Knock sensors

[Frank F Parker]

> I thought I'd get a cheap equalizer (i.e. one that's lying
> around -- free) and use the above-mentioned RatShack audio
> amp to listen in on the engine whilst tuning out the stuff
> I don't want to hear -- until I read Probst's book.
> 
> "some day", but now we're getting serious.  I don't want
> to drive around with a spectrum analyzer lying on my floor
> board -- I can whip up a notch filter and drive an analog
> meter (or, for that matter, an AGC circuit that has a lower
> frequency into it so that I can *hear* some feedback).
> 
> Sorry for the rambling .... has anyone done anything like
> this.  Todd Knighton, and some others, keep knocking me (pun
> intended) to get the commercially available knock detector
> (don't remember who by at the moment  ....  J&S  ?? ) -- but it's
> like $400, as I recall .....
> Tom Cloud

Tom,
 I have done what you thought about with an equalizer and knock sensor
and it was not much better than just listening to bare sensor.
Todd and the others are correct-simple analog filters just do not
cut it- that is why the early knock systems were so sensitive to
location etc. The J&S Electronics unit uses DSP filtering so you are able
to filter freq that are separate but close together- impossible even
with multiple pole analog filters. Best the J&S has a 0-1.25 volt
monitor out so you can drive a LM3914 for a retard display. I also ran
it to a datalogger and took data at 50/sec along with various pressures
and temps on a recent turbo project car.

This is one of those cases where there is not a cheap way to monitor
and control knock- gotta pay your bucks!- Even then location of sensor
is important to the best performance.

Regards,

Frank Parker
fparker at umich.edu
Motorsports Data Acquisition
> 
>