Over the years I’ve acquired a whole bunch of Sulzer OCXOs. One is dead, several have oven temperature control problems, but I have three powered up in the rack that still work amazingly well.
This one, Sulzer #3, is particularly interesting. It’s not on the standard 2.5 or 5 MHz frequency. Instead it runs at 5.006 880 MHz. That happens to be a submultiple of the Cesium transition frequency. And I’ve been told, but unfortunately can’t prove, that this unit came out of NBS in Boulder. Hmmm…
I measured it for 2 hours against an OSA 8607-008 BVA, and for 10 days against an HP5065A to get an idea of aging. Sulzer #3 has been running uninterruptedly for just about 4 years at this point.
That’s not too shabby.
I also did a phase noise measurement against both a Wenzel ULN and the BVA.
Note the steep drop at about 500 Hz. I am guessing there is a crystal filter in the unit to drop spurs. -115 dBc/Hz at a 1 Hertz offset isn’t too shabby; in fact, that’s about what both the ULN and BVA are rated for, so the chances are the actual value is a bit better. The noise floor isn’t all that hot, but let’s give 1960-era designers a break!
I’d really love to know if this oscillator was involved with NBS-2 or NBS-3.