*** Knock, do you have any?

[fishbone]

I wonder if that gigantic correction has been the issue in the past for me which would make the IAM drop from 1. It's been at 1 this whole time, no learned knock in that area that I can tell. The thing is, I have never seen it go as high as it did yesterday but occasionally I do see -6 or -8, very infrequently. I also see a -4 sometimes when I slow down going into a turn and touch the throttle. After completing the turn I'll look down and the AP would show -4. I'll also on rare occasions (emphasis on rare) see a -2 here and there when I am merely slowing down at a light from 40mph or so. I'll see about reducing the timing in that area. I also had a situation two days ago at a stop sign, completely stopped, turning right, touched the pedal and the car shuddered as if it was just about to stall due to misfires, again -4 timing pulled. Weird. I've watched injectors and for misfires in the past, no issues there.

Otherwise, I see no knock activity usually when I WOT and sweep 1-2-3 to redline.

 

Methinks since the load is not particularly high it shouldn't be destructive but I'd still like to address it if possible. I'll go hunting for rattles just in case.

I have a 15PSI actuator for the VF40 and a Perrin 3-port EBCS that I am dying to install, but that's another topic altogether Mark (qikslvr) has been tuning me and really, it took him just one adjustment to get my boost under control. It seems that the car wants to be right around 17.1-17.4 PSI. On the OTS Stg2 Cobb map it would overboost to 17.4, so Mark said OK, we'll make the target boost that and sure enough, she's happy with that value

Edited June 16, 2010 by fishbone

[NSFW]

Far as I know, the ECU drops IAM when it sees lots of learned knock correction. It uses feedback correction for knock in conditions that probably aren't important, and that's where you're seeing the -9 in that log.

 

Again, have you tried turning down timing to see if that helps?

[fishbone]

Haven't tried it yet because I am still on a tank of 93 that I have to run dry.

[Infamous1]

Far as I know, the ECU drops IAM when it sees lots of learned knock correction. It uses feedback correction for knock in conditions that probably aren't important, and that's where you're seeing the -9 in that log.

 

Again, have you tried turning down timing to see if that helps?

Rough correction can be applied at any time. I have seen IAM change a number of times without previous FLKC or FBKC.

[LittleBlueGT]

Rough correction can be applied at any time. I have seen IAM change a number of times without previous FLKC or FBKC.

 

IME that only happens shortly after an ECU reset (which happens a lot when not RT tuning).

 

The ECU is then out of rough correction mode (assuming it wasn't knocking lots just prior) and into fine correction mode.

[Infamous1]

Could be not too sure there, Merchgod explained it on this forum some time ago.

[NSFW]

My experience is the same as LBGT's.

 

Here's more than you ever wanted to know about the knock control system:

http://www.romraider.com/forum/viewtopic.php?f=25&t=1840

[dmanaenk]

-9. Omg wtf bbqthxbye

Log 14 is some of that "cruise knock" I mentioned. 5th, tq locked up, squeeze the pedal to accelerate from 50-ish to 70-ish.

I just saw a -9 (okay, -8.44 but KC jumped by 3 so I'm pretty sure there was a -9 there) on a completely stock map with no mods. So it happens

[fishbone]

^Yeah, I had an occurence of -6 again and engine load was .4. I also had a -4 right after I wound out 2nd gear, let off and the moment it upshifted it "knocked" with my foot completely off the throttle. IAM is 1, no knock activity under WOT and high load, I am not going to lose sleep over it but will try to get it tuned out if possible.

[fishbone]

I have discovered that cyl 1 ignition coil was not properly screwed down, so it was dangling on the block and it was on just enough to make proper contact with the spark plug. I torqued it down and for the time being I have not seen any more big knock corrections. At most I have seen a -2.8 FLKC at very low load.

[rapture]

just installed a walbro and now i have a knock but it seems to only be when im idling ...datalogged the short and long fuel trims and my #'s were 20 short and 14.8 long can i damage my car driving for a little bit like this?

[McDowell Performance Tunin]

Just changing the fuel pump shouldn't cause knock, as it will pump out gas at the rate the ecu dictates, assuming nothing changed on your map. Try reseting your ecu. If your fuel filter was dirty, it could cause some issues. What was the reason for upgrading your pump?

[rapture]

well i just figured i should put it in now instead of waiting till i get the turbo upgraded. my cars at 55,000 so i figured it wouldn't hurt anything i have a bunch of other mods to the air so i figured mod the fuel side of the engine too. now the guy i took it to for tuning said its the fuel pump and to invest in the avo with the housing.he said he has seen alot of problems with the walbro's. i did notice when we modified the housing how much bigger the new fuel filter they give you is and how it gets pinched at the bottom.. we followed the directions and i really don't see what mistakes we made other than keeping the old o ring . but from what i understand the avo is also able to adjust pressure levels to help with idle and times when ur not trying to go full bore.

[fishbone]

It's been a while ...

I think I'm seeing the result of transient temps vs winter gas?

IAM went down to .87, I did some logging and under WOT I see no FBKC or FLKC. But that may be because IAM is down ...

