sharing options...
MrTris Posted August 27, 2014 Share Posted August 27, 2014 Now that you're starting over, you can use oil that won't kill your turbo. Leave that Mobil1 junk on the shelf at Costco and, instead, pickup some Rotella T6 (very popular here) or some other synthetic that contains the proper amounts of zinc for your turbo. If you need your turbo rebuilt, send a PM to @JmP6889928, he does very nice rebuilds. (Ask me how I know!) Link to comment Share on other sites More sharing options...
Trent Bates Posted August 27, 2014 Author Share Posted August 27, 2014 very interesting. First time I hear that bad HG can lead to knock events. wow. You may understand this already but I'm not sure and I don't want to mislead anyone. The head gasket leak didn't lead to actual knock that I know of. Just lots of noise that the knock sensor picked up. Moving the sensor to a different place on the engine isolated it from that bubbling noise and it worked fine after that. I will be moving the knock sensor back to it's factory position with the rebuild. Link to comment Share on other sites More sharing options...
xt2005bonbon Posted August 27, 2014 Share Posted August 27, 2014 Did you ever log the car? like logging misfire per cylinder and stuff like that? Link to comment Share on other sites More sharing options...
MrTris Posted August 27, 2014 Share Posted August 27, 2014 Subaru's Knock Detection system isn't perfect, but it's certainly the most sensitive. It'll pickup ANYTHING as knock. Link to comment Share on other sites More sharing options...
DariusD Posted August 28, 2014 Share Posted August 28, 2014 Hi All, Yes. IN MY CASE (Don't use my experience to diagnose your own symptoms ) this turned out to be a bad head gasket that let exhaust bubbles through the water jacket under the knock sensor. That's apparently noisy. I was able to stay on top of the coolant loss for nearly 2 years (Had too many other things going on to pull the motor and fix) but eventually it bit me in the butt and now I have to replace my turbo AND pull the motor. I would suggest checking for a bad head gasket if the symptoms I had described above occur. But I don't want to say that a head gasket is definitely the problem. I hope that helps! Thanks for the update. So in your gasket leak case, was the leak itself VISUALLY noticeable, around that knock sensor? Link to comment Share on other sites More sharing options...
Trent Bates Posted August 28, 2014 Author Share Posted August 28, 2014 Did you ever log the car? like logging misfire per cylinder and stuff like that? It's been a while since I first encountered this issue but I believe that I did log for misfires, etc. I didn't see anything concerning or I would have addressed that issue back then. The only item I tried replacing was the knock sensor. Thanks for the update. So in your gasket leak case, was the leak itself VISUALLY noticeable, around that knock sensor? No, the spot that the knock sensor bolts to doesn't go all the way through to the water jacket. It's just a threaded hole long enough for the bolt. The only signs of a gasket leak was the bubbles in the coolant and the dark murky color that the exhaust fumes turned it to. Also, it became oilier. Link to comment Share on other sites More sharing options...
iNfEk Posted May 2, 2016 Share Posted May 2, 2016 bump from the dead! LOL! i may have a similar issue happening with my LGT. it's seeing knock at low RPM and just before spool up. here's the learning view dump. Anyway, where is the knock sensor on the EJ255 engine? Link to comment Share on other sites More sharing options...
xt2005bonbon Posted May 2, 2016 Share Posted May 2, 2016 oooh. You definitely have a vacuum leak somewhere. Look at your fuel trims. They are all peggged at +15%. So you are running lean. Be careful. Link to comment Share on other sites More sharing options...
iNfEk Posted May 2, 2016 Share Posted May 2, 2016 disregard. i found it! Link to comment Share on other sites More sharing options...
xt2005bonbon Posted May 2, 2016 Share Posted May 2, 2016 I bet your knock sensor is fine. Search for a vacuum leak first, fix it if found. don't reset the ecu and see if your fuel trims go back to +/-5%. Also your IAM has dropped to 0.875. So hopefully it'll go back up to 1.00. Link to comment Share on other sites More sharing options...
iNfEk Posted May 2, 2016 Share Posted May 2, 2016 oooh. You definitely have a vacuum leak somewhere. Look at your fuel trims. They are all peggged at +15%. So you are running lean. Be careful. hmm. AFR is normal though. the injectors are stock ones and the car was tuned via access port on a dyno before i got to it. my next step is to unmarry the AP and run a stock map to see what's going on. vacuum leak? hmm i've replaced a few hoses including the turbo inlet due to there being a hole in it before the turbo itself. any chance that the stock injectors are max and that's causing the lean condition since it needs more fuel? Link to comment Share on other sites More sharing options...
iNfEk Posted May 2, 2016 Share Posted May 2, 2016 I bet your knock sensor is fine. Search for a vacuum leak first, fix it if found. don't reset the ecu and see if your fuel trims go back to +/-5%. Also your IAM has dropped to 0.875. So hopefully it'll go back up to 1.00. yeah, been looking at the IAM too! it was at 1 before this correction the ECU made which caused it to go tot 0.875. i pull the LV after every trip. seems a bit extreme but we have different elevations and humidity in different areas of the island. thanks Link to comment Share on other sites More sharing options...
xt2005bonbon Posted May 2, 2016 Share Posted May 2, 2016 maybe your AFR looks OK but I bet if you view the A/F correct parameter on your AP, you'll see like lots of live correction, meaning the ECU is trying to add lots of fuel to achieve good AFR. Link to comment Share on other sites More sharing options...
xt2005bonbon Posted May 2, 2016 Share Posted May 2, 2016 yeah, been looking at the IAM too! it was at 1 before this correction the ECU made which caused it to go tot 0.875. i pull the LV after every trip. seems a bit extreme but we have different elevations and humidity in different areas of the island. thanks If the IAM dropped to 0.875, it essentially means that it saw too many knock events on the map and decided to pull timing across the whole map. And as you can see, it still pulled some more timing even with a 0.875 IAM. But yeah, check for vacuum leak first. Do you have an aftermarket air filter? Link to comment Share on other sites More sharing options...
Recommended Posts
Archived
This topic is now archived and is closed to further replies.