[time-nuts] OCXO burn-in
denwood at orcon.net.nz
Thu Sep 15 05:41:29 UTC 2011
Whenever you make changes to the physical setup you will cause mechanical
disturbance to the OCXO which it is likely to take some time to recover
from. A couple of days continuous operating should suffice. Add to that an
hour for every day it's been switched off.
Even tipping the unit upside down changes the G force on the crystal and
affects the frequency. Bumping it could change it for a couple of days.
As for the environment of the OCXO, try to place it where you don't affect
the natural external temperature of the OCXO. Preferanbly don't elevate the
case temperature too much, but definitely don't lag the OCXO, put it in a
Dewar or subject it to draughts (e.g. air from fan). Respect the need for
the OCXO to maintain the internal temperature profile it was calibrated at.
The RAKON/C-MAC OCXOs, especially the CFPO-DO and the single-oven types in
the typical GPSDO have pretty good ambient compensation, but attempts to lag
or cool the OCXO won't be helpful as the thermal profile within the oven
Anybody contemplating making their own high performance oven oscillator
might as well forget it. The design of the crystal, thermal environment and
oven control are all very tightly coupled. These days really good OCXOs,
better than anything you could make, can be had at sensible prices.
More information about the time-nuts