What is normal knock levels in a rotary? I know ranges in a given piston engine, but not a 13b from a fd. I'm curious because i'm monitoring knock as i tune, and I cant seem to get rid of it. I've added fuel and reduced timing, but still randomly i'll get a few cells showing 65+ knock levels. any ideas?

 

Are you seeing those 65+ readings in boost? I think it varies car to car. I see high 20's in cruise/vac, low teens in boost at low rpm and dropping to low single digits at full boost/high rpm.

 

Yeah I'm seeing knock mostly in boost. IDK why either cause when i'm WOT my AFR drops down into the 10's and stays there throughout the entire power band on decent if not low ignition settings. The whole TPS vs INJ thing is kicking my *** too because its adding too much fuel at WOT. My AFR shouldnt be dropping from high 13's or low 14's all the way to 10's at WOT.

 

I've read that up to 60 is acceptable but am not sure.

Using E85 for fuel 14 is the highest I’ve seen on boost.

 

Hmmm idk then. I've been trying to make it right, but nothing seems to be working.

 

hopefully you aren't retaining the timing maps that Apexi loads into the units by default..

as for the knock, every sensor will read differently. the gel that insulates the sensor usually melts and leaks out and the readings begin to climb even when not picking up any real knock levels.

 

Haha no im not on a base map at all by now. Ill look into replacing the knock sensor though. Maybe that's what's wrong.

 

Some cars have knock levels in the 100s. But this is not knock but NOISE.

Real knock values will be higher when first getting on the power but then drop as drivetrian slack is removed. If knock is low under most boost but high in some places, then it may be knock. Remove 2 degrees timing IGL/IGT for the column (N column) that is occurs. The knock value will reduce if it was real knock.

If timing is reduce but knock level does not change, then it is noise.

 

I wouldn't depend on the knock sensor for tuning. As mentioned above the sensor readings will vary depending on its age and how it was torqued into the rotor housing. By the time you see any real knock the damage has been done, or at the very least the seal has been compromised.

 

Yeah theres plenty of "knock" throughout the entire map so i'm guessing its mostly noise because it doesnt really reduce depending on IGL/IGT/Fuel

 

You need to make a chart using the Datalogit software and look at the shape of the knock sensor signal curve.

 

Using the datalogit, should we be concerned with a single Max data knock point or just the average?

I've just realized that all my averages are all low but I do have MAX values that are much higher.

 

i would use peak readings, not an average. as knock is generally a sudden event that goes as quickly as it comes.