|
Hi all,
I've been experimenting with high stability crystal oscillators for methods
of fine tuning by external electronic means. Varicap diodes are nonlinear
for voltage and temperature, so I was wanting to use some other way.
Crystal oscillators usually have coarse adjustment with an NPO trimmer
capacitor, so fine tuning would be acceptable even if it was less than 1 ppm in
adjustment range (sufficient to correct for ageing over a fairly long
period).
What I came up with is variation of emitter load resistance in a Colpitts
crystal oscillator. 20% or so of change in emitter resistance gives smooth
control of fractional ppm change in oscillator frequency. I have been
using a fixed resistor to the emitter that is ten times the installed emitter
resistor, and then a pot (rheostat) to ground also of ten times the emitter
resistor. This arrangement gave 0.4 ppm range for fine tuning with
the home brew 10 MHz oven oscillator being tested. There are minor
effects with thermal equilibrium when emitter current changes, so if a
precise setting is needed then more than one tweaks is needed, but it is
still quite a convenient way of carrying out fine tuning with a panel mounted
control. The lead can be bypassed at the oscillator so the control is
"cold" at RF. Also the control can be a floating with earth
return to the oscillator circuit to minimise a ground loop influencing the
oscillator. Probably a wirewound is the best choice for the type
of pot.
Some types of oven crystal oscillators on the surplus market do not cater
for fine tuning by electronic means (only have screwdriver access to a
trimmer capacitor). If the oscillator housing can be opened up is is
fairly easy to implement a "delta emitter" wire for fine
tuning. Sometimes it is frustrating to adjust the capacitor as the
only means of making a precise setting, as there is inevitable backlash effects,
whereas the "delta emitter" method gives smooth fractional ppm
adjustment (with due regard for temporary thermal equilibrium).
73, Bob ZL2CA
|
