I've been experiencing some consistent feedback knock when I get close to boost under 3k rpm. When boost gets around -2 psi, my car progressively knocks until I let off the throttle and start again. I've seen it knock all the way to -12+ feedback knock. I've noticed this more when I am going up a slight incline where there is more load on the car. This is where I've noticed the more severe knock numbers. **I have never audibly heard any knock from the car - I've only seen this knock on my AP. I usually get "close" to boost (-2 psi) at around 2600-2700 rpm or so, so I'm not in a gear that's too high. We all know it's bad to go WOT under 3k rpm, but my car will knock badly if I just flirt with the idea of maybe getting to even "0" psi under 3k rpm. Above 3k she boosts just fine with no knock. (Well.. she recently will knock around -1 under WOT..). Some irrelevant but maybe relevant facts about this car: -the car is overboosting 2-3 psi (tuner can't tune it out so I just got an EWG setup that will hopefully help with that issue). (I bought it stock and it was overboosting when I got it. boost controller and tune didn't help. wastegate issue?) -the car is leaning out a lot as rpm increases (starts at around 10.8ish AFR but leans out to around 11.4 when approaching redline...) (I am going to clean the injectors and maybe convert to top feed and increase injector cc to try to fix this problem) -my oil gets dark very fast. sign of running rich? i'm confused with this problem - maybe someone may know why this happens. I did a compression test and the compression came back close to perfect. car burns maybe 1/2 quart of oil or less between oil changes. -06 STi - stage 2 - tuned for 20 psi
Generally speaking, I'd only be concerned with FBKC while in boost and with the engine under load. It could be false knock caused by a rattling heat shield, loose intake/intercooler piping, other random stuff loose in the engine bay bouncing around, an incorrectly clocked or incorrectly torqued knock sensor, among other things. Has the car behaved like this since it was tuned or just started progressively getting worse? I'd start by making sure your knock sensor is clocked correctly and is torqued to spec. Have you talked to Graham about it?
I have talked with and sent several logs to Graham about this issue, he doesn't know what it is. it is boost related - if I stay out of boost it doesn't do it, and if I'm in a lower gear it doesn't do it. If I downshift and am above 3k it doesn't do it at all. I tried shaking down my whole bay and I can't find anything that makes any noise.. but I haven't looked at the knock sensor at all yet. the heat shield seems to be in there securely, I did check that. if it were the sensor not properly bolted into place, wouldn't it make lots of noise constantly? and not consistently? I got the tune very shortly after I got my AP and I don't remember if it knocked like this before the tune, but I did have some fbkc before the tune
Stock longblock? Built blocks tend to be "noisier". As for the incorrect knock sensor torque, it may be at a certain load/rpm that replicates the reverberation causing the knock scenario. Hard to say really. More than likely something is wonky with the tune...
As for the leaning out, have you upgraded the fuel pump? 11.4 doesn't seem too terribly lean, but it sounds as though the pump may not be keeping up if it's tapering off at WOT. Any idea what IDC is at under WOT?
the car has a stock block (I wish it was built!). I'll torque down the sensor and see if that helps. I've been wondering about the tune... but I've heard really good things about him and his work. I need a new tune for my ewg so maybe I'll go elsewhere. But I don't know if I'm ready to completely give up on him yet. Any idea why my oil gets dark quickly? I really appreciate all of your input with this
IDC is high (close to or at 100%). it was something crazy like 110% before the fuel pump. that's why I'm wondering about my injectors and if I should either clean the stock ones or just upgrade them to 1000cc for e85
If you go E85 you'll need even more injector. Oil is likely just getting dark quick from heat cycles with the turbo, assuming your oil is pretty clean and your compression is fine, I wouldn't worry about that too much, personally.
