M1 knock sensor frequency bandwidth

Discussion and Support for MoTeC's M1 series ECUs

M1 knock sensor frequency bandwidth

Postby NicD on Thu Oct 12, 2023 5:28 am

I am not sure if this is a hard coded value in all M1s or if it's package specific, but does anybody know the knock sensor frequency bandwidth? For example you want to use 9,000 Hz for your knock frequency but how far does that extend above and below? I believe Haltech and Holley have a 500 Hz bandwidth centered on the chosen frequency so that would be 250 Hz above/below but I just can't find anything with this controller.
NicD
 
Posts: 6
Joined: Fri Sep 01, 2023 4:28 pm

Re: M1 knock sensor frequency bandwidth

Postby Stephen Dean on Thu Oct 12, 2023 10:30 am

Hi,

Similar with the M1, the signal attenuates in the amount of amplification applied on a 45º slope down from the specified frequency by 250Hz on either side, it actually reads further than this but the signal weakens the further from the amplification peak that it goes.

With a 9000 Hz primary frequency, I would use 8500 Hz, 9000 Hz, 9500 Hz and 18,000 Hz as the frequencies monitored. 18,000 Hz is the second order frequency and can be used to help filter real knock from background noise.
Stephen Dean
MoTeC Research Centre Melbourne, Australia
User avatar
Stephen Dean
Site Admin
 
Posts: 1857
Joined: Tue Dec 02, 2014 10:29 am
Location: Melbourne

Re: M1 knock sensor frequency bandwidth

Postby NicD on Thu Oct 12, 2023 11:31 am

Perfect that's exactly what I needed to know, thanks!

I will say that most OEMs use the 2nd order harmonic as their primary/monitored frequency as it has a better signal to noise ratio, at least with the 2 wire wideband flat response knock sensors. 9000 Hz would be the 2nd order, 3rd order would be ~11.5 Hz.
NicD
 
Posts: 6
Joined: Fri Sep 01, 2023 4:28 pm

Re: M1 knock sensor frequency bandwidth

Postby Rub87 on Tue Mar 17, 2026 10:51 am

Is there a way to capture the knock sensors output in the knock window in the ecu in a healthy sample rate (>50 kHz) to be able to do a fft on it?

It would be awesome if the software could generate x amount of cycles w y amount of extra spark and then during this cycles capture the raw signal. Lets say it can capture 100 cycles to its internal memory and it would run each 10 cycles one with 4 or 5 or whatever degree of advance on top. that would allow to do a fft on it and have a perfect view on the optimal frequency with the greatest signal to noise ratio.. at least that how its done when calibrating a OEM ecu but then the knock sensor output is fed into a indication system that usually samples at 0.1* CA
Rub87
 
Posts: 2
Joined: Mon Feb 23, 2026 5:36 am


Return to M1 ECUs

Who is online

Users browsing this forum: Google [Bot] and 4 guests