*** Knock, do you have any?

[BoozeRS05]

There should be some sorta "member donation" button on this site, so people can flick some people a few bucks to say thanks for their help - especially around reading and diagnosing peoples logs - it can be bloody costly through a mechanic. I'm sure gurus would be even more willing to help. God knows I owe a number of people on this site a few beers to say thanks. I get public forums are created for this very reason and there are some generous people around helping effectively complete strangers out, but reading some of these posts though, I get the gut feeling some of the guru's here must get frustrated at seeing the same shit turn up time after time... I got issues with my car at the moment but feel too rude to ask for help trawling through LV and log results time and time again without some sort of kick back for the help. Regardless, big kudos to iNVAR and similar guys for helping out. Blurt over.

 

There's a donation station and you're right, more people should do it.

 

Donate here…

 

probably too early to fully assess. your DAM is at 0.69 & since you mentioned you just updated the maps, you presumably just flashed the ECU & haven't driven enough under load for DAM to increase above the initial value. or, you recently had knock that caused DAM to drop to the current value. also, part throttle logs probably aren't going to be of too much use unless you're looking for something under part throttle conditions (like potential false knock during cruise, etc)

 

Yeah, I literally loaded the new map and took a short warm up drive, then took a crisp run up to 80mph for that log. I'll give it a couple hundred miles and try again with some more throttle. Thanks for the input.

[chimera]

There's a donation station and you're right, more people should do it.

Not just donating towards the site - donating towards individual members for helpful advice. Happy to donate to the site, but without the people this site (and any forum for that matter) = nothing.

[fishbone]

Good point, but consider that those who prefer to charge for their services already do so. Those that don't care about the monetary part, voluntarily contribute to this forum.

[iNVAR]

1) Your log as a PDF is useless. CSV file please.

2) You have a tuner. You should rely on your tuner to make adjustments for you, not us. You're paying for his service - he should provide it to you. Plus, if we give you advice or input and it's not in agreement with the tuner, it sort of puts everyone in an awkward/weird position. I'm willing to help but everyone does things sort of differently as far as their tuning methodology.

[crankshift]

hey all, anyone willing to have a look at this shortened down log for me and let me know if you think I am safe or if this is real knock (FKL logged up to -4.2 via cobb ap v3)?

 

this is revision 1 from a 'pro tune' using a heartbreak dyno to assist

 

thanks for any input, much appreciated, shop isnt responding at the moment

 

not looking for much other than feedback confirming there needs to be further changes or the car is safe and okay to be driving as is. I am not asking for anyone to make these changes for me nor am I asking for any detail about possible reasons why there could be an issue but that would be awesome too

datalog1-short-csv.csv

Edited March 11, 2014 by crankshift

[iNVAR]

Thanks for the CSV. Much easier to see things.

 

A couple things:

It's FLKC and it only occurs that one instance. The other times you see FLKC is the same recorded knock event and it's actually slowly unlearning itself... or rather it's learning to add timing back in as you can see by the values of FLKC slowly going to 0 over time. You can also tell because of the Knock Sum count that goes from 0 to 3, and just stays at 3 all the other instances.

 

It also occurs when you're not in load and also at a low RPM. High load + High RPM + Boost = boom. Low load + low RPM + vacuum = less likely to cause damage, and sometimes can just be a nuisance to figure out.

 

Check for loose objects in your engine compartment and keep an eye on that RPM + load + throttle area in future logs to see if it happens again.

[Lunartick]

Hello all,

I have so far been learning quite a bit from all the great resources on this forum as well as RomRaider's, thanks for all of your hard work and info sharing.

I recently flashed MickeyD's Hawaiian stg1 basemap and have been logging extensively with VAG-COM/RomRaider my daily commutes and relatively short WOT runs. Haven't gotten the chance to log in 3rd gear from 2k RPM to redline simply because I have not so far had the balls nor the opportunity to do so on a public road.

The reason I am writing here today is that I came across one (and only one so far) instance where my ECU used FBKC at high load (~2.5g/rev) to mitigate knock; I've kept a close eye on these two columns and I usually see FLKC used, this has been the first time I see the ECU applying FBKC to timing. Log is attached.

