Hi there
I have knock counts routinely on cylinder 3 and hoping one of you might help me troubleshoot.
It's an FN2 and I am using using the group N tune as a starting point. It's stock aside from an intake snorkel and a panel filter. Have only been calibrating the AFM Flow table and a bit of work on the WOT lambda tables to date. Using MAP to determine WOT. No ignition timing or cam angle changes.
It's got a new fuel filter, fresh spark plugs and it's not the ignition coil. I moved the ignition coil to another cylinder and it is still occurring.
My thinking is it could be:
- a crap tank of fuel
- a blocked injector - should I swap the injector on 3 to another cylinder and see if the knocks move to that cylinder?
- a wrong spark plug and I should put a colder range spark plug on cyl 3?
- maybe adjust the cylinder trim (either fuel or timing)? (done 2 degrees and no luck there)
- I need to adjust timing (retard) in that portion of the map
- that my driving is causing it (you'll notice it's low in the rev range and pretty heat soaked when it happens, there's lots of traffic in my area)
- I have a leak somewhere? after the MAF sensor? I notice I only get 96kpa at WOT
Logs attached and sorry they are snippets of a larger log file to be able to upload them here
Any suggestions would be great!
FN2 cylinder three knock question
FN2 cylinder three knock question
You do not have the required permissions to view the files attached to this post.
Re: FN2 cylinder three knock question
The best thing to do is to remove 2 degrees of timing from cylinder #3. If the knock count deceases or goes away, it was knock. If nothing changes, it is mechanical noise. In that case look for cam train noise & exhaust hitting body etc.
Hondata
Re: FN2 cylinder three knock question
Thanks for responding and I will try that too.
Just an update - I did spend a bit of time analyzing where the knocks were occurring in the datalogs and it seemed to be the MAP values right where the ecu looked like it was referencing/switching from low cam to high cam. I noticed the wot lambda values were quite different in the WOT lambda low table (11.00 say at 2750 rpm) to what the values were in the WOT lambda high table (12.89 at 3000rpm) so the amount of fuel being commanded when it went from low table to high table under load was bit of a step change.
To fix this I tried to match the values as best I could at points where it could be either low or high cam depending on driving situation and then used the (excellent) flashpro interpolation function to smooth the WOT lambda tables and it reduced the knock count pretty dramatically. I still get a few but after a much longer time period and they are shared between cylinder 2 and 3. For example in 25 minutes I was getting 30+ knocks and now getting 6-7.
Having said that at the same time I did a bit more work on the AFM flow table to smooth out the curve in places and also used a different brand of fuel so it might be blind luck. Hard to just vary one thing at a time!
P.S. As an aside, for those with FN2's like me, I recommend using the let MAP determine WOT function once you are happy with the AFM flow table tune. The improved throttle response is great but you can't really get much data above 2.8V for tuning that part of the AFM flow table so I suggest tune that portion then check that feature to on.
Cheers
Just an update - I did spend a bit of time analyzing where the knocks were occurring in the datalogs and it seemed to be the MAP values right where the ecu looked like it was referencing/switching from low cam to high cam. I noticed the wot lambda values were quite different in the WOT lambda low table (11.00 say at 2750 rpm) to what the values were in the WOT lambda high table (12.89 at 3000rpm) so the amount of fuel being commanded when it went from low table to high table under load was bit of a step change.
To fix this I tried to match the values as best I could at points where it could be either low or high cam depending on driving situation and then used the (excellent) flashpro interpolation function to smooth the WOT lambda tables and it reduced the knock count pretty dramatically. I still get a few but after a much longer time period and they are shared between cylinder 2 and 3. For example in 25 minutes I was getting 30+ knocks and now getting 6-7.
Having said that at the same time I did a bit more work on the AFM flow table to smooth out the curve in places and also used a different brand of fuel so it might be blind luck. Hard to just vary one thing at a time!
P.S. As an aside, for those with FN2's like me, I recommend using the let MAP determine WOT function once you are happy with the AFM flow table tune. The improved throttle response is great but you can't really get much data above 2.8V for tuning that part of the AFM flow table so I suggest tune that portion then check that feature to on.
Cheers