TVB noted: If you are using the 10811 as part of a GPSDO the PLL should take care of any number of slow-moving changes in frequency; whether it's temperature, humidity, voltage, OCXO ageing, DAC drift, phase-of-the-moon, etc.
So I don't see a compelling need to "protect" the OCXO in the ways being proposed. If you are in a harsh, or fast-changing environment then do the math to see if your OCXO dF/dt exceeds what the PLL can close relative to the dF/dt of the GPS reference. There are several elements of the GPSDO that can have sever temperature sensitivity: 1. The GPS receiver itself. Of particular concern is the group delay thru the ~2 MHz wide IF filters and also the several MHz wide filters in the RF front-end. The IF filters are often SAW (Surface Acoustic Wave) devices, and the RF bandpass filters are frequently coaxial ceramic devices (functionally similar to tuned cavities, but built inside high dielectric constant ceramic). I have seen tempco's of as much as a few nsec/ºC for some brands of receiver boards. 2. The GPS antenna. Most systems employ patch antennas which are manufactured with ceramic dielectric loading of the patch elements. Since these are located outdoors, they can see several tens ºC temperature swing throughout a day. 3. Most GPSDOs use a fairly long divider to bring the standard oscillator down from 10 MHz to 1 PPS. If you use ripple counters for these dividers, they will show a significant temperature sensitivity. I once used a string of 3½ 74HC390's to bring 10 MHz down to 1PPS and was appalled to find several hundred nsec of delay change with ambient temperature in a trailer. But when you realize that 7 BCD counters with 4 flip-flops each account for something like 72 CMOS gates, each with a tempco ~1 ns/ºC, the results were not surprising. The moral to this story is that the counters in your GPSDO should be synchronous counters. Better still, I found that TVB's single chip PIC divider had unmeasurable temperature effects. I would be surprised if you find a significant humidity or pressure pull unless you have a problem with moisture condensing on some component; the dielectric constant of liquid water (or ice) differs significantly from unity! 73, Tom _______________________________________________ time-nuts mailing list time-nuts@febo.com https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts