Sulzer 5A #494I own three Sulzer quartz crystal frequency standards. When these were manufactured in the early 1960s, they were the best commercial standards in the world. Today, 40 years later, they can still be awfully good.
I bought this unit off eBay and when received it had a problem -- the circuitry that controls the inner and outer ovens had aged out of whack, and the heaters were never shutting off -- they just kept pumping heat into the crystal. I opened the unit up and did some surgery which brought the ovens at least back to nominal function, though I have no idea if the actual crystal temperature is what it's supposed to be.
Here are the results of a 60 hour measurement run on Sulzer 5A serial number 494. My working theory for right now is that the sinusoidal noise visible on the "adjusted phase" plot is due to temperature changes in my basement lab. Although the basement keeps a fairly constant temperature in the summer and winter, this data was taken in October, when the temperatures were changing and heat was soaking out of the building. In addition, the air conditioning was running frequently (we were doing interior house painting at the time) and as a result the temperature varied by about 3.5 degrees during the run.
The noise in the measurement system was about 100 picoseconds, so stability readings below 200 seconds or so are not meaningful.