It seems to me that MUT_72 logs if the Knock is actually present, this is bit0.
The second thing what it logs is if it is in fade out phase that is bit1.

0 - no knock
1 - knock
2 - no knock but decay time - ignition still affected from previously detected knock
3 - knock + decay time timer running

Can anyone confirm?

I will try to log it, hopefully I will get some knock

 

Neat info. We normally refer to the fade out time as decay time, just so we are all on the same page.


Eric

 

So with that said, can the decay 'time' be edited? I think we've all seen the 'decay' in full effect while it peaks at the beginning of a pull and tapers down throughout the ENTIRE pull, basically ruining any hope of power for the entire run.

 

One step closer to "Knock Control"

 

Yes, even knock sum count per one ignition degree (3 in stock) can be edited.
I try to locate the same in some of well known ROMs like 88590015 and get back with locations.

 

Finally! I have been saying 3 knock counts per 1* of timing for years and some people never believed me. It's good that you found that.

 

You said it backwards...maybe thats why no one listened,lol. JK. Its so cool that more and more disam guys are joining the quest to hack the ECU code. I'm dying to understand it completely but I really need my hand held the first time I'd say within a month we are going to have 100% control over the knock control stuff, with Tephra working on a 3D knock map and with the level of ignition retard now user adjustable people will be able to keep their cars in their 'tuned groove' a lot more often

 

On 96530006 there is a big subroutine for knock starting at adr 26360 - however I got lost....
Howoever - MUT72 logs FFFF8A0B; but in the dissassembly - some knock dynamics goes into FFFF8A0A - so we need to log 8A0A as well as another MUT. Am I right?

 

Great stuff!

 

My IX has pulled timing from one count of knock numerous times and at 3+ counts my pull is shot, car falls flat on its face.......my previous VIII was much more sensible when it came to knock vs. retard.

 

I've noticed this too, however maybe it's a logging issue, could be really knocking 3 counts but too fast for the logger to pick up.... maybe.

 

yeah but that is a pretty big maybe.......im inclined to think mitsu made the knock control a bit pickier in the evo IX for one reason or another. Nice thing - its all about to become a non-issue...

 

Has anyone found any evidence that the ECU uses finer increments than 1 degree? Just curious if it interpolates enough to maybe be using 0.1 degree or some other increment? This would explain those who see timing move for 1 knock.

 

I recall someone saying the cam or crank sensor can only read in one degree increments compared to a subie which can do 0.5* increments. I cant confirm any of that, totally from other peoples input on here over time.

 

the decay timer was in that fed up with knock control thread...