I got my ECUflash and evoscan datalogger but want to verify all my maps before i mess with anything.
I am seeing a few random knocks througout the RPMs but not big amounts, maybe a max of 2 but it is not consistent.
Is this normal, what could cause this???

TIA

Disclaimer** I do not have a wideband yet so i am not trying to do any adjustments yet just looking and trying to understand the features.

 

yes that is normal, I see it in the 3500-4500 range from time to time.

 

thats when I see it too. I am wondering if an open element air filter would help cause that.

 

I get pretty frequent knock at part throttle, counts below 7 are pretty normal, its when you have knock that occurs at higher counts, or for sustained periods that you have to be concerned about..

 

Here is my log in 3rd gear. Not sure how to understand the TimingAdv yet but I think it amount advanced over the MAP in the ROM..

RPM KnockSum TimingAdv
3031.25 0 31
3062.5 0 29
3062.5 0 28
3093.75 0 28
3093.75 0 28
3125 0 28
3125 0 27
3156.25 0 24
3218.75 2 19
3187.5 2 17
3250 1 15
3281.25 1 12
3312.5 1 10
3343.75 1 8
3375 1 8
3437.5 1 7
3468.75 1 5
3531.25 1 4
3562.5 1 2
3625 0 0
3656.25 0 -2
3687.5 0 -1
3843.75 0 2
3812.5 0 4
3968.75 0 3
3968.75 0 3
4031.25 0 3
4125 0 3
4156.25 0 4
4250 0 4
4312.5 0 3
4375 0 2
4437.5 0 3
4500 0 4
4562.5 0 4
4625 0 4
4687.5 0 4
4750 0 4
4812.5 0 4
4875 0 5
4937.5 0 5
5000 1 4
5031.25 1 4
5125 1 5
5156.25 1 5
5218.75 1 4
5281.25 0 4
5343.75 0 4
5406.25 0 5
5468.75 0 5
5500 0 5
5562.5 0 6
5625 0 7
5687.5 0 8
5718.75 0 8
5781.25 0 8
5843.75 0 8
5875 0 9
5937.5 0 10
5968.75 0 10
6031.25 0 10
6093.75 0 10
6156.25 0 12
6187.5 0 12
6250 0 12
6281.25 0 12
6343.75 0 13
6406.25 0 14
6468.75 0 14
6500 0 14
6531.25 0 14
6593.75 0 16
6625 0 16
6687.5 0 15
6718.75 0 16
6781.25 0 16
6812.5 0 17
6875 0 17
6906.25 0 18
6968.75 0 18
7000 0 18
6812.5 0 38
6781.25 0 38
6750 0 38
6593.75 0 38

 

Jack how did you come to the conclusion that 7 counts or lower are OK?

I'm curious because I can get timing pulled with as little as one count of knock. I've only seen 7 counts a couple of times and the amount of timing being pulled in relation was crazy.

 

Because it takes sustained knock counts (of any number) before the octane number changes, I log individual counts as high as 5 on occasion with no change in timing.. I have several logs.. I'm talking about part throttle low-medium load here too.. WOT numbers should be kept as low as possibleobviously, but low counts are inevitable..

If you look at the knock filters (discretion maps) it appears that the higher the load, and RPM, the more it reacts to knock..

If you get a chance, take all of those knock filter values and plot them against eachother and 3d, what you'll see is an audio spectrum with peaks in areas where we typically see knock..

 

If you want to see the 3D graphs, look here: https://www.evolutionm.net/forums/at...chmentid=93700

I started a thread in the advanced section on knock control a while back: https://www.evolutionm.net/forums/sh...d.php?t=212476

It's very interesting to see what did with their P0300 flash ... they adjusted the low load/low rpm area of knock filter 5... I haven't had any time to really delve into it beyond that yet, though.

l8r)

 

Did Mitsu make that number higher or lower?

 

Interesting, I wounder if the ecu was detecting knock and pulling timing somehow was causing the missfire code.

 

[QUOTE=MalibuJack]Because it takes sustained knock counts (of any number) before the octane number changes, [QUOTE]



The knock sensor output will also cause real time timing pull porportional to knock sum value. Knock sum value is then loaded into a register that decays over time and this is the octane value. So their are two styles of knock control, short and long term.

 

[QUOTE=dan l][QUOTE=MalibuJack]Because it takes sustained knock counts (of any number) before the octane number changes, That knock sum decay stays too long in the register. someone gotta find the subroutine so that we can control that decay. It is very problematic when you encounter the knock for very brief period of time like fast shifting and there is really no need to pull the timing for entire run ( good for the precaution but bad for the performance )

 

The decay is part of the short term knock protection. I do not suggest messing with this routine, it is a great tool to protect a street car. Instead I've had great sucess trying to tune out this knock. Another option (if your into changing routines) is making the ignition timing go to 10 or 0 when the clutch is in. This provides more boost potential and in my first gen dsm, has eliminated shift knock during flat shifts.

 

Its also good to mention that when the engine "Free revs" while clutching, the load goes way down, at least 3 columns or more...

Also look for things that can cause plinks and clanks when shifting hard, get an engine stabilizer and make sure the downpipe isn't hitting the crossbar, thats the source for most of the knock you'd see when shifting hard.

 

Most knock while shifting you would see are spikes that disappear quickly, in my experience it may affect the short term timing ever so slightly, but it returns to normal very quickly. It doesn't affect the long term value (octane number) at all. When analyzing my logs on my specific car, it had no affect on the actual timing, at least not for a long enough duration for me to be able to detect it.