I've been trying to figure out why I'm getting some high counts of knock when letting off the throttle from a wot run. It happens inconsistently, maybe once out of every 3-4 pulls but always right after lift off.

Well I started looking into knock vs knock voltage thinking maybe I've just got a noisy motor,
So I plotted my knock voltage on a line graph and it seems the voltage of the knock sensor has no correlation with the knock sum.
IE: During the entire WOT run my knock sum is from 1-2v. After letting off the throttle where the Knocksum goes to 5, the voltage is only at .9v

Any other thoughts on high counts of knock upon throttle lift off?

 

if you plotted it and you dont see a voltage spike where the knock is occuring its a good chance its not real knock, but you can always throw in higher octane fuel to confirm that?

similar to other EMS's though if you see voltage spike its a good sign of detonation detected by the knock sensor if it doesnt look inline with the rest of the knock volt curve.

 

Thats what I was thinking, here is a screenshot of the voltage vs knocksum.



If I look at the datlog, where the knock always happens is right when I let off the throttle and I'm still at like 20psi of boost, the timing advance goes to 30ish degrees.

 

Smooth out the transition at high-rpm from high load to low load. Make the changes no greater than 2 degrees per column. Start from the high load and work your way to the left. You'll end up with lower timing at low load, high rpm than you will at low load, medium rpm, but it should help with lift/shift knock. It did in my case.

 

I was thinking of trying this, Doesn't seem like there should be any problems with running lower timing in the high rpm's/ low load cells as you won't be there except for on decel anyways.

I will try it out thanks.

 

Well, I tried out Raptord's idea and lowered and smoothed the timing in the low load high rpm but to no avail.
I'm beginning to think this is a vibration issue that the knock sensor is picking up when I let off the throttle.

Here is my timing map, does anyone see anything crazy here that I'm overlooking?




Heres a datalog of 1st, 2nd, 3rd pull. You can see the knock jumper after first gear and is not present at all after 2nd and 3rd.

 

I am following this, because I am working on a similar, but not exactly the same issue. From all that I have read, if you have a knock count like that all by itself, or erratic its not really knock and nothing to worry about. If you have that value surrounded by a bunch of similar values, like I do, then something is going on. I have those jumps, but then they slowly fade down and out.

Have you tried running with some 100 oct to see if you can repeat that value? Also, I have found I can induce this on track, and never on a street pull. I'd be curious what happens when you run third gear relatively hard, similar to a pull, but then slowly back down, like you were driving a tight corner on track. When I log that, then knock is pretty active for me. Its decal...ish, and part throttle, but really active.

 

I used to occasionally get throttle release knock as well, but not any more. Not sure what I did to get rid of it. There's definitely nothing wierd about your timing map, but the log that you show does suggest that the increase in knock sum could have been induced by the sustained higher knock voltage during the pull. What is the voltage range in that region before the knock?

 

I see what your saying, it just doesn't make any sense that in 2nd and 3rd gear I've got voltage up to 2.2v's and not seeing any knock.

So should I be concerned about the motor with this kind of knock? I would guess no but,
I'm wondering if i'm getting this due to my noisy input shaft bearing.

 

^This. If you're leaning towards thinking it's real knock get some det cans on it. I bet if you look at the knock filters the at the point of knock the noise is enough to tip it above threshold, the bearing is more audible on the lift than during acceleration and that's what you're picking up.

 

The knock voltage will change in response to real or false knock noise, and I don't see anything that stands out in the knock voltage to have caused the knock sum change except the sustained higher voltage during the pull. Maybe a degree of timing needs to be pulled across the map?

 

So does it seem the knock voltage I'm getting is higher than normal overall?

I guess I will try pulling some more timing but I believe I have tried pulling some timing up top and it didn't help.
Seems like I'm already not running much timing compared to other maps I've seen.

 

Depends on mods and quality of the gasoline. I think I was only able to run 10 deg up top when I had a nearly stock exhaust.

 

Ok, well I guess I should just try throwing in a tank of race gas and see if that clears the knock up or not.

I'd be ok to throw 5 gallons of race gas in a tank of 93 right?

Thanks for all your help BTW, with not only my problem but for everything you've done for the community.