[time-nuts] need example frequency vs temp equation

Jim Lux jimlux at earthlink.net
Thu Oct 13 00:12:44 UTC 2011


I'm putting together some examples of oscillator behavior vs temperature 
and I'm looking for some plausible coefficients and simple equations to 
use to generate nice looking curves.

oscillators are TCXO and run of the mill whatever they use for computer 
clock oscillators (AT cut?).

Something that gives me decent PPM vs degrees C for a range from cold 
(-50 to -40) to hot (say +70 to +80)

I've got tons of measured data, but before I spent the time to try and 
do a curve fit (implying that I can actually read the data and not have 
to copy it by hand from a table or translate some oddball log file)

I've got Vig's tutorial: If you look at page 2-7, it's those curves I 
want to generate.

(or if you look at 4-43 in the tutorial)

Maybe the Army book by Bennett?

Even better if I can replicate the hysteresis with a simple model.

I've got Frerking's equation
delta F/F = A1* deltaT + A2*deltaT^2 + A3*deltaT^3

deltaT = T - 25C

but what's some good numbers for A1, A2, and A3..  I found one reference 
that cites Gerber and Ballato
A1 = -5.08E-6 * angle of cut relative to 35.25 degrees
A2 = -0.45E-9
A3 = 108.6E-12


that gets me a generic AT cut..

But what about TCXOs (which have more lumps in the curve...)





More information about the time-nuts mailing list