Welcome to the North American Subaru Impreza Owners Club Thursday July 24, 2014
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 > Commercial ECU Reflashes

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 08-17-2007, 01:00 PM   #1
Zumble
Scooby Specialist
 
Member#: 34860
Join Date: Apr 2003
Chapter/Region: Tri-State
Location: NY
Vehicle:
2005 Forester CGM

Default Low load pulling timing?

My ECU has decided it wants to pull 2 degrees of timing in the 3200-4000 rpm in the low load areas. Areas below .85 load. I've tried pulling timing in those areas and nothing. It's not pulling timing between .85 and 1.0 load. The ECU doesn't make fine corrections below .75 or something with the stock tune. I'm still running the stock pistons and such. Anyone have experience with this? Should I just adjust the fine correction range to turn on at .90 and not listen below that? I was thinking of maybe throwing some torco in the tank to see if the problem still persists.

Mods as follows:

Stock EJ255 Block
TD05H-16G Turbo
TXS Downpipe to SMB Metal Substrate Hi-flo cat to STi Axleback
Cusco Heat Shield
Cusco Catch Can
STi Uppipe
DW 740cc Injectors
Perrin Turbo inlet hose
Zerosports Intake Elblow
APS 65mm CAI
Perrin TMIC
* Registered users of the site do not see these ads.
Zumble is offline   Reply With Quote
Old 08-17-2007, 01:52 PM   #2
Tea cups
Scooby Specialist
 
Member#: 103136
Join Date: Dec 2005
Default

Quote:
Originally Posted by Zumble View Post
The ECU doesn't make fine corrections below .75 or something with the stock tune.
It doesn't make adjustments to the fine learning table below .75 g/rev, but that doesn't mean that fine correction isn't being APPLIED. If you are using Enginuity, you would need to look at the 'Fine Correction Columns (Load)' values. The first value (.85 g/rev factory value on the 05 fxt) is the range 0-.85 g/rev. So, any fine correction adjustment made between .75-.85, in your example, would be applied in the range 0-.85 g/rev (except for the idle range for which these corrections are NOT applied). So, if you believe this to be based on false knock, you would want to raise the 'Fine Correction Range (Load)' enable value above .85 g/rev so that effectively no adjustments are made in the 0-.85 g/rev range.
Tea cups is offline   Reply With Quote
Old 08-17-2007, 01:58 PM   #3
Zumble
Scooby Specialist
 
Member#: 34860
Join Date: Apr 2003
Chapter/Region: Tri-State
Location: NY
Vehicle:
2005 Forester CGM

Default

Yes, I am using enginuity for the basemap flashes. I understand about the fine correction table etc. What are the chances of actual knock in those areas? I've tried stock timing/avcs/fueling tables in this area, I've tried STi stage 2 values, I've tried Cobb VF39ed FXT values and nothing gets rid of it. I think I'll try the torco tonight then. My TXS knocklite is also quiet in these areas.

Quote:
Originally Posted by Tea cups View Post
It doesn't make adjustments to the fine learning table below .75 g/rev, but that doesn't mean that fine correction isn't being APPLIED. If you are using Enginuity, you would need to look at the 'Fine Correction Columns (Load)' values. The first value (.85 g/rev factory value on the 05 fxt) is the range 0-.85 g/rev. So, any fine correction adjustment made between .75-.85, in your example, would be applied in the range 0-.85 g/rev (except for the idle range for which these corrections are NOT applied). So, if you believe this to be based on false knock, you would want to raise the 'Fine Correction Range (Load)' enable value above .85 g/rev so that effectively no adjustments are made in the 0-.85 g/rev range.
Zumble is offline   Reply With Quote
Old 08-17-2007, 07:19 PM   #4
Tea cups
Scooby Specialist
 
Member#: 103136
Join Date: Dec 2005
Default

Get some det cans if you want to be sure. False knock at low loads seems to be fairly common from what I've seen. You need to also look at feedback correction as fine correction is applied across a rather wide load/rpm range and is learned correction under very specific circumstances whereas feedback correction can be interpreted as immediate negative correction due to what the ecu perceives as knock.
Tea cups is offline   Reply With Quote
Old 08-17-2007, 08:50 PM   #5
Zumble
Scooby Specialist
 
Member#: 34860
Join Date: Apr 2003
Chapter/Region: Tri-State
Location: NY
Vehicle:
2005 Forester CGM

Default

I'm going to have to read up on det cans. You've lost me on the feedback correction and fine correction.