 

I would like to fix this potential issue so if you could please help me with which table/cells to adjust... Is it base timing at that nasty knock RPM/load area?

 

Btw I'm driving 05 LGT Ltd. 5EAT since April 2009, just in February I put in a new CHRA from ebay, new OEM banjo and OCV myself since the stock VF40 snapped at the shaft close to the turbine wheel (go figure). Thought about a BNR 16G but was out of my budget at the time.

 

Again, thanks much.

romraiderlog_20140311_182613.csv

[serx7]

Your AF Learning #1 indicates that it's pulling a small amt of fuel but is within accepted normal range I think (my tune disables this at high load). Do you usually see FLKC being applied in the same load & rpm range (~2.5 g/rev, 4200 rpm region)? If so, you might want to consider reducing timing slightly in that load/rpm region (Base Timing table) and smooth the values in neighboring cells.

[crankshift]

thanks for the feedback..

 

tuner is back now and has been helping me out some. he sent me a low boost map to get another log and another perspective to check things out.. I am waiting for him to call me back now and will let you know. he suspects ringland issues

[Lunartick]

Your AF Learning #1 indicates that it's pulling a small amt of fuel but is within accepted normal range I think (my tune disables this at high load). Do you usually see FLKC being applied in the same load & rpm range (~2.5 g/rev, 4200 rpm region)? If so, you might want to consider reducing timing slightly in that load/rpm region (Base Timing table) and smooth the values in neighboring cells.

 

I do not usually see FLKC at that RPM/load but then again I have not been driving or logging at WOT that often to know for sure. What I do know for sure (w.r.t. FLKC) is that it occurs at ~1-1.5 g/rev load / 2k-3.5k-ish RPM. Some LTV table snapshots are attached.

 

I would like to take back what I said regarding FBKC showing up only once so far. I went through my log history and found more, which is probably a REAL cause for concern. Please see attached. At one instance even knocking twice or more in a row (lower load than 2.5g/rev) !!!

 

Also attached is a log with both FLKC and FBKC being made: I am suspecting the FLKC's are learned corrections since knock sum does not increment. It does increment for the FBKC however.

FBKC_20140305_130659.csv

FBKC_20140307_070820.csv

Both_FB.FLKC_20140313_180435.csv

[Lunartick]

Please take a look at this log with FBKC also (attached). Seems that I definitely need to adjust timing at 4k to 4.4k RPM and 2.45 to 2.70 g/rev load. Sorry I didn't log FLKC on this one.

 

I would really appreciate if someone knowledgeable could advise how much to subtract from Base Timing cells.

 

I guess good news is that I am not seeing any FBKC at lower RPM's and high load. I'll try to keep the RPM's down during daily driving until I have adjusted base timing @ 4k and above RPM and high load.

 

Thanks!

romraiderlog_20140306_182559.csv

[iNVAR]

Some of that knocking is really bad, some isn't. You shouldn't be going nuts mixing your logs if they're on different tanks of gas. Your 2 logs that start with FBKC show a lot of timing being pulled, AND your IAM is already below 1, at one point hitting 0.6875, which means you were knocking a LOT.

 

Did you refill with new gas after that? Because I suspect you had crap gas in there.

 

If it was bad gas, that sort of only leaves us with 2 log files. The first log file you posted (in your first post) only shows one FBKC knock event and it occurred when you lifted off throttle. Your second log (BOTH FBKCFLKC) shows OLD FLKC (because Knock Sum doesn't increase) which is probably from the bad tank of gas. The FBKC in that log is also in a completely different area than the very first one you posted.

 

So reeeeeeally, if it was bad gas, you have 1 log showing FBKC at WOT... And maybe not even that because there might still be some bad gas mixed in.

 

If none of this was bad gas, then yeah, you'd have a problem. Get back to me.

