The above is incorrect.where the strike through was added. The ECM sends the data out as packets at all times over the CAN buss. It just how far in between each packet comes out from the ECM. While the ECM is fast when you add everything it must do and the priority at which its does, things can get pretty far out of wack. Since scan data is the lowest priority everything else gets done first. So your first packet of scan data comes and it might take as much as 100 ms before the second packet comes, you just never know for sure.

 

I maybe should have started with the basics first by looking at KNOCK COUNT... which is not one of the standard enabled channels on the PV, so I didn't see it. I enabled Knock Count and have ZERO knock (both with and without adaptive knock table zeroed). I've been looking at Spark Knock Retard, which apparently returns a value even when knock count = 0.

Does this mean the ECM senses knock beforeit happens? I understand there are tens of thousands of missing data points in my batch of 30,000 and a knock event could have happened in that missing info. I would expect ONE knock event to occur outside the missing data.

Having said that, the bike feels more alive with the adaptive knock table zeroed. Then I input the previous values into the adaptive knock table and the bike feels as good as it did with the adaptive knock table zeroed.The Spark Knock Retard numbers were drastically lower and less frequent with the Adaptive Knock table zeroed. My previous example of cruising at 75mph, 6th gear, 2950 RPM, 55 kPa MAP would result in 2.25 deg F / 1.75 deg R spark knock retard. With the Adaptive Knock table zeroed, I'm getting ZERO Spark Knock Retard...

Will the bike re-learn the adaptive table over time?
Based on my zero knock events, should I just ride and forget about it?

 

By putting "0" in the adaptive knock tables you limit the Adaptive Spark Retard but leave the Knock system working. This allows you to see what is happening at the time you are testing without any of the learned knock jumping in. So if you ride the bike with Adaptive set to 0 in the limit tables it will help the Adaptive learn it way back to 0 while the knock system still functions and removes knock as needed. It's not a good idea to leave it that way permanently.

 

That’s great info Steve! I’ll zero the adaptive knock table for a little while (200 miles? 500?) and see where that takes me.

Instead of zeroing the table, couldn’t I just put the minimum MAP for adaptive knock enable to 500?

Either way, the bike feels so much better already. I zeroed the adaptive knock table, ran an hour long “auto” tune session (ie, correcting VE tables), and noticed a 5% average change in VE.

Re-enabled the adaptive knock table and rode 35 miles and the bike feels good. Very windy today and the bike isn’t working so hard anymore.

Now what? How do I know if the timing isn’t pulled too much? Keep an eye on engine temps? The bike has always seemed to run hot (65F ambient, 330 engine temp at 75mph highway cruising).

Maybe I should just stop asking questions.

 

nah...you on a roll..and i'm learning something.

 

I had a revelation. Running 87 octane will fugg up your adaptive knock retard…

I got into the habit of running 87 because it is the only available non ethanol here. That might explain a little bit of the situation here.

Since tubing, I have ONLY run 91 E10, but still, I have to wait for the bike to relearn the adaptive knock…

 

If you screw with the knock system it shuts it ALL down. Then you have no knock system, period. The Adaptive is there to learn but when you change the bike around it may have learned the wrong thing and over time it will learn back to "0" if the current calibration is not knocking.

 

This is what it’s all about. My ride home from work. Highest ever by 10 mpg: 51 mpg. We’ll see how accurate that is when I fill up. Power is up and the bike is more efficient: win-win!