The false knock at low loads being common is kind of reassuring.

In your experience have you had to up the fine correction threshold to avoid false knock?


Quote:
Originally Posted by Tea cups View Post
Get some det cans if you want to be sure. False knock at low loads seems to be fairly common from what I've seen. You need to also look at feedback correction as fine correction is applied across a rather wide load/rpm range and is learned correction under very specific circumstances whereas feedback correction can be interpreted as immediate negative correction due to what the ecu perceives as knock.

Last edited by Zumble; 08-17-2007 at 08:56 PM.
Zumble is offline   Reply With Quote
Old 08-17-2007, 09:40 PM   #6
Tea cups
Scooby Specialist
 
Member#: 103136
Join Date: Dec 2005
Default

I don't worry about it at those low loads and haven't had an issue with fine correction, although I have an 02 wrx. Feedback 'knock' correction is another parameter you can log with Enginuity. It is one of the three elements that make up KC (the other two are fine learning correction and dynamic advance). Fine correction is learned correction meaning that just because you are logging negative correction doesn't mean the ecu is seeing knock at that point in time. It could have any time in the past (however table is reset after an ecu reset or when IAM changes). However if you are logging negative feedback correction means the knock signal is set at that point in time (with a slight delay afterwards after the knock signal is clear).
Tea cups is offline   Reply With Quote
Old 08-17-2007, 09:59 PM   #7
Zumble
Scooby Specialist
 
Member#: 34860
Join Date: Apr 2003
Chapter/Region: Tri-State
Location: NY
Vehicle:
2005 Forester CGM

Default

I'm actually using street tuner to log. I use a ScanMaster3 to watch the KC in real time as well as using ST realtime dashboard. I know I keep seeing -2 because there is no dynamic advance in the map in that area but there is -2.10 in the knock correction table. Usually when I get a neg value in the KC table I can pull some timing or richen it up and it will learn to zero out the KC table again in a few passes of the load/rpm area but it's not happening here.
Zumble is offline   Reply With Quote
Old 08-18-2007, 09:35 AM   #8
Tea cups
Scooby Specialist
 
Member#: 103136
Join Date: Dec 2005
Default

I'm not sure what Cobb calls what, but I'm pretty sure you can't log feedback correction with ST. You can infer it from by looking at your dynamic advance map, logged knock correction and fine learning correction (what you called knock correction table). However, I think it is much easier to just log feedback correction separately. Think of feedback correction as a knock light when the ecu perceives knock. Think of fine correction as more blunt correction over specific load/rpm ranges due to perceived knock in the past under more specific circumstances. That is, a number of requirements have to met in order for the fine learning correction table to be ADJUSTED (notice I said adjusted but not APPLIED). Whereas feedback correction, as long as you are within its designated load/rpm ranges, it will correct for perceived knock. It is more useful than logging the 'knock signal #1' bit as there is a slight delay where feedback correction remains negative even after the knock signal is clear, so you don't have to worry about "catching" it when logging.
Tea cups is offline   Reply With Quote
Old 08-19-2007, 08:11 AM   #9
Zumble
Scooby Specialist
 
Member#: 34860
Join Date: Apr 2003
Chapter/Region: Tri-State
Location: NY
Vehicle:
2005 Forester CGM

Default

I've never used enginuity to log. I'm going to have to try that. I just found out that I can use the tactrix cable for street tuner so I've been doing that.

I got 1 step colder plugs put in yesterday and a compression test done. 150psi through all 4 cylinders so at least my engine doesn't seem blown. The mechanic said my plugs were pretty fouled and asked if I used any additives. I had used torco 3-4 times in the car. I was told that the torco doesn't stay mixed with the gas very well and separates and fouls the plugs. I actually have some torco in the tank now so I'm trying to flush it out and dilute it by topping off the tank often.