[UnluckySkmr]

Im new to tuning and the accessport all together. I have been trying to learn for the most part what I am looking for, as far as knock and fuel goes. I have a 05 LGT flashed to stage 2 about a week ago. Everything seems to be leveled out so I did a 3rd gear pull. I couldn't get another pull that day which I was bummed about. What I am looking for is if this looks ok for a couple months till I can get a custom tune. I live on the east coast of NC so there are not really any tuners I trust around here with my subaru. Prob. going to go with infamous but open to suggestions as well. I also didnt get to redline and had to pull off throttle due to a sunroof flying off the car that was about 10 car lengths in front of me.

datalog 3rd gear pullDM.csv

[cryo]

Im new to tuning and the accessport all together. I have been trying to learn for the most part what I am looking for, as far as knock and fuel goes. I have a 05 LGT flashed to stage 2 about a week ago. Everything seems to be leveled out so I did a 3rd gear pull. I couldn't get another pull that day which I was bummed about. What I am looking for is if this looks ok for a couple months till I can get a custom tune. I live on the east coast of NC so there are not really any tuners I trust around here with my subaru. Prob. going to go with infamous but open to suggestions as well. I also didnt get to redline and had to pull off throttle due to a sunroof flying off the car that was about 10 car lengths in front of me.

 

Looks to be about on par for an ots map. Your fuel trims are in line for the most part and thinks look ok. Your fine until you figure out what you would like to do.

 

Have a great day.

 

Dave

[Lunartick]

Some of that knocking is really bad, some isn't. You shouldn't be going nuts mixing your logs if they're on different tanks of gas.

 

Thanks iNVAR, lesson learned!

 

Your 2 logs that start with FBKC show a lot of timing being pulled, AND your IAM is already below 1, at one point hitting 0.6875, which means you were knocking a LOT. Did you refill with new gas after that? Because I suspect you had crap gas in there.

 

I believe you are right on with your observation. I started thinking the same thing yesterday before you posted this and you are probably correct! I try to consistently refill with Mobil's 93 octane gas now which imho is the best I can find around here. However I do believe, beyond a reasonable doubt, that around that timeframe (March 4th,5th,6th logs -- flashed the stg1 basemap on March 3rd) I ALREADY had 91 octane in the tank and wasn't paying attention to this detail. I also filled a half-tank sometime last week from a gas station (not Mobil) in the ghetto, which was labeled 93 octane but who knows what crap they actually got in there.

 

If it was bad gas, that sort of only leaves us with 2 log files. The first log file you posted (in your first post) only shows one FBKC knock event and it occurred when you lifted off throttle.

 

I beg to differ. Snapshot of the first log I posted in my first post romraiderlog_20140311_182613.csv is below. I try to lift or press very gently on the throttle, but this knock happened at constant WOT. Throttle opening angle was held at 100% for another ~0.3 seconds after the knock occurred.

 

I believe you were referring to the knock shown on FBKC_20140305_130659.csv due to the abrupt change in throttle, I can't disagree on that one. Snapshot is below.

 

Your second log (BOTH FBKCFLKC) shows OLD FLKC (because Knock Sum doesn't increase) which is probably from the bad tank of gas. The FBKC in that log is also in a completely different area than the very first one you posted.

 

I agree, the knock and ensuing FBKC is at low load RPM/throttle so I'm not too worried about that. More snapshots shown below.

 

So reeeeeeally, if it was bad gas, you have 1 log showing FBKC at WOT... And maybe not even that because there might still be some bad gas mixed in. If none of this was bad gas, then yeah, you'd have a problem. Get back to me.

 

I am at least 85% sure it was bad gas (91 octane) but since I was freaking out yesterday I re-flashed mickeyd's stg1 map w/ a modified base timing table. Snapshot of the changes made is below.

 

I reset my ECU, filled-up on 93 octane gas from Mobil (tank had less than a quarter of the older gas) drove at WOT a little bit on my way home and logged it. Please see my yesterday's post-flash logs attached and let me know what you think. I had to shorten down the 0_WOT_log... size was too big since it was my whole commute.

 

EDIT: I am posting another post-flash log from yesterday where there is knock. Please take a look, thanks! romraiderlog_20140314_183923

0_WOT_log_20140314_192820.csv

mild_acceleration_20140314_194153.csv

romraiderlog_20140314_183923.csv

Edited March 15, 2014 by Lunartick
Posted 1 more log with FBKC from yesterday

[UnluckySkmr]

Looks to be about on par for an ots map. Your fuel trims are in line for the most part and thinks look ok. Your fine until you figure out what you would like to do.

