Interesting. On the one hand it would explain why some cars dont really ever seem to knock and others are phantom knockers from H. Have to do some testing and see.

What I have learned from code is in a Mitsu it never does exactly what it says.

Aaron

 

Because there is always another algorithm that acts besides the one you are trying to adjust or bypass. Something is always compensating.

 

Hey, I am not compensating! A 600whp car is just fun, not compensating!

 

This doesn't make much sense. My low octane and high octane maps are only a few degrees different in areas I'm currently experiencing 5-8 counts of phantom knock.

Data logging today to see if re-torquing and properly aligning the knock sensor made any difference.

 

Make the low octane timing map the same as the high octane timing map, then do some logs.

 

I don't want to take this thread off topic, but I will say I've tried this a couple times now I either have some extra engine noise with this new built block, the exhaust leak I have is registering as knock, or the knock sensor I rigged to work from a similar car (same NTK part #) really isn't working that great Though it was initially, so

 

Please start a different topic. Rigging a knock sensor isn't the greatest idea without some real insider knowledge of how to tune those tables... even torqueing the factory sensor wrong can lead to false readings.

 

I was just replying to the guy as to not be rude. No need to try to undermine my credibility. I've been in this game for 5+ years now, and I'm most certainly aware of what you mentioned in your post.



Anyway, directly related to Evo X Knock Algorithms and Maps, has anybody found the "Knock Load" map for the Evo X - this pretty much was the silver bullet for curing tip-in knock on the earlier Evos. I know there is a map that seems to correlate with load thresholds for affecting the learned octane level, but that's not what I'm looking for.

 

ok so I did some more code debuggeration and have found that:

1) when the above the knock load threshold table (which you are most of the time) AND when your high/low timing map delta * 3 > KNOCK_SUM (which again is most of the time) you are basically asking for increased sensitivity to knock

so by setting the high/low octane maps the same you are in effect desensitising the knock sensor

2) still working on #2

 

thank you for keeping up on this tephra

 

Yeah I found this out a while ago through some map testing. Keeping the high/low maps the same and adjusting the multipliers got rid of much of the phantom knock.

 

Interesting... man some people have some scary knock sums. I made a few pulls today... saw nothing more than a 1 count.

 

I had the same prob with my evo x knocking here and there and sometimes would get 3-8 counts at 6500 to 7000 rpm but it would come and go my first couple logs would be 0 knock. This was on a Cobb ots map so I set the low octane timing map the same as the high and all that knock went away I only get 1 count sometimes also the car runs better pulls harder and timing logs match the map. Only thing is now I get the rear o2 code that I never got before. Btw this was how my evo 8 was setup

 

I forgot to mention the ecu was pulling timing with no knock present even at low rpm so strange its like wtf is going on there obviously must be more undefined tables in the ecu

 

I think you are doing it wrong.... and possibly going to blow up your car. Maybe you should contact COBB about the maps being incorrect or check to see if you are using the correct OTS map...