VXOs, the ability to vary the rock is directly tied to the fixed C (holder and crystal)
and can be expressed by PPM so naturally higher you go the more you can pull
a rock expressed in Khz.
The big problem is for 100 crystals from 100 lots even from the same vendor
that variability will be individual to the each crystal. That's never minding how
close they are to the nominal frequency.
Since the SI570 and SI5351 I stopped bothering with rocks for variable oscillators.
For filters The criteria is different and matched is the first factor. The frequency
for a given bandwidth broadly determines the termination impedance as well
as widest attainable bandwidth with good shape factor. I've found for SSB
work 5 crystals are at the low end of acceptable and shave several radios
where the filter is a cascade of 7 or as many as 10 to get decent RX skirts
at SSB widths. CW 4 will do for a 500hz filter. Lots of Superhet radios
over the years so lots of filters built.
The yabut is that using Tayloe base band converter and SI5351 an analog
phasing design works well and solved the crystal question totally. Opamps
are cheap and a 3rd order all-pass is cheaper and easily reproduced.
Allison
--
Please reply on the group, no private emails we as a group share info that way.