 

Have a great day.

 

Dave

 

Thanks for taking a look. Hoping to get off the ots map as soon as possible.

[Lunartick]

A snapshot of that last log I posted is below.

I see severe knocking at around 3K RPM / 1.4 g/rev combo, total ignition timing is pulled way low to only 11 degrees. -2.46deg FLKC is applied as well as FBKC, meaning that this is not the first time during the commute that the ECU sees knock at this range. This is with good gas, IAM is 100% and A/F learning is 0.

 

How should I address this issue? How much timing would you advise I should pull from Base Timing?

[serx7]

A snapshot of that last log I posted is below.

I see severe knocking at around 3K RPM / 1.4 g/rev combo, total ignition timing is pulled way low to only 11 degrees. -2.46deg FLKC is applied as well as FBKC, meaning that this is not the first time during the commute that the ECU sees knock at this range. This is with good gas, IAM is 100% and A/F learning is 0.

 

How should I address this issue? How much timing would you advise I should pull from Base Timing?

 

The base + advance timing values in the 1.4 g/rev 3000 rpm area that you're seeing the knock, appears to be pretty much the same as the stock 05 LGT AT ROM, at least from what I'm seeing.

 

Are you certain there isn't anything rattling around underhood that might be tripping the knock sensor? I saw funky knock behavior like that on my car when the motor was still somewhat cold, but I have forged internals.

 

Maybe flash back to the stock ROM and see if you see the same behavior. If you're running good gas and still see that kind of knock response on the stock ROM, maybe the knock sensor is marginal or mis-torqued?

[Lunartick]

Good idea, I'll flash back to the stock ROM and take some logs and see what we're dealing with.

 

Is there any other sure-fire way to find something that may be rattling underhood?

 

If not something rattling, then do you know how easy is it to get to the knock sensor to inspect it? I found its position in the service manual, not sure I understand the removal process.

[serx7]

Good idea, I'll flash back to the stock ROM and take some logs and see what we're dealing with.

 

Is there any other sure-fire way to find something that may be rattling underhood?

 

If not something rattling, then do you know how easy is it to get to the knock sensor to inspect it? I found its position in the service manual, not sure I understand the removal process.

 

Cool, let us know.

 

I'd probably just pop the hood (and get under the car) and bang on everything I could get a fist on. Don't worry too much about stuff that isn't somehow directly connected to the motor. On my prev car (Stg 2 05 Saabaru 9-2X Aero), I had a rattling turbo heatshield that was causing all kinds of timing to be pulled. Couldn't hear it w/ the windows up, but the first time I had the windows partway down and accelerated/decelerated through a 'problematic' RPM range, the rattle was pretty much unmistakable.

 

As for removing the knock sensor, not sure. I've never removed mine myself but when I was contemplating it, I seem to recall seeing a few decent writeups (some w/ decent pics/diagrams) here or on nasioc. I don't think there's much to inspect, unless you find a stray bolt flopping around where the knock sensor is located.

[Lunartick]

I know for sure it's not the turbo heatshield since I made sure that's installed securely (with ALL of the fasteners that come with it) when I serviced the turbo circa mid-February.

I'll try the fist-bang method, and also drive next to a wall with the windows down. Thanks.

[serx7]

I know for sure it's not the turbo heatshield since I made sure that's installed securely (with ALL of the fasteners that come with it) when I serviced the turbo circa mid-February.

I'll try the fist-bang method, and also drive next to a wall with the windows down. Thanks.

 

I was just using the turbo heatshield as an example.

 

Last yr I was diagnosing a very consistent shift knock problem. I tried reducing timing in the next gear landing zone, fiddled w/ tip-in, even put some Torco in the tank to get octane up to around 100 (my car is pro-tuned to run somewhat fat but I didn't have a WBO2 on hand, so I figured if it was a temporarly lean condition b/n gears, the higher octane might reduce the knock tendency). Still had single or double count FBKC on almost every 1-2 and 2-3 upshift. A month or so later I had the car on ramps and noticed the rear diff was wet. Turned out my Nameless midpipe was resting against the diff casing, and the heat transfer had compromised the diff's main seal. Suspected it was also banging against the diff b/n shifts at high-ish RPM. I got the diff leak fixed and the pipe off the diff, and the shift knock almost completely went away.