I've heard mixed opinions, just like any tuner though. I had a bad experience, but I can't say for certain the tune was entirely at fault. I can say that the amount of timing my car was running, it was no wonder I had issues (lots of timing produces lots of heat, which can cause detonation aka: knock). IMO, if he can't figure it out, I'd go elsewhere. Heard good things about Devin at JM. That's for sure why your afr is leaning out with IDCs that high 1000cc will be fine for stock turbo and e85. But good quality aftermarket side feed injectors are pretty much non-existent. Best bet would be to convert to top feed injectors, but you're looking at more money there too (top feed conversion rails, fuel lines and then injectors). Injector Dynamics seem to be the go-to for quality top feed injectors...they only make top feeds though. Lots of tuners will tell you to stay way from DW injectors (and aftermarket side feed injectors in general)
yeah as I said earlier in my initial post, I would convert to top feed if I upgraded my injectors. side feeds tend to clog, are more prone to leaks, don't have bigger size options, can't stabilize fuel pressure as well, among other things. I looked into all of that as well. I've already looked at all the pricing etc for top feed and it's not a cheap conversion. but it's one I'll probably have to do if I want to meet my power goals. IDs are for sure the injector brand I am going with. I have to look into rails and lines more to know what to go with for them. you have been extremely helpful! I guess there isn't anything really obvious causing this issue, and I will check the sensor and see if that solves everything!
1k injectors and e85 should work fine, as mentioned that's likely the cause of your lean condition. Overboost that can't be tuned out sounds like a mechanical issue somewhere. Possibly wastegate, maybe boost control solenoid? What are you running for that? (may have missed in your post) but may be somewhere in the tune too. As far as other tuning options, I was extremely happy with my tune from Shane at DBPerformance in Rogers. He's safe above all else and didn't seek to add power until he had an air fuel curve he liked, and then pulled 40 more HP out of the car after just an intake. Otherwise JM Auto in Osseo has a dyno as well and Devin tunes there.
my other Subaru was tuned by DB and I was very happy with it! however, I admittedly didn't watch my ap (I had the v2) like I do now so I guess I can't really compare it genuinely. maybe it was knocking etc and I had no idea. I have lots to say about JM that isn't good, so I don't plan on getting a tune from them. with that being said, I hear endlessly about how great they are so I'm sure they are good. I just personally don't want to go there due to past experience. I should probably give them another try! I'm not trying to taint their name by saying that i'm running a grimmspeed ebcs
Harder to tune, not as well flow matched between injectors, calibration datasheet isn't as good, worse atomization/spray pattern etc. They used to be (maybe still are) just modified stock injectors which isn't as good of a spray pattern as good injectors (like ID). I personally will only run ID. I got a set of the new ID1050x (stainless internals for E85 protection) for my CTS-V and a lot of tuners won't even touch any other injectors
I'm so glad I went ID when I did my e85. I can't say how well other brands would be but ID are pretty much spot on with their flow sheet. Shoved the numbers in my map and only needed a little latency. Inj scale was perfect and my trims are only +-3% on pretty much any ethanol content.
From what I've learned on my car, a few oddball things can trigger fbk. I noticed it more on e85 than anything. I noticed when I was too aggressive on my intake cam adv in the cruise range, it would knock once and a while. I found that the sudden change from 10-17 degrees to 35 was about where the load would increase. Most of the time it was when I was going from a decline to incline in 5/6th. Anywho, the ignition timing also played a factor as the cam advance would not always perfectly match up with ignition while driving. If you look at any map, there are some drastic changes in that low/very low load range.. Fbk can definitely happen in decline to incline changes. For the leaning, if IDC is that high at wot I'd suggest starting with seafoam. Easy and cheap test that never seems to hurt in the long run. It's also worth checking the maf and seeing how high it is when IDC is maxing out. If it seems low, it might a fuel delivery issue. Kind of random but I know someone else tuned by Graham and he had the similar issues where knock seemed all over the place. There are many contributors but it might be something consistent with the tuner. Other people have mentioned it but if knock isn't happening under obvious load, with the exception of jabbing the throttle (which I found can cause a knock event for several factors too) it should be fine.. The car knocks all the time but I think subaru only shows fbk if dynamic ignition advance is being applied to the final timing value. That can happen at just the right time during cruise or coming onto boost. Not to throw more out there but there's generally a known problem on cars that uses fpr that references manifold pressure. Basically I found even the stumble delete on the sti doesn't completely remove the weird fuel pressure drop around 0 psi. Lastly.. It might be worth watching knock sum cyl 4 and keep an eye on how it grows. It's a little overwhelming to watch if you aren't familiar with it but it gives you an idea on how knock is seen. Thats my $0.02. Idk how accurate it is but it's what I know.
very helpful information. I really appreciate you taking the time to give me all of these ideas. does the fpr issue apply to my year? or only 08+? I haven't heard of that before. for knock sum - should I watch cyl 4 over others because the knock sensor is on cyl 4? why cyl 4? why watch that if it doesn't affect DAM? sorry if that's a silly question..
