Feedback Knock Question!

Hello everyone! I have an 05 LGT with 1000 miles on a new OEM rebuild, BNR 16G, and Invidia from the Up-pipe back. Other than that she's stock for now. I notice every so often if I'm just pulling out from a stop light or sometimes when shifting my feedback knock will hit -2 then go back to zero for the remainder of my drive. But today I was doing a pull in 1st and noticed when I went into 2nd it hit -11.2 then dropped back to zero.. I haven't seen any leaks and from the what I've read it could just be false knock but I'm definitely not the best at this kind of stuff. Oil still looks brand new and still keeping steady pressure. Fine Knock stays at 0 and I'm still hitting target boost. Any ideas on what else it could be and if it's something that will cause me to think of another rebuild in the near future?

You should definitely start by getting a log to your tuner right away.

That was my plan but it doesn't seem to be doing it consistently. I've driven 50+ miles today with no numbers coming up but heck, hopefully he'll see something with another value that I'm not seeing. Thank you for the advice!

Read this, it should help you understand how knock is detected http://www.romraider.com/forum/topic1840.html

not sure if you have an accessport or running opensource, if you want some really good articles from Cobb about knock in addition to the romraider post above, check out these:

 

SUBARU KNOCK MONITORING AND DIT DAM STRATEGY CHANGES

http://www.cobbtuning.com/subaru-knock-monitoring-and-dit-knock-logic-changes/

(all info applies except for the last section about the new DIT engines, the rest of the article is still applicable to you)

 

Knock Monitoring FAQ

https://cobbtuning.zendesk.com/hc/en-us/articles/200025084-Knock-Monitoring-FAQ

 

How to Datalog

https://cobbtuning.zendesk.com/hc/en-us/articles/200025134-How-to-Datalog

 

you might want to touch base with your tuner to make sure they are in the loop and can review your logs and offer advice.

Read this, it should help you understand how knock is detected http://www.romraider.com/forum/topic1840.html

Thank you for the link. That cleared up my mind on a lot of things.

not sure if you have an accessport or running opensource, if you want some really good articles from Cobb about knock in addition to the romraider post above, check out these:

 

SUBARU KNOCK MONITORING AND DIT DAM STRATEGY CHANGES

http://www.cobbtuning.com/subaru-knock-monitoring-and-dit-knock-logic-changes/

(all info applies except for the last section about the new DIT engines, the rest of the article is still applicable to you)

 

Knock Monitoring FAQ

https://cobbtuning.zendesk.com/hc/en-us/articles/200025084-Knock-Monitoring-FAQ

 

How to Datalog

https://cobbtuning.zendesk.com/hc/en-us/articles/200025134-How-to-Datalog

 

you might want to touch base with your tuner to make sure they are in the loop and can review your logs and offer advice.

I'm using the AP as of right now. From what I read it seems to be normal for the -2 to be pulled since it's mainly while going from a stop and from shifting. I'm still going to do a log and send to my tuner because the -11 still scares me a little, even though it was a 1st gear pull. Now is there a way to shorten logs from the AP? I figure I'll have to log maybe 3-5 minutes at a time since the FBKC only occurs every so often.

yeah its hard to sift through really long logs. the typical logs you want to take when dialing in a car is 1) idle log for a minute or two 2) cruise log normal driving for a minute or two, no full throttle, just cruise. 3) 3rd gear pull - as described in the "how to datalog" video. that should capture everything necessary. anything else that happens is usually a one-off, or mechanical thing like a loose bolt causing false knock, etc.

yeah its hard to sift through really long logs. the typical logs you want to take when dialing in a car is 1) idle log for a minute or two 2) cruise log normal driving for a minute or two, no full throttle, just cruise. 3) 3rd gear pull - as described in the "how to datalog" video. that should capture everything necessary. anything else that happens is usually a one-off, or mechanical thing like a loose bolt causing false knock, etc.

Okay thank you for that information. Never really had to do any logs besides one for seeing if my front O2 was bad so I'm still getting used to the whole when to log thing. This greatly helps my understanding.

I usually just shorten the logs by opening/editing them in Excel (preserve the column headers, of course).

 

I chased down similar FBKC behavior several years ago. In my case it was false knock mainly due to my IPR TMIC having been positioned such that it just barely rested on top of the pitch stop mount. The FBKC didn't occur often at all, and I had to log several entire 20+ min drives before I was able to see a pattern that led me to where to look in the engine bay.

 

IME you'll probably want to log full time until you see it happen, at which point you can stop the logger. Every time I tried to do logging over shorter intervals, I would see the big FBKC show up very shortly after I had stopped logging, or would see it show up as I was just getting ready to start logging

 

Edit: I forgot you said the big FBKC occurred at an upshift. Could also be shift knock, or something rattling under the hood as a result of the upshift. I chased down a pesky shift knock issue a few yrs ago as well, it was a tedious exercise.

I usually just shorten the logs by opening/editing them in Excel (preserve the column headers, of course).

 

I chased down similar FBKC behavior several years ago. In my case it was false knock mainly due to my IPR TMIC having been positioned such that it just barely rested on top of the pitch stop mount. The FBKC didn't occur often at all, and I had to log several entire 20+ min drives before I was able to see a pattern that led me to where to look in the engine bay.

 

IME you'll probably want to log full time until you see it happen, at which point you can stop the logger. Every time I tried to do logging over shorter intervals, I would see the big FBKC show up very shortly after I had stopped logging, or would see it show up as I was just getting ready to start logging

 

Edit: I forgot you said the big FBKC occurred at an upshift. Could also be shift knock, or something rattling under the hood as a result of the upshift. I chased down a pesky shift knock issue a few yrs ago as well, it was a tedious exercise.

That's what I was gonna do:lol:

I figured it was just a rattle or false knock since the car drives perfect from my understanding. I took it out to my tuner 2 weeks ago and mentioned it and he drove it for an hour and said it drives great. Just big numbers scare me when Knock is in the heading:(

I'm gonna start logging my daily trips so I can hopefully catch it and start looking under in the bay to see if I can find the pesky rattle. Thank you!

Assuming tuner retained the original FBK Retard Increment Delay (which governs how long the knock signal must be clear before the ECU starts adding the pulled timing back in) and the original FBK Retard Increment (how much timing is added back at a time once the ECU decides it can safely start adding the pulled timing back), big FBKC hits like what you saw can easily keep timing reduced throughout the entire next gear (assuming the FBKC is applied right after an upshift like you reported). I'd be willing to bet if you saw -11 deg FBKC right after a 1-2 upshift and stayed in the throttle, the ECU probably would have added only ~half of that pulled timing back before you finished winding out 2nd gear. IME a quick hit of -2 deg that decays isn't very noticeable seat of the pants. A hit of -11 deg right after an upshift is -very- noticeable.