I will soon replace my 1965 Nova's 194 ci engine with a reworked 250 bored 0.030 over (3.905). The head is from the original 194 with combustion chambers opened to 66cc. The deck has been milled to provide quench (squish) of 0.041 (thickness of the head gasket). The dished pistons give 9.3:1 static compression ratio, and a mild Crower cam produces a Dynamic CR of 8.825:1.
I plan TBI via Megasquirt. A
Knock Sensor will be used because of the rather high CR. Since that was never part of a pre-70's engine, I've been doing a lot of research, but still have questions.
The
Knock Sensor will go in the block drain, unless I am convinced otherwise. The rationale is; that's where GM does it on their V8 engines (until recent high performance ones). Also, I was told by my machinist, that the block drain works well because the Sonar effect transmits
knock impulses from any and all cylinders.
As I understand it, the
knock frequency is a function of bore size. I believe a 3.905 bore should put the frequency at 5700 Hz using
http://www.phormula.co.uk/KnockCalculator.aspx . Results of this link match my hand done calculations exactly, and it is very easy to use.
Based on that calculation, I searched for a matching sensor that would screw into the NPT drain. I found a sensor that resonates between 4800 5600 Hz, and another from 5600 6400 Hz and others that are too far from 5700 Hz to be considered.
The 4800-5600 sensor fits 1996-1999 Chevrolet C2500 Suburban 5.7 L V8 (and others) 4.0 bore.
The 5600-6400 sensor fits 1996-2005 Buick Park Avenue 3.8 L V6 (and others) 3.8 bore.
My first reaction is to select the 5600-6400 unit because 5700 is within that range. However, my bore is almost exactly half way between the Chevy and the Buick, so I'm not 100% sold. Notice how the bigger the bore, the lower the frequency (like an organ pipe).
My questions are: 1) Which of the two sensors have others used successfully, or perhaps another one I didn't find. 2) If someone has a better location than the drain plug hole, where did you mount it. Naturally, I welcome any comments, especially from someone who has done something similar and discovered what works and what doesn't.