There are a lot of theories for the stumble that I've read up on. The 08 was the worst because it had dual fprs that have a reference Vaccuum line between themselves and cyl4. I think 02-07 had it referencing the center of the intake and maybe just one. Either way, I think it's a global problem on turbo cars with different degrees of symptoms. Someone else might know more about that though. Im only familiar with the 08-14 sti. So for knock.... Definitely not a silly question. There's a bit too it but it's all intertwined. The sensor is closest to 4 and most events are assigned to cyl 4 count. I think the ecu counts potential knock based on crank position too, but not entirely sure. The easy answer is that number counts up the fastest/most. I use knock sum 4 to watch how my driving habits affect knock. Basically I can see 20 knock counts in about an hour of driving. Some cars might be 1, some might be 100.. Its more about averaging over time. When it increases, I typically check trims and maf to make sure thinks look consistent. Onto dam... Dam stands for dynamic advance multiplier. Dam is multiplying the dynamic ignition advance and the result is added to timing. If your dam is 1 and your dynamic advance is 5°, your timing will be base timing + (dynamic ign x dam). Base timing (cell where rpm and calculated load intersect) could be 19 so your car is actually applying 24° ignition timing. If your dam is 0.75 and dynamic ignition is 5 with base 19, 19 + (3.75) is 22.75° total timing. Feedback and fine knock learn work the same way.. Sorta. The values you see is the amount of ignition degree that is pulled from the dynamic ignition advancement that is being applied. Any time a knock sum goes up on any cylinder, it pulls timing with feedback knock. If it sees another noise from the knock sensor before the last knock event decayed (time it takes for the ecu to forget the last event), it will pull additional timing until the noise stops or behavior is modified (lifting off the throttle). If the ecu recognize several feedback knock events under a certain calculated load and rpm range, it will automatically add the fbk timing into fine knock learn. When the car gets back into a range with fine knock, the ecu pulls timing regardless of if the knock sensor went off or not. If it successfully pulls timing and does not hear knock, it tries to lower the amount of timing decrease. (aka add advanced timing back). If the ecu sees a lot of knock in a short time, it will drop dam to minimize additional timing that may be causing knock... Whew.. Anywho... Now that knock has been explained... Let's get into real world situations. Noises that sound like pinging/knocking can cause the sensor to trigger a knock count increment.. Like exhaust banging around or even the ac compressor magnets click on. I noticed low rpm "increasing" load coming into boost, tends to be where the subi rumble causes things to vibrate the most. This is annoying because everything inside my car squeaks and rattles. I use to see 7° fbk on the highway on ots tunes and I found it was that I didn't have enough timing and the noise the ecu heard was getting worse as it pulled more timing. (running too rich). It can also happen if dynamic compression (aka avcs/cam advancement) and timing arent happy together. Hope that makes sense. Let me know if there's something that doesn't.
Derps got good points here too. Sorry just catching up on posts I missed. If the tunes wgdc were too aggressive on the low rpm range, it can bring boost on time in the summer (as more wgdc is needed) but cause tons of issues when air is denser. You can't tune it out but you can definitely tune it in. If it's a stock boost control solinoid, it could be a plugged wg boost line. Ive seen oil in my intercooler so I wouldn't be surprised if some oil makes it to that line as well. Ewg is never a bad idea but this plus top feeds are going to add up quick and your are getting close to the max torque that block can take. Ymmv, literally. Lastly, if you are still concerned, I'd say there's always the driver mod. I found myself less worried about blowing my car up when I stick to rolling into boost instead of jabbing the throttle.