However, at low load I see a fair amount of FLKC activity. I haven't pulled a LV yet, I will keep an eye on things.

Two questions:

1) does low-load knock cause the IAM to drop?

2) which upper threshold for load is considered "safe" to disable knock learning? Is 1.4 OK?

I still have not figure out exactly what event causes FBKC sometimes to pull a -6 degrees suddenly and out of the blue in city driving with barely any throttle.

 

I swear this tune has been running top notch for 2 years now. Must be a bad tank of gas or like I said, warm temps, wrong gas?

[LittleBlueGT]

It's been a while ...

I think I'm seeing the result of transient temps vs winter gas?

IAM went down to .87, I did some logging and under WOT I see no FBKC or FLKC. But that may be because IAM is down ...

However, at low load I see a fair amount of FLKC activity. I haven't pulled a LV yet, I will keep an eye on things.

Two questions:

1) does low-load knock cause the IAM to drop?

2) which upper threshold for load is considered "safe" to disable knock learning? Is 1.4 OK?

I still have not figure out exactly what event causes FBKC sometimes to pull a -6 degrees suddenly and out of the blue in city driving with barely any throttle.

 

I swear this tune has been running top notch for 2 years now. Must be a bad tank of gas or like I said, warm temps, wrong gas?

 

 

1) not really, I usually change the rough correction range so that it only changes IAM with knock above about 1.8 load or so. Heck, my DA table doesn't have much below 1.5 or so load.

 

2)you may have written that wrong, but, the LOWER limit of about 1.4 load is fine IMO. I don't have much problems right now, but when it is cold, there are certain low load ranges that knock a lot, and it doesn't matter what I do with fueling, and what I do with timing, so essentially disable it.

[Spec B]

^ I disagree with #1. Flkc is the only type of knock that lowers IAM. Flkc is also the type of corrections that deals with low rpm, low load cruise knock. So if your getting a ton of low load knock that falls within the flkc correction ranges, it will drop your IAM pretty quickly.

 

EDIT: this is all assuming your flkc ranges are within the same low load range in which you knock. I use almost the same flkc ranges as lbgt, I have it set to ignore anything below 1.0.

Edited April 13, 2011 by Spec B

[LittleBlueGT]

^ I disagree with #1. Flkc is the only type of knock that lowers IAM. Flkc is also the type of corrections that deals with low rpm, low load cruise knock. So if your getting a ton of low load knock that falls within the flkc correction ranges, it will drop your IAM pretty quickly.

 

EDIT: this is all assuming your flkc ranges are within the same low load range in which you knock. I use almost the same flkc ranges as lbgt, I have it set to ignore anything below 1.0.

 

You are correct that FLKC is the only form that causes IAM to fall (once ECU is out of rough correction mode after a reset).

 

There are some min values in the ROM, and I can't say I completely understated them, but I thought that your DA table had to call for a min of 4.0 degrees or else the IAM would not change. You will see the table, read the description and see what you think.

 

I know that I had FLKC values of -11.75 at low load before in the winter months. It was my clutch, but IAM never fell.

 

To be honest though, I don't remember if my coarse low load limit was already raised.

 

At any rate, I would raise it to 1.6 load or so if it were me.

[fishbone]

I suspected as much. I get constant knock right around 1-1.2 load if I "pinch" the throttle while cruising. So as high as 1.8 would be safe? By high I mean 0 load being low and going up from there so I guess I am looking at it in reverse I will try 1.4 and monitor. Last LV was clean except for one cell in the first column which had a -1.4. I forget exactly where but it was at low load.

I very rarely take it above 4k, it's 80% grandma driving in city so I am not surprised IAM dropped.

I have not seen fbkc nor flkc at wot ever in any gear, except 3rd and letting off abruptly at redline which is nothing abnormal. I try to avoid it but sometimes forget to upshift before letting off Kinda rich though, 11.0.

Edited April 13, 2011 by fishbone

[fishbone]

Also call me crazy but I've got some solid reasons to think the auto tranny or rather the presence of the stiff mount insert is inducing some of that low load knock.

[LittleBlueGT]

Why not build some det cans and listen for knock, then you will know for sure.

 

Yes, I think it is safe.

[fishbone]

I'd have to look into it since I know nothing about it. Might give it a try.

[spec.B Dream]

For anyone reading this not as familiar, it is worth it to read (and periodically re-read) Bill's explanation of Subaru's knock control strategy on RomRaider. http://www.romraider.com/forum/topic1840.html

[fishbone]

Thanks for the link, I'll look it over again.

I looked at some DetCan DIYs and looks like a fun project

I think I want to go the electronic route rather than the tube route unless it doesn't work as well, which I doubt.

[sumfoo1]

I have knock and an ecutek tune that i can't friggin change...

 

Next time... i'm bringing romraider with me when i buy a subaru.... lol

[Legend]

I have knock and an ecutek tune that i can't friggin change...

 

As one who previously had an ecutek tune, I feel your pain.

 

"Ecutek - The Flashed and The Furious"