How to detect when real knock occurs

Turboroe

Member
Joined
Jul 8, 2019
Threads
6
Messages
36
Reaction score
20
Location
Porto Alegre, Brazil
Vehicle(s)
Honda Civic Touring 1.5L 2018
Country flag
Can anyone tell me how to know when knock is actually happening in a datalog?

I get many knock counts on some trips, and little in others, and I can´t seem to find a pattern of causes or consequences to increases in knock counts.

The information is very limited on this subject, I only hear people saying that Knock counts are not a good indicative and that I should watch KControl.
I have tried looking for increases in KControl, but then there are times where i´d get many knock counts and KControl does not change, and others where KControl changes without any knock count alteration, even if i'm driving normally.

it would be great if someone could explain to me the logic around knock on this engine?

I have attached a datalog example a commute on a very hot day here where I got about 30 knocks and a situation with KControl above 80%.
Sponsored

 

Attachments

gtman

Senior Member
First Name
Mitch
Joined
Oct 27, 2015
Threads
328
Messages
16,661
Reaction score
24,351
Location
USA
Website
www.civicx.com
Vehicle(s)
2017 Cosmic Blue EX-L Sedan
Vehicle Showcase
2
According to the pros here, "Knock count is misfire count on this setup and unless it is so bad that you are throwing a check engine light, it is usually best ignored". (source: 360glitch)

Some people have reported multiple "knocks" going over a pothole! Concentrate on K. Control which goes up or down based on factors like temps, load, octane etc., and increases or decreases timing accordingly.

I did notice something that I hope is just a glitch in the datalog. Your short term fuel trims were great but your long term fuel trims were locked at -100 the entire 40 minutes. :eek:
 
Last edited:
OP
OP
Turboroe

Turboroe

Member
Joined
Jul 8, 2019
Threads
6
Messages
36
Reaction score
20
Location
Porto Alegre, Brazil
Vehicle(s)
Honda Civic Touring 1.5L 2018
Country flag
According to the pros here, "Knock count is misfire count on this setup and unless it is so bad that you are throwing a check engine light, it is usually best ignored". (source: 360glitch)

Some people have reported multiple "knocks" going over a pothole! Concentrate on K. Control which goes up or down based on factors like temps, load, octane etc., and increases or decreases timing accordingly.

I did notice something that I hope is just a glitch in the datalog. Your short term fuel trims were great but your long term fuel trims were locked at -100 the entire 40 minutes. :eek:
Yeah I checked that with hondata and turns out that my ECU does not have LTFT, i have no idea why, so it is always -100.

About those misfires, what are the causes and consequences of them? I ran some datalogs when my ECU was stock and i never got any knock counts whatsoever.
 

Jpierro79

Senior Member
First Name
John
Joined
Mar 6, 2019
Threads
33
Messages
398
Reaction score
273
Location
Myrtle beach South Carolina
Vehicle(s)
2018 type R 2015 c300
Country flag
Yeah I checked that with hondata and turns out that my ECU does not have LTFT, i have no idea why, so it is always -100.

About those misfires, what are the causes and consequences of them? I ran some datalogs when my ECU was stock and i never got any knock counts whatsoever.
If your getting knock counts after tune your fuel is more likely insufficient.poor 91 or 93 octane. Knock is not always what you think but noise detected from the engine based upon the level of noise the ecu reduces timing. It doesn’t necessarily mean you have actual knock. The stock settings are very high to prevent issues if you get a bad tank of fuel. By increasing turbo pressure and timing can cause “knock” or engine noise. Yes you can get knock reading going over pot holes cause it send vibration through the whole car vibrating the actual knock sensor giving a false reading. Running a open downpipe can cause the same thing if the outlet is close enough to the engine block. Do NOT decrease the knock sensitivity!!!! I can help you with this provided I have a copy of what time your using your mods and what fuel. I’m talking from 21 years of professional experience. I’m not someone who learned all my information from the internet but real world.
 

Andre80

Senior Member
First Name
white civic
Joined
Feb 23, 2018
Threads
7
Messages
338
Reaction score
88
Location
Europe
Vehicle(s)
Civic X 1.5 sport +
Country flag
Yeah I checked that with hondata and turns out that my ECU does not have LTFT, i have no idea why, so it is always -100.



Have you tried an odb2 diagnosis?
 


hobby-man

Senior Member
First Name
Mike
Joined
Jun 20, 2018
Threads
50
Messages
2,867
Reaction score
2,739
Location
GTA
Vehicle(s)
2021 WRX STi, 2018 Sport Touring Hatchback (sold)
Build Thread
Link
Country flag
One thing I'll just add on, is that KCon. will increase no matter what, if you go above a certain RPM (5k?) on a Non-Si ECU. It's just a safeguard. So if you're seeing KCon. rise and you're hitting those high RPM, that's why.
 

Jpierro79

Senior Member
First Name
John
Joined
Mar 6, 2019
Threads
33
Messages
398
Reaction score
273
Location
Myrtle beach South Carolina
Vehicle(s)
2018 type R 2015 c300
Country flag
One thing I'll just add on, is that KCon. will increase no matter what, if you go above a certain RPM (5k?) on a Non-Si ECU. It's just a safeguard. So if you're seeing KCon. rise and you're hitting those high RPM, that's why.
Yes it’s an increased knock sensitivity multiplier to keep with the demand at high rpm. Good point thanks. It’s actually incremental on type r ecu. On this ecu I’m not sure if it increases the amount of time the ignition is retarded when knock is detected.
the stock ignition charts will make a LOT of burst knock between 2200-4K. I actually had to reduce as much as 4 degrees in the lower load to stop burst knock so it would build boost better. Burst knock is when you rapidly depress the pedal and knock occurs cause the sudden burst of air into the cylinder is more rapid and uneven.
I’ll say this there’s a lot of head room in Hondatas OTS. 93 map. I’ve added up to 4 degrees of timing up top
I run to 7200 no problem no drop off
I run 24-25 psi from 3500 fading out to 20. The stock turbo flows more than most people think.
I run slightly richer in the lower rpm but not too rich as it increases injector pulse width and the direct injection window is tiny. I will not run 11.8 at 3500 it’s a waste of fuel you don’t make any more power than with 12.5. If anything you slow turbo spool cause cooler fuel burn. Lean is hotter.
I also adjusted cam timing on spool charts.
If you felt a custom tune vs OTS you’d be shocked how much room is there.
Sponsored

 


 


Top