Originally Posted by Red Dragons

How do you know this?

Originally Posted by Lonewolf176

From tuning for 15 years? Try it. when you see 6 in your logs try removing 2 and retest.

Originally Posted by b0fh

it can be made less aggressive via some parameters (I do that), but you are right it has a tendency to over-react. if the adaptive one prevents knocking to boot there is no need for the quick retard to even kick in

I keep the adaptive permanently disabled, but since my tune is 'done' for at least 1 year by now I may enable it limited to 1 or 2 degrees to see how it behaves. in the beginning it was annoying AF, always interfering with the logs (on the PV it used to show a retard of -3 all the time)

Originally Posted by b0fh

it can be made less aggressive via some parameters (I do that), but you are right it has a tendency to over-react. if the adaptive one prevents knocking to boot there is no need for the quick retard to even kick in

I keep the adaptive permanently disabled, but since my tune is 'done' for at least 1 year by now I may enable it limited to 1 or 2 degrees to see how it behaves. in the beginning it was annoying AF, always interfering with the logs (on the PV it used to show a retard of -3 all the time)

Originally Posted by b0fh

tough crowd here, I missed you guys

yes they are *different* things, and they show differently in different log elements! :P

for the M8 ECU:

knock retard is a short lived event, ECU detects the knocking using the new piezo sensors and pull timing quickly, and then adds it back (also quickly, unless knocking is detected again then it keeps going up and down - not good). in a properly tuned engine it should be a short lived event that happens on hot days under acceleration only.

adaptive knock retard is a totally different beast. the ECU decides based on 'X number of knock events' (those described above) that you are running **** fuel and pulls up to X degrees or timing pretty much forever. you can see this on logs as a permanent difference between your spark advance table and the actual spark advance. those small tables is what controls it, and I run my bike with them zeroed out, because if there is a problem with my timing I'd rather adjust it on the main tables.

Gas here in brazil is crap everywhere, but in the US where you can easily get different 'grades' of gas this adaptive thing is a good idea - I just wish it would reset faster, in my experience it never 'unlearns' (perhaps it does after 50 key-on cycles like mentioned before, but thats like an eternity).

Originally Posted by stroisi12

What I can see in the logs is the difference between Advance spark (logs) and advance spark from spark advance table, constantly changing, up and down. I have not been able to correlate it with other log data.

I’m still trying to figure out what/ where in the logs can identify a knock event? Knockcnt doesn’t seem to be it. Spark intensity? Not that I can see as it’s all over the map. No correlation with Spark Knock which I think is timing being pulled (short term? Or a real time knock event?).

here is what I was trying to ask in my OP:
what excactly are the knock related signals reporting:
knockcnt
advance spark
spark knock
spark intensity

Originally Posted by Steve Cole

A lot of this depends on the year and model bike your working with as they are all not even close to the same. That said you also must be aware that the data being recorded you are basing your decisions on is not telling you the truth! The bike at an idle of 1000 RPM fires each spark plug 8.33 times per second and as RPM increases so does the Spark plug firing, when you reach 6000 RPM its now at 50 timer per second per cylinder. So let's say your at 3000 RPM and you see knock occur. The question you have to ask yourself is that report true? The answer is absolutely not true, due to the speed that the recording is being made at. All it shows you is after it has already happened.

Originally Posted by Red Dragons

I'm pretty sure most tuners are aware that it may not be at the EXACT spot it's being recorded but it's pretty close. When I adjust my spark tables I'm using an aggregate of reported knocks to generally locate a trouble area, and pull timing accordingly. I obviously know if it works or not based on the next test run.

Originally Posted by Steve Cole

All depends on what one calls as close. I've seen it be off as much as 600 RPM when the data recording speeds are the way HD has them to begin with. Then the newer the model years and software, the worse it gets. As an example, the software HD release for the 21 model year is terrible it can and does cause offsets from what the data shows by over 1000 RPM. This is the very reason HD toss that software out and replaced in in the fist 3 months after release of the 21 model year bikes, but many are still using it in the field.