[Lunartick]

Ok so based on your rear diff experience, we can safely conclude that the knock sensor is prone to drivetrain noise as well. How about road noise or 'clunking' in the suspension? Could that kind of noise also trigger the sensor?

 

Assuming that the knock my ECU is seeing is noise: Does anyone foresee any serious consequences for the engine while the knock control sys is pulling timing (through either FB or FLKC) while it didn't necessarily have to, i.e. there was a false incerement (or a few false incerements) of knocksum due to drivetrain or rattling noise?

 

I remember reading on this post on RomRaider forums by NSFW says less timing is better. Is less timing always better?

[serx7]

Ok so based on your rear diff experience, we can safely conclude that the knock sensor is prone to drivetrain noise as well. How about road noise or 'clunking' in the suspension? Could that kind of noise also trigger the sensor?

 

Assuming that the knock my ECU is seeing is noise: Does anyone foresee any serious consequences for the engine while the knock control sys is pulling timing (through either FB or FLKC) while it didn't necessarily have to, i.e. there was a false incerement (or a few false incerements) of knocksum due to drivetrain or rattling noise?

 

I remember reading on this post on RomRaider forums by NSFW says less timing is better. Is less timing always better?

 

The banging of the midpipe against the rear diff isn't really drivetrain noise per se (a noisy clutch would be a better example). My tuner and I do suspect drivetrain noise is causing or at least contributing to ongoing intermittent instances of knock right after some shifts. I have datalogs w/ identical shift points and landing RPM but completely different post-shift engine loads, recording single counts of FBKC at the same RPM in the next higher gear (eg: upshift 2nd to 3rd at 5k rpm going WOT into the next gear, upshift 2nd to 3rd gear at 5k rpm going half throttle into the next gear). Just sitting at idle w/ the clutch pedal in, I hear a very audible, sharp metallic 'clack' when I quickly let the clutch pedal out. I'm pretty sure that isn't helping.

 

I doubt road noise or most suspension noise would trigger a response from the ECU since the ECU is listening for noise that's characteristic of detonation. I've hit a few nice potholes w/ the logger running and the only thing I had to worry about was a bubble in the sidewall

 

I don't think you have to worry about the timing pulls causing long term problems. Esp not the FBKC since that decays to 0 after a few seconds even when it pulls a lot of timing. Also, note that an increase in knocksum does not always result in the ECU pulling timing.

 

I think what NSFW was referring to w/ the 'less [timing] is better' idea, was slightly reducing the highest timing values in the lower load/middling high RPM cells so that the timing delta that occurs when you mash the throttle at say 4k RPM isn't as large. The large and rapid timing change going from lower load @ cruise to high load @ WOT can cause knock. You can see how much and how quickly the timing changes by looking at logs where you hold a given RPM at part throttle for a short time before going WOT, esp at RPM where the turbo spools quickly. You can also see this just by looking at the base timing table, going across it left to right from say ~1 g/rev & 4k rpm to ~2.5 g/rev & 4k rpm. In the stock 05 LGT AT base timing table, you're looking at a base timing swing from ~40 deg to ~14 deg. Granted at 1 g/rev & 4k rpm there's 0 in the Advance table and ~4 deg Advance in the high load cells, but overall it's still a pretty big change in timing.

[chato]

In my extensive searching on the issues of false knock, there have been instances where suspension noise was triggering it, usually bad swaybar endlinks.

 

my turbo heat shield was on tight but was actually touching the bellhousing which was causing some vibrations.

 

Best way I found to check is to have someone else rev the car while you check all around for rattles and other out of place noises. Focus on things attached to the engine.

 

as for removing the knock sensor, it was kind of a pain, it's below the throttle body. I couldn't get a socket in there so I ended up just removing the TB to take it off and inspect. I did find some random crap sitting on top of the engine while I was at it. Nothing to really see on the knock sensor unless it's cracked or improperly torqued.