Do you know for sure that it is knock would be my first question...I think some tuners actually have knock detectors (basically stethoscopes) and will be able to properly listen for it. Keeping in mind that knock sensors are nothing more than vibration sensors (piezoelectric IIRC), with some filtering applied to the signal (i.e. digital signal processed), proper knock detection/alarming is mostly dependent on Mitsu's ability to write the filters properly. Basically, false knock. LOL.

Having said that, is the knock consistent? You said yes and no...I'd say, post your log to let the gurus on the forum have a look.

verkion

 

The existence of knock is consistent. The counts are not.

And yes, what you mention is exactly what I'm looking into now. We're also going to check if the knock (or vibration) that is being detected is limited to one cylinder or not. We'll also see if hitting the block with a wrench causes the sensor to detect knock

 

You can do this while stationary. If I free rev my car right, I can get a single count or so. Funny stuff.

- Lou

 

Has anyone solved this problem I too have the same issue with knock in the 6500-7000 rpm range it's strange some times it's zero then other times it's 3 or 6 one log showed 9 I have reduced timing no change seems to come and go at random. I think it may have to do with heat my first run is always just 1-2 counts after that I get 3-9 counts.

 

FWIW, I've noticed more knock with certain types of gas vs others too. Lots of things can cause it. Needless to say, I'm a pretty firm believer in buying good gas vs just buying cheapy stuff.

 

I have tried different gas stations no change wonder if it's even real knock

 

How are you tuning your car?

 

Sounds way rich if your wideband is accurate.

 

Are you guys logging misfire incidents and calling that "knock"? Could it be a lean tip-in or lean open loop transition that's causing the misfire events?