Welcome to the North American Subaru Impreza Owners Club Thursday September 3, 2015
Home Forums WikiNASIOC Products Store Modifications Upgrade Garage
NASIOC
Here you can view your subscribed threads, work with private messages and edit your profile and preferences Home Registration is free! Visit the NASIOC Store NASIOC Rules Search Find other members Frequently Asked Questions Calendar Archive NASIOC Upgrade Garage Logout
Go Back   NASIOC > NASIOC Technical > Engine Management & Tuning > AccessPort

Welcome to NASIOC - The world's largest online community for Subaru enthusiasts!
Welcome to the NASIOC.com Subaru forum.

You are currently viewing our forum as a guest, which gives you limited access to view most discussions and access our other features. By joining our community, free of charge, you will have access to post topics, communicate privately with other members (PM), respond to polls, upload content and access many other special features. Registration is free, fast and simple, so please join our community today!

If you have any problems with the registration process or your account login, please contact us.
* Registered users of the site do not see these ads.
Reply
 
Thread Tools Display Modes
Old 08-04-2012, 09:49 PM   #1
cueball89
Scooby Specialist
 
Member#: 90513
Join Date: Jul 2005
Chapter/Region: Tri-State
Location: NY
Vehicle:
2003 wrx wagon
black

Default Hybrid knock, false knock?

Trying to get a break in e tune on my hybrid. Tuner thinks it's false knock. It was pulling 2 degrees of timing, he'd remove 2 on the next revision and then the car would pull another two so on and so forth. Since then I've been chasing down rattles and anything else in the engine bay that could cause the issues. I do have a noisy throw out bearing on deceleration.

Unbolted the knock sensor and reinstalled it as per fsm and now I'm getting crazy high fbck and little to no flkc, the dam is unaffected. I'll change something in the engine bay and think I have it fixed then a ton of knock. Sanded down all the grounds on the car. Have rubber on anything I think that is even remotely capable of knocking.

99% of the knock is under 2800 and under load values of 1. I changed out my lw crank pulley for the stock one and got some fbck at loads of 1.1. These are the 2 latest logs, one is 50% throttle through the gears shifting at 4k rpm. That's the most aggressive I've been on the new motor since I installed it. No knock, nothing.

https://docs.google.com/open?id=0B-f...mpXLUxseHcwdVk

The next log is me driving around like I normally do. Some times the knock is there and sometimes it not. The load and rpm sites are similar but I can't always reproduce it. In other logs I have it pulls up to 12 degrees of timing.

Its a long log, it's pulling timing at 4, 13, 85, 197, 208 and 286 seconds. Sometimes the conditions are similar other times they aren't, sometimes the conditions stay the same and the knock stops for itself.

https://docs.google.com/open?id=0B-f...Ho2UUN5TTk5cnM


Is this just noisy phantom knock on a hybrid?



Mods
sti shortblock
asf chamber matched 2 liter heads with valve job, pocket port and sti cams.
grimmspeed 3 port boost controller
grimmspeed aos
grimmspeed up pipe
td04 turbo
catless tbe
perrin turbo inlet
cobb ap2
* Registered users of the site do not see these ads.
cueball89 is offline   Reply With Quote
Sponsored Links
* Registered users of the site do not see these ads.
Old 08-04-2012, 10:06 PM   #2
BlackJavaPearl04
Scooby Specialist
 
Member#: 198522
Join Date: Dec 2008
Chapter/Region: Tri-State
Location: Up State NY
Vehicle:
98 LGT Limited
00 LGT -WRX swap

Default

I have the same problem but not in that rpm range. We had knock feedback with my hybrid build (forged pistons) between 5500-6k. Its a funky thing with the 2.5 block and 2.0 heads I guess. The first time we discovered it was with a fully built motor with big cams/ built heads, 35r turbo. So I made what are called 'det cans' you can google how to make them, but what they are if you dont know already is basically like a mechanical means of listening to the motors vibration under load. In order to get the ecu to stop pulling timing we had to dumb down the knock sensor and use the det cans to listen. We never heard anything aside from the typical piston noises. Now im on another build...same block but with stock heads and a vf39...still getting knock in that rpm range, so we did the same thing and all is well. The only thing is there is no fail safe in place with the sensor dumbed down if I were to have an actual knock event. BTW if you can swing it I recommend upgrading from the td04 to a vf series turbo as the little td04 doesnt perform as well on a 2.5 as it does with the 2.0.
BlackJavaPearl04 is offline   Reply With Quote
Old 08-05-2012, 09:22 AM   #3
Cobb Tuning
NASIOC Vendor
 
Member#: 4803
Join Date: Mar 2001
Location: Austin, DFW, Portland, SOCAL
Default

Quote:
Originally Posted by cueball89 View Post
Trying to get a break in e tune on my hybrid. Tuner thinks it's false knock. It was pulling 2 degrees of timing, he'd remove 2 on the next revision and then the car would pull another two so on and so forth. Since then I've been chasing down rattles and anything else in the engine bay that could cause the issues. I do have a noisy throw out bearing on deceleration.

