Knock event tracing from csv output

Miha__

Member
Joined
Sep 10, 2017
Threads
10
Messages
34
Reaction score
5
Location
Slovenia
Vehicle(s)
civic
Country flag
Hi,

I want to track back the exact moment that knock has been detected
Can you help me to recreate values within ignition table:
Honda Civic 10th gen Knock event tracing from csv output upload_2019-12-14_13-19-4


Goal is to recreate this table in excel, add use X in exact field that knock has happened.

For input data, i will use data logs in CSV.

So fields that are available are

Time(s) | RPM | MAP(mBar) | MAP(Conv) | TC,Target | TC,Press | MAF,v | MAF | PWMDC | TPS | TPS,CMD | CAM,CMD | CAMA | EXCAM,CMD | EXCAMA | AIGN | FDC | FP1 | STFT | LTFT | LAM | LAM,ADJ | LAM,CMD | FUEL,STAT | KNK,C | KNK,CTRL | ECT(C) | IAT(C) | VSS(KPH) | GEAR | BAT | VTEC

Does anybody know how to calculate tabulated ignition timing ?

Also X axis (LOAD) I want to confirm that it represents MAP(mBar) ?
Sponsored

 

kytos

Senior Member
First Name
Marcos
Joined
Apr 18, 2018
Threads
7
Messages
601
Reaction score
416
Location
Madrid - Spain
Website
www.instagram.com
Vehicle(s)
CIVIC SPORT 1.5t 2017 EUDM
Country flag
Knock count doesnt count knock event but missfire events. Monitor knock control movement.
 
OP
OP

Miha__

Member
Joined
Sep 10, 2017
Threads
10
Messages
34
Reaction score
5
Location
Slovenia
Vehicle(s)
civic
Country flag
"Knock Count" = This parameter is a reflection of the ECU's misfire monitor. It is NOT an accurate reflection of anything other than misfire events. When will you see these normally? Startup, rapid throttle closures, rapid throttle inputs, rapid release of the clutch pedal, and sometimes at idle (especially if you have a lighter flywheel/clutch combo, and/or plugs that have been gapped down to prevent spark blowout at high load). A completely stock car, operating perfectly will still accumulate these counts. In my experience with these engines, you will see more activity on cylinder four compared to others, as the crank position sensor is physically located closest to there on the block. This usually isn't a concern area unless you see a steep count in a very, very short time (like 50+ in a few seconds), which will likely trip a P0300-variant DTC code.

"Knock Control" = This parameter is the ECU's determination of fuel quality. Movement here indicates the knock sensor hears what it thinks is knock activity, and reports to the ECU to apply a steeper ignition retard to avoid continued knock activity. This value is dynamic, and WILL move from time to time. On Civic Si models, there doesn't seem to be a forced rise at play at WOT like the non-Si 1.5T ECU's (which naturally rise above 5,200-5,400rpm regardless of sensor input). Movement that goes up and up and up and never comes down is more concerning than movement alone. Knock control can typically be manipulated down by driving the car in a higher gear at lower engine speeds and targeting atmospheric pressure on the MAP sensor reading."

SO

I'm interested it the points that knock count is produced. If the event is constantly repeated at the same location you have an issue. If not you can safely register it as assumed misfire.
 

D-RobIMW

www.imwtuned.com
First Name
Derek
Joined
Oct 15, 2017
Threads
1
Messages
379
Reaction score
1,075
Location
Carlisle, PA
Vehicle(s)
2017 Civic Si, 2019 Accord Sport 2.0T, 2001 Integra Type R, 2019 RDX A-Spec
Country flag
"Knock Count" = This parameter is a reflection of the ECU's misfire monitor. It is NOT an accurate reflection of anything other than misfire events. When will you see these normally? Startup, rapid throttle closures, rapid throttle inputs, rapid release of the clutch pedal, and sometimes at idle (especially if you have a lighter flywheel/clutch combo, and/or plugs that have been gapped down to prevent spark blowout at high load). A completely stock car, operating perfectly will still accumulate these counts. In my experience with these engines, you will see more activity on cylinder four compared to others, as the crank position sensor is physically located closest to there on the block. This usually isn't a concern area unless you see a steep count in a very, very short time (like 50+ in a few seconds), which will likely trip a P0300-variant DTC code.

"Knock Control" = This parameter is the ECU's determination of fuel quality. Movement here indicates the knock sensor hears what it thinks is knock activity, and reports to the ECU to apply a steeper ignition retard to avoid continued knock activity. This value is dynamic, and WILL move from time to time. On Civic Si models, there doesn't seem to be a forced rise at play at WOT like the non-Si 1.5T ECU's (which naturally rise above 5,200-5,400rpm regardless of sensor input). Movement that goes up and up and up and never comes down is more concerning than movement alone. Knock control can typically be manipulated down by driving the car in a higher gear at lower engine speeds and targeting atmospheric pressure on the MAP sensor reading."