If this started some time after the tune then something mechanical is going on not tune related. Feedback knock is typically detonation or noise being picked up (loose exhaust bolt, intercooler bracket loose, hood prop rattling) and not actual "rod knock". It's going to be hard to diagnose without someone inspecting it in person with data logging with experience because it can be 1 of many things! If you don't have the ability to go to your tuner or don't want to but feel comfortable with Shane and DB I would bring it to them. Shouldn't be too hard for them or any tuner to diagnose in person. Good luck!
I understand the concept of knock very well. Although knock numbers may be a sign of something bad, it does not at all mean that there's a blown motor - especially if that is the only symptom. Subarus have an especially sensitive knock sensor as well, as I'm sure you and everyone else know. As for tuners - I have no idea what I am doing. I need to do more research, and I may even be willing to check you guys out again. The last time I reached out was back in early 2015, so it has been a good while! I plan on checking the actual sensor to see if it's loose and then cleaning my injectors (something my tuner advised to try doing - he thinks this is causing the knock) before blaming my tuner for the problem. If that doesn't help, then I may seek out someone else edit: as stated above this problem has always been an issue with this car, ever since the tune. the car knocked some before the tune as well, but I can't remember if it was this specific problem. I feel like it was random knock before the tune vs now it is rpm and boost related and I can replicate it every single time
If you still have the ap flash the standard stage two program and see if you still have issues. You can always reflash back. Just a thought. I didn't read all of the posts here but good luck. Weather always plays a role in this as well
This isn't a bad place to start if you haven't talked to a tuner about the issue. This will help determine if the tune has too much timing or is too lean.. But you might not see any improvements because ots can generate a lot of noise too. Maybe less which is a good start. A different tuner can try a few things with the map to see if the behavior changes with different values. It's been brought up before but having someone look at the data logs with a comb will probably be the fastest route. In person is best because a stethoscope can be used.
The car is actually in storage so I can't do this until spring, but I may try this. The car did knock before the protune however. If I remember it correctly though, it was more random than it is now. It did it ever since I got the new tune, so I don't think weather is a factor in this. It happens every time, in spring, hot humid summer, and cool fall (I waited all the way until November to put it away this year) and the knock was always there every time cool air could totally be a factor in the overboosting though, as we all know. the car however, like the knock, has always overboosted around 2-3 psi - even on cobbs stge 1 tune when the car only had an intake on it
I saw this issue with Cobbs stage 1 tune yes, I just can't remember if it was always under 3k or not. I remember trying out different Cobb tunes to try to stop the knock but it always knocked
I've been sending logs over to the tuner for awhile now, he's sort of saying it's either nothing, or it's my injectors. I've always heard to never run cobbs off the shelf tune - i'm sort of surprised i'm getting so many recommendations to run it haha. I could try to post a log on here? since the car is in storage I only have old ones but that's okay
You could post it to google docs and then share the link to the spreadsheet. Has your DAM been consistently staying at 1.o (or 16 if you have a 16bit ecu...not sure what year your car is) or does it drop on occasion?
It is an 06 sti so it's 1. Before I got my tune, when I was running the Cobb stage 1 off the shelf tune for like two weeks or so, my car knocked so bad that it pulled timing and DAM dropped to like 0.025. My car hated the cobb tune. After the new protune it hasn't affected DAM. I put in bad gas one time and it did then - but it was due to the crap gas, not due to this problem.
I probably wouldn't worry too much if it was me. If it happens when you change load (like starting up a hill in high gear) just downshift don't lug the engine. Even if it isn't knocking don't lug the engine. I sometimes saw some random cruising knock/tip in knock like that on my 07 WRX if I was constantly monitoring knock retard and I just ignored it because it was in low load and not in boost. My car had a bigish 20G turbo on the stock motor and lasted a long time I never cracked a piston from knock (I lifted a head a little bit from too much boost but that is a different story). It could be "tip in" knock which is kind of hard to tune out. If you are really concerned and think its a problem and your tuner cant fix it then maybe just take it to JM Auto and have their tuner retune it, I have heard good things about him.
Ok Hondaboi. I read. I still wouldnt worry about it. It’s a subaru so it’s going to blow up anyway. But if you care so much and your tuner can’t figure it out just take it to Shane and retune it when you get your EWG to cure your boost creep issue.
