Page 1 of 1

Is this false knock?

Posted: Wed Feb 11, 2015 10:28 pm
by shind3
Attaching my map and logs for what I am suspecting is false knock. It mostly tends to happen around 1300RPM and 3000RPM at around 50kPa. I only use 93 octane (R+M)/2. Only mods are a 70mm TP.

This forum collectively has seen many logs over the years and I have little experience with this particular engine so I'm hoping this forum can help me with its vast experience. Thanks.
Shek 70mm TP, 39.fpcal
datalog0035.fpdl

Re: Is this false knock?

Posted: Thu Feb 12, 2015 10:40 am
by Hondata
Nope, looks like real knock. The easiest way is to graph the knock control value and looks to see where it increases.
Most of the knock is around 3000 rpm @ 55 kpa, with a little at 2000 rpm @ 50 kpa.

Re: Is this false knock?

Posted: Thu Feb 12, 2015 12:43 pm
by shind3
Thanks for confirming! I'll keep pulling timing out over there. It's remarkable how much less timing there is there compared to stock. I think I've pulled 4 degrees out so far. I guess that's a good sign. The 70mm TP must be increasing VE there?

And yes, that's how I look for knock. I zoom in wherever the knock count increases from a previous value.

Re: Is this false knock?

Posted: Fri Mar 06, 2015 8:01 am
by shind3
I pulled more timing out and was able to get logs with almost no knock. You can see this in the logs for "Shek 49" calibration.
Shek 70mm TP, 49.fpcal
Shek 49 (1).fpdl
Shek 49 (2).fpdl
Shek 49 (3).fpdl
Shek 49 (4).fpdl
Based on this, I pulled a little bit of timing in the areas as indicated by the logs, made some slight changes to certain cells to make the AFR more flat (pull up 12.8s to 13.1s etc.) at WOT, and adjusted the throttle calibration table.

Then I noticed logs with some ridiculous knock counts that seem to occur all at once when in part throttle. For example, knock counts went from 7 to 36 within 2 seconds at half throttle at 4000RPM.

You can use the 'compare' feature in FlashPro to compare "Shek 50" calibration to "Shek 49" and see how little was changed. Then look at the logs below.

Re: Is this false knock?

Posted: Fri Mar 06, 2015 8:04 am
by shind3
Compare this to "Shek 49".
Shek 70mm TP, 50.fpcal
Shek 50 (1).fpdl
Shek 50 (2).fpdl
Shek 50 (3).fpdl

I can't figure what is going and was hoping someone on here would be able to help. Is this false knock? I started suspecting the Cam Chain Tensioner but not sure if I'm hearing anything abnormal with the engine to indicate that. Has a failing CCT been known to cause false knock?

Engine is completely stock with 98K miles. Sparkplugs are tight. Oil level is indicated full on both sides of dipstick. Valve Lash (0.009"/0.011" I/E) was checked 2000 miles ago.

Engine after 30 mins drive in 20F weather so it should be at full operating temp.
https://www.youtube.com/watch?v=ev_fnJLXCog

Engine after sitting in a 60F garage overnight so ECT should be about 60F
https://www.youtube.com/watch?v=1rhOibOTVLU


Anyone hear a failing CCT/TCT in these videos? If it isn't the CCT/TCT then what is causing this? Is this real knock? If so,what's causing it?


Thanks for your help.

Re: Is this false knock?

Posted: Fri Mar 13, 2015 11:09 am
by shind3
OK, update. Turns out all of this WAS false knock. Caused either by a aft LH differential mount with 98K miles on it (looks and feels fine) or more likely, a Miata sway bar I installed at the rear that might have been rubbing against the aft differential mount nuts.

I removed the rear swaybar, and replaced the aft LH diff mount and the knocks went away with no tune changes. I have about an hours worth of datalogs to prove it.

Then I added back all the 'Ignition Low' I had removed (also changed TCT as a precaution) and still no knocks! In some cases I had removed 10 degrees of timing compared to stock. Car feels much more lively to drive at low RPM.

It's amazing that noise from a rear swaybar was being communicated all the way to the microphone bolted to the block...

Re: Is this false knock?

Posted: Fri Mar 13, 2015 3:10 pm
by Hondata
Thanks for the update. The usual culprit is the exhaust header hitting somewhere.