Knock count... What am I doing wrong?

[ggenovez]

Hi all,

I did some data logging last night, and when I got home I looked at the knock accumulator history table.

Then I went and looked under my truck, and the engine, transmission and differential are all hooked up, and nothing was dragging on the street. EVERY Cell I was driving in had about 40 knocks

So this morning I was playing back my datalog, and was looking at the knock accumulator, and what is going on is the knock count is being placed being copied into the RPM/MAP data instead of taking the value that is in the box, and incrementing if there is a knock.

I tried 3 adx files $42-1227747-V4.2.adx, $42-1227747-V3.adx and $42-1227747-DashHistory.adx and they all do that.

So am I looking in the wrong area? bug? What's going on...

[RobertISaar]

that's normal operation for a history table. i know of no way to have only the cells where the knock was encountered be incremented.

certain masks don't run an accumulating knock counter, instead they only show knock counts when knock is happening, then revert back to 0. those would be the only masks i can think of that would work with a history table.

[ggenovez]

So what do you do for configuring the SA table? How do you know what cells need to be adjusted?

What do you think of this? how's about creating a self adjusting SA table.
Basically I would set the spark retard to 10 degrees, leave the attack rate at .229 deg/msec and set the recovery rate to 0 output the data to a historic table. Now bump the SA table by 5'

If there is a knock it will auto retard until it stops knocking and stay there. I can then take the table an input it as my new main spark. Bump it up by 2 degrees and run it again to identify any false knocks. repeat until I have a stable table, then sub 2' for safety.

What do you think??? am I asking for trouble?

[dave w]

I use Excel to help tune out knock. The screen shots below shows examples when knock happen, then adjust the timing in SA table.

dave w

[EagleMark]

The 1227747 will always have knock counts in every cell and when it does the ESC knock test a couple cells will increase, it's an accumulater. It is not a great knock sensor circut and system like 8192 Baud systems. But I have just found a way to hack bin to output knock retard in ALDL output which when implemented would see exactly how much is being retarded in what cell and make a way better history table.

For now you look for cells that increase, if it is bouncing around in cells and all are 40 then all is OK, if 40 starts going up then that is where knock is being found.