I did see a instance of "feedback correction". While driving home I saw the car pull 3 degrees of timing under low load so I let off the gas. I saw this maybe 2 more times and nothing showed up in the knock learning (learning correction) table. Currently the knock learning table is clean. I'm not sure what kept it clean, new plugs or the torco.
Zumble is offline   Reply With Quote
Old 08-19-2007, 11:03 AM   #10
Tea cups
Scooby Specialist
 
Member#: 103136
Join Date: Dec 2005
Default

I wouldn't worry about it. You are always going to have some false knock regardless even on a completely stock setup. From what I've seen, this is most likely to occur when lugging the engine or just spordically under light throttle off cruise. What you really need to worry about is when you are under moderate to high loads.
Tea cups is offline   Reply With Quote
Old 08-20-2007, 07:16 AM   #11
Zumble
Scooby Specialist
 
Member#: 34860
Join Date: Apr 2003
Chapter/Region: Tri-State
Location: NY
Vehicle:
2005 Forester CGM

Default

My tune seems pretty clean under moderate/high loads. Thanks for the help. I'm going to keep and eye on it over the next few days.
Zumble is offline   Reply With Quote
Old 08-20-2007, 07:46 PM   #12
PerformanceAutoSolutions
Former Vendor
 
Member#: 853
Join Date: Feb 2000
Chapter/Region: Tri-State
Location: Deer Park, NY, USA
Vehicle:
2011 STI Sedan
Satin White Pearl

Default

whats your afr--with using a sepeate afr sensor-not the oem one?also whats your a/f correction when this happens
PerformanceAutoSolutions is offline   Reply With Quote
Old 08-21-2007, 11:06 AM   #13
Freon
Scooby Specialist
 
Member#: 88322
Join Date: Jun 2005
Chapter/Region: MWSOC
Location: Indianapolis, IN
Vehicle:
2009 BMW 135i

Default

You will log some negative KC around the cruise range, ~ 0.2-0.8 load, 2000-3000rpm, if you lug the engine.

I've logged as much as -8 while also listening to the knock sensor audio over my stereo and there is no audible knock.

I've for certain found what is logged as KC by the ECU with my disassembly work. I haven't revisited trying to find out what causes this negative KC to display even when there is seemingly no knock. Before I thought there was a compensation map that got rolled into it. I'll look at it again.
Freon is offline   Reply With Quote
Old 08-21-2007, 11:28 AM   #14
Tea cups
Scooby Specialist
 
Member#: 103136
Join Date: Dec 2005
Default

Quote:
Originally Posted by Freon View Post
I've for certain found what is logged as KC by the ECU with my disassembly work. I haven't revisited trying to find out what causes this negative KC to display even when there is seemingly no knock. Before I thought there was a compensation map that got rolled into it. I'll look at it again.
KC is feedback correction + fine learning correction + dynamic timing advance with IAM applied. Negative KC could result at low loads and assuming no noise that the factory knock sensor would determine as knock, if current learned fine correction is negative, then you would have negative KC with no knock (since dynamic timing advance is typically zero at low loads). That is why it is important to log each element of KC separately, which you can do with the latest Enginuity logger definitions. There's no previously undiscovered compensation table for KC.
Tea cups is offline   Reply With Quote
Old 08-21-2007, 11:33 AM   #15
Zumble
Scooby Specialist
 
Member#: 34860
Join Date: Apr 2003
Chapter/Region: Tri-State
Location: NY
Vehicle:
2005 Forester CGM

Default

I have a scanmaster 3 connected to a LC-1 so I see AFRs right next to the knock correction value. AFRs are in the 14s when this happens. Correction, not sure but its probably around +/-5 with a learned trim of -5.

Quote:
Originally Posted by M.S.P.T View Post
whats your afr--with using a sepeate afr sensor-not the oem one?also whats your a/f correction when this happens

Last edited by Zumble; 08-21-2007 at 11:41 AM.
Zumble is offline   Reply With Quote
Old 08-21-2007, 11:40 AM   #16
Zumble
Scooby Specialist
 
Member#: 34860
Join Date: Apr 2003
Chapter/Region: Tri-State
Location: NY
Vehicle:
2005 Forester CGM

Default

Today I was seeing -1 and -2 and nothing in the knock correction (fine learned correction?) table. I'm still working on flushing the torco out of the tank though.

