[JohnSub2013sti]

(The PDF attached includes all below information in addition to the results
If your interested in reading its best to open the pdf)

&& Thank you to anyone that does read

Intro:

• I wrote a VBA program to analyses when FKL or FKC is first seen (initialized) and then matched it to the corresponding cell in the timing table
• Below data does not show the FKL and FKC returning to 0. I figured this information would be of little value. To me its more important where the knock started
• The log consists of 3 hrs of data on the highway
• The timing table should be assumed to be cruise as I babied the car for this log
• Another thing to not is that I lowered the FKL load activation limits. There are pictures of this below. This was done because I saw it in a stage 1 tune & id rather my engine not knock at all even at low rpm. I am aware that under low load maximum in cylinder pressures will be lower which significantly decrease the dangers of knock
• One other thing to note is that my engine does have forged internals and cobb seems to think that built engines should have decreased knock detection here. At least in the higher rpms 3200+

Questions:

• Can a parameter be made to log whether the vehicle is in cruise or not? I looked for this in RR but could see anything that would allow me to log which table was in use
• Is this much knock a concern?
• Could I reduce the base timing in cruise to possibly decrease how often the engine knocks? E.g., for the cells shown below that show knock

The greyed cells could become

​

47

42

28

18

Effectively decreasing the maximum in cylinder pressure in these cells (please confirm I should reduce the degrees before BTDC)

 

[mheyman]

You should have "no" knock at "cruise" . . . which most of the time means light loads. But, I don't see enough info to surmise anything . . . and I don't know anything anyway . . .

 

[JohnSub2013sti]

This is 3hrs of data summarized....

Sent from my SM-S901W using Tapatalk

 

[kopele]

I'm assuming this is the number of events?
Usually you'll see the ECU pull timing and then slowly add it back in, so most of those events are not very important.
More important is how much timing the ECU is pulling when it detects knock, which is the initial value.
From my experience anything less than 2.8 degrees is not a concern, more than 4 should be investigated.
Focus on the areas of the initial event and try reducing the timing to see how it affects your engine.
This tool can be helpful and one of the tabs can do what you're doing above: https://www.romraider.com/forum/viewtopic.php?f=32&t=10481

 

[JohnSub2013sti]

Ya this is number of events. (Everything’s timing was pulled I recored it). I didn’t record adding back timing… bc to me it’s more important where knock was initiated

Also your comment about how much knock is a concern is inline with what a tunner I’ve been talking to said

Thank you for the feedback! I’ll check out that sheet when I get home

Also just one comment about knock… if knock is always a concern then Subaru wouldn’t have added activation limits. Knock isn’t recorded below .8 g/rev on stock cars…. Likely bc even with knock the in cylinder pressures won’t get as high

A lot of the time you’ll be under this .8g/rev limit at cruise

Sent from my iPhone using Tapatalk

 

[kopele]

By "add it back in" I meant removing less and less timing, so you'll see -1.4, -1.0, -0.7, -0.3, and back to 0.
My assumption is that your count includes all those events, when you really only care about the initial -1.4.
If all of the 135 events are the initial timing reduction I would be a bit concerned.
As a reference, I get maybe 1 or 2 total initial events per 30min drive.
You're right about the activation limits, at low load knock is not as harmful and Subaru decided not to monitor it.

 

[JohnSub2013sti]

So say a knock was initiated…/ and the timing pulled went like this

(-1.4, -1.4, -1.05, -1.05, -0.7, -0.7, -0.35, -0.35)

I recorded this as 1 initiated fkl event
In my 3 hr drive I got 22 initiated events

Keep in mind though I also lowered the fkl activation to .6g/rev

Sent from my iPhone using Tapatalk

 

[kopele]

Oh I misread your data, sorry about that.
Yes, you're correct.
You can disregard some of those and focus on the ones above 1g/rev.
Depending on what your total timing is you can try reducing by 1 degree to see if it makes a difference.

 

[praha-sti]

You should have "no" knock at "cruise" . . . which most of the time means light loads. But, I don't see enough info to surmise anything . . . and I don't know anything anyway . . .

You can always get knock at low rpm/cruise, especially on decel ... it's not that big of a deal, unless you really have a crap tune or fuel .... on WOT, you don't want much or bye-bye ...

 

[praha-sti]

Hopefully you don't have one of those stupid "I want flames from my exhaust" tunes as well ... fuel blowing up in the exhaust will ping the det sensor as well ...

 

[JohnSub2013sti]

No flames, not to worry. I just want a nice and safe reliable tune.

Also, this knock I'm trying to fix normally wouldn't even be present anyhow... with the stock fklc limits

Sent from my SM-S901W using Tapatalk

 

[praha-sti]

No flames, not to worry. I just want a nice and safe reliable tune.

Also, this knock I'm trying to fix normally wouldn't even be present anyhow... with the stock fklc limits

Sent from my SM-S901W using Tapatalk

Certainly no reason to make them more sensitive everywhere ... you should only adjust them where boost comes on, depending on your turbo ...that transition is usually where the most "false" knock is .... anyway, good luck! Else just turn it off