DIY Knock Ears Recording - Keen For Feedback

Had my VQ37 swap project car on the dyno yesterday and made a knock recording, which I'd be keen to have listened to by some experienced tuners.

To capture this I'm tapping into the two OEM knock sensors (one per V bank) and passing this through Equalizer APO on my laptop which is removing everything but 5.5 - 6.5kHz ... bore size calculation says knock should be at 6kHz.

Data log wise this was a clean run with no knock present; however I'm keen to hear from some folks as to if this is an expected sort of sound from the recording ... as to my untrained ear when listening in real time it sounded like a thousand seagulls screaming into my ears at once :D

For reference also the wiring is as per this link, basically a small cap wired in else its connected directly: https://uprev.com/documentation/Knock%20Listen%20Device.pdf

Here is the recording: https://youtu.be/IBDqXTiVnF0

Hmmm, to be honest that recording is probably not going to be overly useful to you as it's quite 'noisy'. What you really need to do to validate the sound of knock is to purposely induce some light knock so that you can compare what the sound is like. You may find you have better results if you focus on the second order frequency (12 khz +/- 1khz for example). This often produces a better signal to noice ratio. One of the advantages of the professional knock detection systems is that they usually include a range of filter options you can apply which also improve the signal to noise ratio.

David,

I find much of the value comes from comparative analysis of the waveforms. You can slowly sweep through the RPM in neutral, then do a dyno run with a known safe tune, then add timing until you reach MBT or the waveform amplitude significantly changes in an area. Like Andre said, the second order frequency generally gives a signal that's a bit less dirty to work with.