I've definitely seen examples of seeing negative KC because of a negative value in the learned fine correction table.

A simpler example would be when I had -2 in one of the cells in the mid load level. I pulled 2 degrees of timing and richened that area a bit. I'd do a pull past that area and the KC would still be 7 out of 9. Do a pull again, it will be 8, again and it would be 8.5 then finally it would go all the way up to 9 and the value in the fine correction table will be 0. So yes, even if it's not knocking the fine correction table will remove timing until it is satisfied that area is clear.

Quote:
Originally Posted by Tea cups View Post
KC is feedback correction + fine learning correction + dynamic timing advance with IAM applied. Negative KC could result at low loads and assuming no noise that the factory knock sensor would determine as knock, if current learned fine correction is negative, then you would have negative KC with no knock (since dynamic timing advance is typically zero at low loads). That is why it is important to log each element of KC separately, which you can do with the latest Enginuity logger definitions. There's no previously undiscovered compensation table for KC.
Zumble is offline   Reply With Quote
Old 08-21-2007, 11:55 AM   #17
Tea cups
Scooby Specialist
 
Member#: 103136
Join Date: Dec 2005
Default

Quote:
Originally Posted by Zumble View Post
A simpler example would be when I had -2 in one of the cells in the mid load level. I pulled 2 degrees of timing and richened that area a bit. I'd do a pull past that area and the KC would still be 7 out of 9. Do a pull again, it will be 8, again and it would be 8.5 then finally it would go all the way up to 9 and the value in the fine correction table will be 0. So yes, even if it's not knocking the fine correction table will remove timing until it is satisfied that area is clear.
There are very specific conditions that must be met before an adjustment to the fine correction table and a delay (over which the knock signal must remain clear) before a positive adjustment is made.

The problem is that, unless you are listening like Freon with det cans you can't tell what was false knock and what was real knock. Further compounding the issue is that the fine learning correction table could consist of corrections based on false knock and real knock (plus the fact the rpm/load ranges are not very narrow). That is why it is more useful to look at feedback correction (along with fine learning) rather than just fine learning table by itself.
Tea cups is offline   Reply With Quote
Old 08-28-2007, 08:49 PM   #18
PerformanceAutoSolutions
Former Vendor
 
Member#: 853
Join Date: Feb 2000
Chapter/Region: Tri-State
Location: Deer Park, NY, USA
Vehicle:
2011 STI Sedan
Satin White Pearl

Default

how did you make out--you have a log? of the issue
PerformanceAutoSolutions is offline   Reply With Quote
Old 08-28-2007, 08:50 PM   #19
Zumble
Scooby Specialist
 
Member#: 34860
Join Date: Apr 2003
Chapter/Region: Tri-State
Location: NY
Vehicle:
2005 Forester CGM

Default

It seems like the issue was caused by bad plugs. Installed some new plugs and everything has been in the clear since!
Zumble is offline   Reply With Quote
Old 08-30-2007, 06:59 AM   #20
cucamelsmd15
Scooby Specialist
 
Member#: 122980
Join Date: Aug 2006
Chapter/Region: South East
Location: Raleigh, NC
Vehicle:
2010 F150, 2013 CX-5
92 Miata 15 TTF/PTF/E3

Default

Quote:
Originally Posted by Freon View Post
You will log some negative KC around the cruise range, ~ 0.2-0.8 load, 2000-3000rpm, if you lug the engine.

I've logged as much as -8 while also listening to the knock sensor audio over my stereo and there is no audible knock.

I've for certain found what is logged as KC by the ECU with my disassembly work. I haven't revisited trying to find out what causes this negative KC to display even when there is seemingly no knock. Before I thought there was a compensation map that got rolled into it. I'll look at it again.
I see this alot at a 70mph cruise on the highway, usually a random 2-3 degrees pulled, and nothing from the knocklite.

