Knock sensor min & initial changes , now cell ?

[Maxspeed96ct]

I've determined the knock I get during a hard launch is false after logging several runs of roll runs and digs
I adjusted my knock sensor sensitivity . Ls2 e40 Ecu

My initial value was 400 so I only bumped it to 410
My minimum was 50 so I bumped it to 100 like others do .

After a few hours of driving I got a dtc code for knock sensor immature or something along those lines ( I forget the exact wording )

Did I increase the minimum to much ? Is there another value I need to change ?

Any help would be appreciated .

[mowton]

I am not sure why you coded with that change, but I haven't changed those values for your stated issue, or any other issue that I can remember. I normally use the multipliers and then only after determining by reducing timing in the affected area, determining it is not Burst Knock (PCM's predictive action to possibly oncoming knock) and verifying it is truly false. Sounds like what you have done already, but not sure to what extent. As yours is via transient (tip-in), then I would start by increasing the transient multipliers by say 10-15% in the applicable rpm cells. Cylinder 1-8 multipliers will adjust the steady state sensitivity.

What is you cars setup....

Ed M

[Maxspeed96ct]

Awd trailblazer ss - Headers , mild cam with the other related bolt ons and a fast 102

So I dont acutualy have the touch cylinders 1-8 tables ?

[mowton]

Awd trailblazer ss - Headers , mild cam with the other related bolt ons and a fast 102

So I dont acutualy have the touch cylinders 1-8 tables ?

Depends on where you are seeing the knock. You stated it was during launch so I would look at burst and transient first. I am a proponent that GM did a pretty good job of characterizing the Knock values on their engines and then we go and screw it all up with mods :-) Again I try to make the least amount of changes as necessary and preclude false knock. Burst knock I normally will disable or reduce to 2 degrees max. If during a wot pull from 2K to redline, you see knock, then again make sure the timing isn't attributing to it and then you could desensitize the Cylinder 1-8 values in that area.

Ed M

[Maxspeed96ct]

I read the ls2 and ls3 knock sensors are to sensitive . There's a few threads around that guys adjusted the values .

The knock is not exactly off the launch .
When I do a 1/4 mile run I get 2-3 degrees of knock but im comparing the logs and in my 3rd gear highway pulls im hitting the same cells with 0 knock .

[schpenxel]

I read the ls2 and ls3 knock sensors are to sensitive . There's a few threads around that guys adjusted the values .

The knock is not exactly off the launch .
When I do a 1/4 mile run I get 2-3 degrees of knock but im comparing the logs and in my 3rd gear highway pulls im hitting the same cells with 0 knock .

Leave the initial/min values alone. Put them back to stock. I get errors every time I try to touch those

I have had better success with changing the multipliers instead. Log knock for each cylinder and create a histogram that shows knock vs. engine speed/cylinder airmass (same as the knock sensor threshold multiplier table) and figure out at what cylinder airmass / RPM you are seeing knock, and on which cylinders

From there, if you are 100% sure it is false, you can slowly increase the multipliers in that area until it stops detecting false knock. (they are under spark-->knock sensors-->multipliers)

Again, this is assuming you are 100% sure it is false.

It is also assuming that you are sure it isn't burst or transient KR. There is a multiplier table for transient threshold as well that you may want to take a look at.

Also, you should be logging every type of knock retard there is to make sure it's not coming from something else. There is piston slap retard, burst knock retard, total knock retard, and I believe one more I am forgetting.

Lastly, is your tune scaled at all?

[Maxspeed96ct]

I am positive it's not burst , not sure if it transient kr. I have not messed with that or the multipliers yet .

What do you mean by scaled ?

[mowton]

If your injectors and/or airflow are greater than the hard coded limits of the software, then you have to proportionally reduce (scale) them to go below the limits. The overall goal is to maintain the same afr.

Try with our suggestions and see how you do..

Ed M

[schpenxel]

If your injectors and/or airflow are greater than the hard coded limits of the software, then you have to proportionally reduce (scale) them to go below the limits. The overall goal is to mai tajn thw same afr.

Try with our suggestions and see how you do..

Ed M

Beat me to it

But yeah, you didn't give a whole lot of info on your setup, so we had no idea whether you have a supercharger, large injectors, Nitrous, heads, cam.. no clue.

If it IS scaled you have to adjust the multipliers accordingly too, but if you don't know what that means and no one else tuned it, then it shouldn't be anything you have to worry about.

[Maxspeed96ct]

I'm running ls2 vette injectors , 32lb vs the 28lb that come stock in the Tbss . All tables were copied over . Fuel is in check and consistsnt per wideband .
The truck is in speed density , I'm not positive on the scaled part . I'm unfamiliar with it to be completely honest .

Id like to adjust the multipliers . Increasing the numbers decreases the knock sensor sensitivity In the cyl tables and the transient tables correct ?

[mowton]

With that setup I would doubt the tune is scaled. Yes, increasing the multipliers, decrease the sensitivity.

Ed M

[schpenxel]

I'm running ls2 vette injectors , 32lb vs the 28lb that come stock in the Tbss . All tables were copied over . Fuel is in check and consistsnt per wideband .
The truck is in speed density , I'm not positive on the scaled part . I'm unfamiliar with it to be completely honest .

Id like to adjust the multipliers . Increasing the numbers decreases the knock sensor sensitivity In the cyl tables and the transient tables correct ?

You would only need to go the scaling route if you need to use injectors larger than the computer will allow you to put values in for (somewhere around 63 #) or airflow #'s are over what it can understand. So, not an issue in this case.

Yes to higher = less sensitive

Basically the knock sensors are like a microphone. More noise = more output voltage from them. What you are setting here is how much noise (voltage) it takes for the computer to say OK, that was real knock..

The higher the threshold, the more noise you need to get it over the knock threshold

[Maxspeed96ct]

Thanks for all the help guys, I started with a 5% increase in the 0-4500 rpm range in transient multi. And 5% in the .7 to .9 g/cyl on cyl 1 though 8. I'll do some logs as the weather clears , I was getting random 1-2 deg of knock so Ill be a happy camper to see 0.

One other thing that I will be fixing prior to doing some more WOT pulls is my passenger header touches the frame slightly. I plan on putting the hammer to it....
I think this could be giving me false knock readings during shifts but I could be over thinking it

[schpenxel]

Thanks for all the help guys, I started with a 5% increase in the 0-4500 rpm range in transient multi. And 5% in the .7 to .9 g/cyl on cyl 1 though 8. I'll do some logs as the weather clears , I was getting random 1-2 deg of knock so Ill be a happy camper to see 0.

One other thing that I will be fixing prior to doing some more WOT pulls is my passenger header touches the frame slightly. I plan on putting the hammer to it....
I think this could be giving me false knock readings during shifts but I could be over thinking it

Can't be helping and isn't a good thing regardless.. so I'd fix that first

Some things I've read say the knock sensors are supposed to be able to filter out noises that aren't the same frequency real knock, but I have no idea how true that is