Unbolted the knock sensor and reinstalled it as per fsm and now I'm getting crazy high fbck and little to no flkc, the dam is unaffected. I'll change something in the engine bay and think I have it fixed then a ton of knock. Sanded down all the grounds on the car. Have rubber on anything I think that is even remotely capable of knocking.

99% of the knock is under 2800 and under load values of 1. I changed out my lw crank pulley for the stock one and got some fbck at loads of 1.1. These are the 2 latest logs, one is 50% throttle through the gears shifting at 4k rpm. That's the most aggressive I've been on the new motor since I installed it. No knock, nothing.

https://docs.google.com/open?id=0B-f...mpXLUxseHcwdVk

The next log is me driving around like I normally do. Some times the knock is there and sometimes it not. The load and rpm sites are similar but I can't always reproduce it. In other logs I have it pulls up to 12 degrees of timing.

Its a long log, it's pulling timing at 4, 13, 85, 197, 208 and 286 seconds. Sometimes the conditions are similar other times they aren't, sometimes the conditions stay the same and the knock stops for itself.

https://docs.google.com/open?id=0B-f...Ho2UUN5TTk5cnM


Is this just noisy phantom knock on a hybrid?



Mods
sti shortblock
asf chamber matched 2 liter heads with valve job, pocket port and sti cams.
grimmspeed 3 port boost controller
grimmspeed aos
grimmspeed up pipe
td04 turbo
catless tbe
perrin turbo inlet
cobb ap2
No preview for the logs so I am unable to view, but false knock is common with a hybrid set up on these cars. You are changing the noise characteristics of the motor relative to the original 2.0L block, for which the OEM calibration of the knock sensor is designed for. The tables that govern this are available in the AccessTUNER software, but it is something that can be complex to tune and likely would require that the tuner have the vehicle in person (and some means to determine knock from false knock, such as det cans). Note that this is different than shutting off the knock response in certain area - this would actually be effectively manipulating the noise thresholds that determine a knock event based on the new noise characteristics of the motor.

Bill
Cobb Tuning is offline   Reply With Quote
Old 08-05-2012, 10:10 AM   #4
cueball89
Scooby Specialist
 
Member#: 90513
Join Date: Jul 2005
Chapter/Region: Tri-State
Location: NY
Vehicle:
2003 wrx wagon
black

Default

Thank you for the responses guys. I'm off to lowes to make some det cans today.



50% throttle with preview
https://docs.google.com/spreadsheet/...FlpeHpBWGNaTkE


regular driving where the knock has been thus far with preview.
https://docs.google.com/spreadsheet/...TlkWVFqeExOcXc
cueball89 is offline   Reply With Quote
Old 08-07-2012, 12:30 AM   #5
cueball89
Scooby Specialist
 
Member#: 90513
Join Date: Jul 2005
Chapter/Region: Tri-State
Location: NY
Vehicle:
2003 wrx wagon
black

Default

How close does the knock sound to these samples?
http://theknockbox.com.au/the-knock-...nd-recordings/

Is it a large increase in sound like in the 2500rpm sample? I've heard knock described as frying bacon, welding and bb's in a can. I can't say that I've heard knock on this motor. It's not as clean sounding as the samples with no knock but it's not like the ones with knock either imo.
cueball89 is offline   Reply With Quote
Old 08-14-2012, 04:31 AM   #6
BlackJavaPearl04
Scooby Specialist
 
Member#: 198522
Join Date: Dec 2008
Chapter/Region: Tri-State
Location: Up State NY
Vehicle:
98 LGT Limited
00 LGT -WRX swap

Default

you will hear alot of tapping and vibrating but yes the knock will sound like bbs in a can.
BlackJavaPearl04 is offline   Reply With Quote
Old 02-02-2013, 07:32 AM   #7
cueball89
Scooby Specialist
 
Member#: 90513
Join Date: Jul 2005
Chapter/Region: Tri-State
Location: NY
Vehicle:
2003 wrx wagon
black

Default

Bumping this back up.

I've been doing some reading on tuning the knock sensor and such. I see a few ways I can possibly change the knock sensor so that it isn't so dramatically affected.


Feedback Knock Retard Activation (Min. Load) I can up the minimum load in which the fbkc happens. This should prevent most major knock correction from happening and being learned to the flkc? I'd be driving around with no knock protection/detection under .8 or .9 load.

Fine Knock Learning Modify( Load Range) Again up the minimum load range to .8 or .9. I'll still have instantaneous knock correction, non of it will be learned by the flkc. I'll have to log the car more, but fbkc doesn't seem to be happening a lot during driving these days. A lot of flkc. I think this might be my best option.

Knock Threshold Level Final Limit (Min). Increasing this value will desensitize the knock sensor, it will become more numb over its entire operation. I don't think I want to go this route because it's only under light loads I'm getting weird knock. I don't want to sacrifice the knock sensor where its working properly.


There are a few other tables but I'm not exactly sure how they work, background noise and knock sensor calibration.