Ive also had the ECU pull 8.1 degrees once, as I happened to be looking at the laptop, but I had also just hit an expansion joint on a bridge. I did see a bit of knock on the knocklite, but I was also at less than 10% throttle, and extremely low load since I was going downhill.

My full throttle tune is knock-free, and like the OP, no matter what timing I tried, I always had the random "knock", so I wrote it off as just noise.
cucamelsmd15 is offline   Reply With Quote
Old 09-01-2007, 07:51 AM   #21
Zumble
Scooby Specialist
 
Member#: 34860
Join Date: Apr 2003
Chapter/Region: Tri-State
Location: NY
Vehicle:
2005 Forester CGM

Default

I spoke too soon. After all the torco had been flushed out of the tank the negative fine correction values at low load are back. I thought I was safe because I had topped off the tank a few times to dilute out the torco. Since this is happening, does that mean it's real knock?

Same thing here as the poster above. 2-3 degrees at cruise, nothing from the knocklite. My knocklite is spliced in to the stock knock sensor.
Zumble is offline   Reply With Quote
Old 09-09-2007, 12:58 PM   #22
Beanboy
Scooby Specialist
 
Member#: 4814
Join Date: Mar 2001
Chapter/Region: NESIC
Location: Beantown
Vehicle:
2005 LGT Wagon 5MT
Garnet Red Pearl

Default

Any follow-ups?

Getting KC values of 0, but ignition timing is being yanked hard (18.5 to -6.5 then back to 19) for one recording instance (about 120ms between them) at low load (0.25-0.33) and RPM (800-2000) values.

I kinda-sorta understand why KC values can go negative at light loads, not sure about actual ignition timing, especially when KC is rock solid at 0.
Beanboy is offline   Reply With Quote
Old 09-09-2007, 01:54 PM   #23
Zumble
Scooby Specialist
 
Member#: 34860
Join Date: Apr 2003
Chapter/Region: Tri-State
Location: NY
Vehicle:
2005 Forester CGM

Default

I just raised the fine correction threshold to .85 load on the latest basemap reflash. I'm going to see if that helps anything. Last night while driving home I was seeing the ecu pull 4 degrees of timing at very light load at highway cruise and it was actually causing the car to run slightly rich! I was getting 13.x afrs while the timing was being pulled.
Zumble is offline   Reply With Quote
Old 09-09-2007, 07:51 PM   #24
HndaTch627
Scooby Guru
 
Member#: 6551
Join Date: May 2001
Chapter/Region: MWSOC
Location: Carol Stream, IL
Vehicle:
'01 GC8 Dinged STM
'09 Concours 14 ABS Black

Default

my fix: more timing, what freon said was right. It's bascially the engine lugging, adding 2-3 degrees seemed to really take care of it.

Then again i am running E-85 now and not premium
HndaTch627 is offline   Reply With Quote
Old 09-10-2007, 11:52 AM   #25
PerformanceAutoSolutions
Former Vendor
 
Member#: 853
Join Date: Feb 2000
Chapter/Region: Tri-State
Location: Deer Park, NY, USA
Vehicle:
2011 STI Sedan
Satin White Pearl

Default

whats your air fuel correction when this happens.
PerformanceAutoSolutions 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

Similar Threads
Thread Thread Starter Forum Replies Last Post
Idle timing/ low load ignition timing adjustment after cam install? NITROS Engine Management & Tuning 2 03-18-2009 02:32 PM
Timing pulled at low loads & throttle tip in ronzogonzo Open Source Reflashes 9 06-10-2008 05:54 PM
Pulling timing at low RPMs? beavis820 Engine Management & Tuning 8 09-10-2007 09:44 AM
UTEC pulling timing but low knock counts TurboRex UTEC 15 08-21-2006 01:25 PM
Timing at low load/high RPM sites Kha0S Engine Management & Tuning 6 01-08-2004 08:34 PM


All times are GMT -4. The time now is 12:35 AM.


Powered by vBulletin® Version 3.7.0
Copyright ©2000 - 2014, Jelsoft Enterprises Ltd.
Powered by Searchlight © 2014 Axivo Inc.
Copyright ©1999 - 2014, North American Subaru Impreza Owners Club, Inc.