Forum » General Tuning Discussion » Plex knock monitor operation

Plex knock monitor operation

General Tuning Discussion

Discuss all things tuning in this section. News, products, problems and results. 

Page 1

Hi Andre,

I'm using my v2 Plex knock monitor on an evo x with forged pistons. I have the knock sensor mounted on the engine head.

I have a question concerning the knock frequency in relation to the audio filter frequency. Do you have to always select the audio filter closest to your knock frequency?

I am currently using the second harmonic knock frequency for the evo x engine since the sensor is not mounted in the most ideal place together with the fact that this car has forged pistons. I am new to tuning so deciphering knock with my ears via the headset is the main problem. I currently have the ignition timing in the car quite retarded....I took around 6 degrees out of the wot area of the stock JDM map.

With the knock frequency set on the 2nd harmonic I've cycled through all the audio filters and I've found the 9khz bandpass filter is giving me the least amount of "noise" through the headset...With all the others it sounds really noisy and messy and I don't think if the engine was knocking that I can pick out the difference.....This is what I mean the 9khz is sounding the cleanest. Can I make a decision like this by just cycling through all the filters ? Or do you need to actually know what knock sounds like and listen to the engine knock under all the different filters before deciding on a filter ?

To me using the 9khz during a ramp run I just hear the background noise getting louder and louder as I go through the rev range but I don't hear anything distinctly different....Together with the fact that I said I've removed around 6 degrees of timing.... I think what I'm hearing is all background noise and no knock.

I've tried the 10khz high pass and with this I can definitely hear some different sounds through out the run but I am pretty sure this can be knock because it's mean the engine was knocking quite a lot and with so much timing removed it's highly unlikely.

The other filters are just "louder" and as I say deciphering Anything from them...To me.atleast seems impossible.

I'd appreciate some feedback on this, thanks

While I know this question is addressed to Andre, I have found some useful info here:-

"The engine tuner will advance the timing until best torque output is found, or he can hear a faint irregular spattering against the background noise of the regular sound of reciprocating engine components in the headphones. This widely spaced irregular spattering sound is the onset of knock and most un-tuned road engines will sustain light knock for extended periods without damage.

It is the irregularity of light knock that we are listening for, rather like the sound of the first drops of rain on a tin roof. Advancing the ignition to light knock has the same characteristic irregularity as sparse raindrops, then passing into more knock events until there is a storm of spatters as ignition is advanced. When there are so many knocking events that it sounds like a hailstorm, back off immediately! The idea is to listen for the onset of knock, (this first few spits per second) and then back off a degree of timing or so and run again to check the setting; but more of that later."

Also there are sound files available that clearly show the difference between detonation and no detonation, which are available here:-

Many years ago I learned what knock sounded like by experimenting on my old VH Holden Commodore by manually adjusting the distributor a few degrees at a time until knock got worse and worse.. Those old engines made such low specific power they could knock all day and not get damaged.. :)




Thanks a lot for that mate it'll definitely be helpful.

Much appreciated.

Bump for this...Anyone knowledgeable on the Plex or audio knock detection ?

When you're selecting an audio filter I'll start with the narrow pass filter closest to the calculated knock frequency for the bore size, however this isn't always the most clear filter frequency in my experience. I'll typically cycle through the different filters and test to see which one gives me the best signal to noise ratio. the important part though is to purposely induce some light knock to ensure you can accurately distinguish it. You should be checking this in conjunction with the plot of background noise vs rpm to ensure the knock monitor is simultaneously displaying the knock that you're hearing.

Andre, from cycling through all the filters to me the 9khz sounds like it has the best SNR. I'll do what you recommend and try to induce some light knock at low rpm.

The issue with looking at the rpm vs noise plot is that I am tuning this car on the road and my innovate inductive clamp is a bit clumsy to have in the engine bay with the hood down.

Today, i am going to spend some time making some connectors for using the CAN, Most of my cars are Haltech and Hondata, try and keep most of the wiring in-car.