knock voltage vs knock count
that make sense, rather than looking at one voltage, i should focus on spikes on the voltage curve. i will add some fuels near the spikes and try to smooth out the knock voltage.
Chong
Chong
Using both values is the only way to know exactly what is going on..
First, you should not be determining knock by only looking at voltage, watching the trend is a good point as made earlier, but the voltage is all of the noise picked up, the sensor is a microphone.. If you only have the voltage to go by, then you'd notice as engine speed increases, background noise (and voltage increases) spikes outside the constant background noise are generally considered knock, but the only way to know for sure is to determine what frequencies these spikes occur at. A loud clunk may be drivetrain noise, but a loud plink might be detonation, they may both result in identical spikes, but occur at two completely different frequency ranges.
This is where the knock filters and knock sum comes into play,
As far as we can tell, each filter represents a freqency range, and each cell within that range represents a threshold, the combination of all of the filters likely represents a signature that is what Mitsubishi has determined to be knock.
The only way to know for sure without using complex signal analysis is to listen when the engine is running for these noises and see what they actually are. Since most of us aren't equipped to do that, your second best option is to use the knock sums.
Although using a threshold for voltage is a good indicator, and it works, its not the ideal method for determining true knock.
Using both and comparing the two, is a great indicator though, as becomes immediately apparent that knock sums almost always occur at voltage spikes, but voltage spikes don't always result in knock sums.
First, you should not be determining knock by only looking at voltage, watching the trend is a good point as made earlier, but the voltage is all of the noise picked up, the sensor is a microphone.. If you only have the voltage to go by, then you'd notice as engine speed increases, background noise (and voltage increases) spikes outside the constant background noise are generally considered knock, but the only way to know for sure is to determine what frequencies these spikes occur at. A loud clunk may be drivetrain noise, but a loud plink might be detonation, they may both result in identical spikes, but occur at two completely different frequency ranges.
This is where the knock filters and knock sum comes into play,
As far as we can tell, each filter represents a freqency range, and each cell within that range represents a threshold, the combination of all of the filters likely represents a signature that is what Mitsubishi has determined to be knock.
The only way to know for sure without using complex signal analysis is to listen when the engine is running for these noises and see what they actually are. Since most of us aren't equipped to do that, your second best option is to use the knock sums.
Although using a threshold for voltage is a good indicator, and it works, its not the ideal method for determining true knock.
Using both and comparing the two, is a great indicator though, as becomes immediately apparent that knock sums almost always occur at voltage spikes, but voltage spikes don't always result in knock sums.
thanks for the quick info MJ.
sorry for the newbie question but...how can I get knock sum? as previously posted, i should use 2 logging system, ECU+ and ____. what's the option? there are so many and name sounds so familiar. EVOscan? Mitsulogger?
CP
sorry for the newbie question but...how can I get knock sum? as previously posted, i should use 2 logging system, ECU+ and ____. what's the option? there are so many and name sounds so familiar. EVOscan? Mitsulogger?
CP
It seems that the ECU+ knock voltage operates along the same lines as chiptorque knockbuffer. I run that on my xede and I also logg knocksum through tactrix cable/evoscan/logworks. The KnockBuffer shows knock as voltage spikes.
I have noticed that that 3-4 knock counts are rarely picked as spikes in the voltage of the Knockbuffer. Higher KS are usually picked as spikes in the KnockBuffer especially when the car hits 5000 rpm or more.
Generally, I trust the KS more than the voltage spikes of the KnockBuffer
I have noticed that that 3-4 knock counts are rarely picked as spikes in the voltage of the Knockbuffer. Higher KS are usually picked as spikes in the KnockBuffer especially when the car hits 5000 rpm or more.
Generally, I trust the KS more than the voltage spikes of the KnockBuffer
Thread
Thread Starter
Forum
Replies
Last Post
bsbllfit7
General Engine Management / Tuning Forum
20
Sep 28, 2015 05:07 PM
jameswwt
Evo Engine / Turbo / Drivetrain
14
Jan 8, 2012 05:12 PM





