[time-nuts] Re: Why do have OCXO a Vref output?

[Tom Van Baak]

Attila,

 > Why do have OCXO a Vref output in the first place?

The purpose of Vref is to optionally allow the oscillator to be 
fine-tuned using 3 wires to an external user-supplied multi-turn pot, 
often hiding behind an access hole at the rear of the instrument.

Note that the internal Vref needs to be good enough not to degrade the 
performance of the crystal oscillator, but not much better.

For example, if your OCXO is stable to 5e-10/day and has a tuning range 
of 1e-7 with an EFC range of -5 to +5 volts then 1 mV stability on Vref 
translates to 1e-7 * 1e-4 = 1e-11 in frequency, far more than adequate 
(50x) for the OCXO. Check my math.

By contrast you would not want a cheap Vref that drifted by, say, 50 
mV/day because then half the drift of the oscillator might be due to 
Vref and half due to the crystal; a waste of a good crystal.

I'll let volt nuts run the actual numbers for Vref that are found inside 
various OCXO. But it sounds to me like a Zener would be fine for almost 
any OCXO that you own, especially since it's inside an ovenized box. 
Using a metrology-grade voltage reference would be a waste of a good 
reference.

Here's how to get actual data: disconnect Vref, ground EFC, and then 
measure both Vref voltage and Fout frequency once an hour for a few 
weeks and compare their normalized daily drift rates.

/tvb