After building an enclosure for the board and allowing it time to settle, the clock stability is remarkably improved over the previous revisions. Here's the settled performance over 12 hours:
The enclosure slows the temperature ramps sufficiently enough that I've been able to relax the proportional coefficient in the PID controller which makes the VCTCXO adjustments far smoother.
Discussions
Become a Hackaday.io Member
Create an account to leave a comment. Already have an account? Log In.