[R-390] Crystal substitution puzzle

[Tim Shoppa]

Bill writes:
> Seems to me that crystals have a rated socket capacitance. The
> newer crystals could be rated for lower capacitance. OTOH, 0.01%
> seems like a lot of pull for a crystal.

A few kHz change in resonance is entirely typical for a
16MHz crystal being loaded differently than specced.
So if it was supposed to be series resonant and you put in
a parallel resonant crystal, a few kHz shift is expected.

The sub-dollar crystals they produce by the bazillions are
only out of factory spec by a rate of less than one out of a hundred.
I know, because I've bought them by the hundreds and sorted through
them for homebrew crystal filters in the past few years, and only once
did I see one that was out of spec.

Example: Of a lot of 200 14MHz crystals I sorted through, 90% of
them had their series resonant frequency matched to +/-70 Hz,
which is +/-5ppm.

That does not mean they are all identical. Different brands and different
lots from the same brand will have non-specified parameters (Q; motional
Capacitance; motional Inductance; Series resistance; etc.) that are often
completely unspecified (and these in large part determine the pullability
aka difference between series and parallel resonance etc.)

Tim N3QE