PDA

View Full Version : knock frequency formula



Igor
07-29-2004, 02:57 AM
I've tried to find one on the web but came up empty. Does anyone of you know of a formula that will calculate the fundamental frequency of the knock signal based on boresize (and maybe something else)?
I want to build a knock monitor based on a tuned filter, but we don't have an engine on the dyno at the moment and I don't wish to pay Cosworth the big bucks to cough up the numbers.

Igor

Igor
07-29-2004, 02:57 AM
I've tried to find one on the web but came up empty. Does anyone of you know of a formula that will calculate the fundamental frequency of the knock signal based on boresize (and maybe something else)?
I want to build a knock monitor based on a tuned filter, but we don't have an engine on the dyno at the moment and I don't wish to pay Cosworth the big bucks to cough up the numbers.

Igor

John Bucknell
08-01-2004, 04:07 PM
Knock frequency is actually pretty easy. Take a stab at speed of sound at the in-cylinder densities when knock occurrs (ie near peak cylinder pressure). The frequency will be the number of times that sound wave bounces back and forth across the bore (ie 1 cycle is twice bore).

For reference a 77mm bore is like 14 kHz. However, this only works for in-cylinder pressure transducers. Your block will resonate at a different frequency (usually quite a bit lower) if you intend to use a knock sensor. And that is so complicated it is usually determined experimentally.

Igor
08-01-2004, 11:32 PM
Thanks John,
Interesting that what you measure on the inside doesn't match what you hear on the outside. You learn something new here everyday.
Well, I guess I'll just slap on a scope when we've got something on the dyno.
If you see clouds of smoke coming out of the Spyker at Le Mans next year, you'll know it didn work :-)

Igor