I think the FLKC load range changes sounds like my best option?
cueball89 is offline   Reply With Quote
Old 02-02-2013, 10:06 AM   #8
Cobb Tuning
NASIOC Vendor
 
Member#: 4803
Join Date: Mar 2001
Location: Austin, DFW, Portland, SOCAL
Default

Quote:
Originally Posted by cueball89 View Post
Bumping this back up.

I've been doing some reading on tuning the knock sensor and such. I see a few ways I can possibly change the knock sensor so that it isn't so dramatically affected.


Feedback Knock Retard Activation (Min. Load) I can up the minimum load in which the fbkc happens. This should prevent most major knock correction from happening and being learned to the flkc? I'd be driving around with no knock protection/detection under .8 or .9 load.

Fine Knock Learning Modify( Load Range) Again up the minimum load range to .8 or .9. I'll still have instantaneous knock correction, non of it will be learned by the flkc. I'll have to log the car more, but fbkc doesn't seem to be happening a lot during driving these days. A lot of flkc. I think this might be my best option.

Knock Threshold Level Final Limit (Min). Increasing this value will desensitize the knock sensor, it will become more numb over its entire operation. I don't think I want to go this route because it's only under light loads I'm getting weird knock. I don't want to sacrifice the knock sensor where its working properly.


There are a few other tables but I'm not exactly sure how they work, background noise and knock sensor calibration.


I think the FLKC load range changes sounds like my best option?
When dealing with this kind of problem, you are going to have to have some way of distinguishing false knock from actual knock. Usually this is accomplished via "det cans". This is especially important if you are going to adjust knock sensitivity.

The easiest course of action is to modify the fine knock learning and/or feedback knock correction load or RPM ranges, depending on what would work best for what you are seeing. If you disable fine knock learning at lower load by raising the load limit, then feedback knock correction will still come into play in its place when perceived knock is present (assuming you didn't also raise its min. load). That might be better, though, than a learned correction.

Bill
Cobb Tuning is offline   Reply With Quote
Old 11-04-2013, 09:26 PM   #9
cueball89
Scooby Specialist
 
Member#: 90513
Join Date: Jul 2005
Chapter/Region: Tri-State
Location: NY
Vehicle:
2003 wrx wagon
black

Default

Bit of an update to this thread. I sorted out the knock issues on the TD04 and was able to get successfully tuned. I install an 18g xtr and on wastegate pressure everything seems ok. Start to turn up the boost and then I start getting weird knock that wont go away again, thinking false knock. On the latest revision I was able to get one clean run in with no knock, probably the only pull. At this point the clutch would intermittently slip so I replaced it thinking the slipping clutch could have been contributing to the false knock. The latest revision is a log from today, I'm basically back where I was preclutch install, knock.

I don't have the afr in the logs, but I have been/ am being tuned with a wideband, just didnt have it in for these runs. It's tuned at 11.0 afr and it hits ~11.0. I know I'm at the limit of the maf, the plan is to run sd in maf mode to gain headroom once I get this knock sorted out.


I think it's a relatively conservative tune thus far, relatively low boost and low timing. It "shouldn't" be knocking like this on 93oct. Looking for ideas. Pretty stumped. Thinking of seeing how the car acts if I relocate the knock sensor up a few inches to where I had my det can installed.

Mods
sti shortblock
asf chamber matched 2 liter heads with valve job, pocket port and sti cams.
grimmspeed 3 port boost controller
grimmspeed aos
grimmspeed up pipe
18G XTR
catless tbe
perrin turbo inlet
cobb ap2
08 STI TMIC
walbro fp
ID1000
PTP turbo blanket
1 step colder Iridium plugs

TD04 final tune
https://docs.google.com/spreadsheet/...rive_web#gid=0

18G wastegate pressure
https://docs.google.com/spreadsheet/...=sharing#gid=0

Rare clean run pre clutch install
https://docs.google.com/spreadsheet/...rive_web#gid=0

Latest run post clutch install
https://docs.google.com/spreadsheet/...rive_web#gid=0
cueball89 is offline   Reply With Quote
Old 03-23-2014, 08:34 PM   #10
2slofouru
NASIOC Supporter
 
Member#: 90539
Join Date: Jul 2005
Chapter/Region: South East
Location: somewhere east of CA
Vehicle:
95 L with ej257k
powa, Raven R6

Default

Have you tried a different knock sensor? Maybe the one you are using is hyper sensitive at the peak frequency. What about a bandpass filter to eliminate the wrong frequencies from making it to the ecu, or maybe the ecu's filter isn't sharp enough?

Also, what Cobb stated is correct. Your block having larger pistons will have a different knock frequency / Hz and your knock sensor may be designed to pick up the 2.0 ecu and block knock for example.

Last edited by 2slofouru; 03-23-2014 at 08:40 PM.
2slofouru is offline   Reply With Quote
Reply

Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Forum Jump


All times are GMT -4. The time now is 02:55 AM.


Powered by vBulletin® Version 3.7.0
Copyright ©2000 - 2015, Jelsoft Enterprises Ltd.
Powered by Searchlight © 2015 Axivo Inc.
Copyright ©1999 - 2015, North American Subaru Impreza Owners Club, Inc.