Knock control and knock count

Hello,

While tuning ignition and datalogging, if you find knock in a certain ignition cell when playing back the log, would you retard that individual cell or the entire row of that cell? Based off the logic that, if knock is happening in one cell, it's most likely happening in that entire row of cells? if say knock is happening at say 203mbar at 2750 rpm, then most likely it is happening throughout the entire row for that cell (118mbar 500rpm - 1026mbar 8000rpm)?

I found that, when I advanced my high load values, I didn't detect any knock in those areas and knock control was fine; but in the cells where no advance was made - i found knock count go up. Some instances, I retard the timing by .5, log again, and find that same cell is still detecting knock. How much timing would be safe to remove?

Also, there is a safe amount of knock that does occur; from my understanding more important than knock count is knock control. Going through Hondata, Ktuner forums/Q&As etc, I found that even if you get say 6 knock counts in 15 mintues of driving but still retain a low knock value; even that shouldn't be a worry as knock control value plays a more cricial role in determining specific knock.

Some clarity would be appreciated, thanks :)

I may certainly be corrected on this, but my 5 cents worth.

At different rpm, and loadings/boost, the engine's combustion characteristics will change - this can be due to many things, such as camshaft timing variations, inlet and exhaust breathing characteristics affecting cylinder filling, combustion chamber shape and clearances, etc.

Because of this, the MBT and 'knock' timing may vary from the expected values (even cylinder to cylinder), and there may be points where more timing is required and times less. If you find one cell needs to be trimmed back at one load and rpm site, it may just be due to the engine's characteristics at that point. 0.5 degrees isn't going to make a lot of difference, I'd suggest pulling 5 and then adding it back in 2 degree increments until it's detected, then backing of to the last 'safe' point. I'd also suggest double checking the adjacent cells as they may be on the threshhold and could be an issue if fuel quality drops slightly.

That said, are you tuning to the knock threshhold, which may not be optimal, or to MBT and running into knock limitations?

Thank you so much for your insight I'll try your suggestions.

Currently unable to get to a dyno, so I am road tuning. In terms of what you had asked, I generally advance timing until I see knock or an increase thereof. Where I advance (1-2 degrees) in the high load cells in WOT I do not run into knock; suggesting to me that I have not reached MBT in those areas of advance. I see knock mainly in three different areas - when rpm drops down after high speed in the lower load area of the cells, sometimes at a cruising speed on 5/6th gear, and lastly, start of WOT pulls (upon initial acceleration of rpm). Ill go ahead and try what you suggested and see the results from datalogging. What does confuse me though is that id understand seeing knock in the high load where I've tuned the ignition map, the low load area has not been advanced and is running the basemap, yet im seeing a knock or two being detected from the basemap ignition degrees in cruise.

I am mainly looking at the knock control value and how its fluctuating (generally I range from 0.20 on the low cruise area and 0.50 ish on the high), and the number of knock counts the engine is detecting at the certain cell in the ignition table; where then Id make adjustments based off the knock readings. I should go ahead and say that I'll be looking at a knock sound detection device for more accuracy, although the oem sensors installed is quite modern and accurate in itself for the readings.