SO

I'm interested it the points that knock count is produced. If the event is constantly repeated at the same location you have an issue. If not you can safely register it as assumed misfire.

You can accumulate "knock count" in KTuner without moving the control value. Again, "count" is misfire monitor, not knock sensor activity, so don't use that as a current means of determining knock activity,
 
OP
OP

Miha__

Member
Joined
Sep 10, 2017
Threads
10
Messages
34
Reaction score
5
Location
Slovenia
Vehicle(s)
civic
Country flag
So why would Honda add this specific parameter and name it knock count ?
What is the benefit of monitoring misfire vs monitoring knock ?
How it is then used in ECU logic? What is this knock count driving?

Or maybe Ktuner can rename it to “Misfire count” and this will solve the confusion?
 
Last edited:


D-RobIMW

www.imwtuned.com
First Name
Derek
Joined
Oct 15, 2017
Threads
1
Messages
379
Reaction score
1,075
Location
Carlisle, PA
Vehicle(s)
2017 Civic Si, 2019 Accord Sport 2.0T, 2001 Integra Type R, 2019 RDX A-Spec
Country flag
So why would Honda add this specific parameter and name it knock count ?
What is the benefit of monitoring misfire vs monitoring knock ?
How it is then used in ECU logic? What is this knock count driving?

Or maybe Ktuner can rename it to “Misfire count” and this will solve the confusion?
It's simply how Honda named the parameter.

Misfire counter exists to determine if the engine has a misfire condition, as stated in the text I wrote (which you quoted/posted).
 

xcivicx

Senior Member
Joined
Jan 14, 2016
Threads
69
Messages
933
Reaction score
490
Location
whycanogaparkofcourse.
Vehicle(s)
19 SI coupe in black
Country flag
I'm monitoring such data before I tune. I've noted just as said above that count goes up under the conditions stated above. They seem to stay until cleared somehow.
Control will climb to 69 and keep there for some time until?, have seen it drop quickly driving up my hill.
91 arco gas.
 

kytos

Senior Member
First Name
Marcos
Joined
Apr 18, 2018
Threads
7
Messages
601
Reaction score
416
Location
Madrid - Spain
Website
www.instagram.com
Vehicle(s)
CIVIC SPORT 1.5t 2017 EUDM
Country flag
Knock control goes down over time, the fastest method to drop it is just target 0 or 1 psi with your foot, specially in non-si ecus that have an artificial rise past 5300 rpm or so.

Miha, why not just watch knock control movement?
 

varge

Senior Member
First Name
George
Joined
Oct 7, 2017
Threads
3
Messages
348
Reaction score
478
Location
Greece
Vehicle(s)
Civic 1.5T 4D Elegance EUDM
Country flag
I would think that Miha wants to tune ignition timing and wants to correlate actual knock events with specific rpm/boost/IGN/AFR data in order to determine how far he can push things. I think that when working with a non-Si , where K.Control will just rise once you get past 5300rpm you will be working more or less in the dark as there doesn't seem to be any tangible information coming from the ECU (unless someone is willing to fill up with crap gas, push boost to the max and add 10 degrees to IGN to see if K.Control will go nuts before the engine is blown :p )
 


D-RobIMW

www.imwtuned.com
First Name
Derek
Joined
Oct 15, 2017
Threads
1
Messages
379
Reaction score
1,075
Location
Carlisle, PA
Vehicle(s)
2017 Civic Si, 2019 Accord Sport 2.0T, 2001 Integra Type R, 2019 RDX A-Spec
Country flag
I would think that Miha wants to tune ignition timing and wants to correlate actual knock events with specific rpm/boost/IGN/AFR data in order to determine how far he can push things. I think that when working with a non-Si , where K.Control will just rise once you get past 5300rpm you will be working more or less in the dark as there doesn't seem to be any tangible information coming from the ECU (unless someone is willing to fill up with crap gas, push boost to the max and add 10 degrees to IGN to see if K.Control will go nuts before the engine is blown :p )
You can still discern the "normal" rise from abnormal rise.
 

varge

Senior Member
First Name
George
Joined
Oct 7, 2017
Threads
3
Messages
348
Reaction score
478
Location
Greece
Vehicle(s)
Civic 1.5T 4D Elegance EUDM
Country flag
You can still discern the "normal" rise from abnormal rise.
It is relieving to know that someone has witnessed an abnormal rise (presumably due to pushing things too far, or crap gas), most probably meaning that those of us who don't are more or less on the safe side :)
Sponsored

 
Last edited:


 


Top