Previously, I blogged about the Orion's Frequency Calibration. I was able to tweak the Orion's TCXO to within a couple of Hz at 15 or 20 MHz, after warmup. The ambient room temperature was 80 F.
So now, after about 3 months, I recheck and find the oscillator has drifted about 14 Hz low (dial readings 14 Hz high) at 15 MHz. The room ambient is 73 F. This is still within +/- 1 ppm, and the oscillator is spec'd at +/- 3 ppm over 0 - 50 C.
The upshot: just because you can adjust the oscillator to +/- 0.1 ppm with care doesn't mean it will hold up for too long, as the oscillator ages and ambient conditions change.
Note: I'm getting ready for the ARRL Frequency Measurement Test tomorrow. The Orion's slow drift doesn't really matter. I will have to measure the the apparent WWV frequency anyway and apply the correction to the measurement. It doesn't matter how closely I can set the TCXO, I still have to do the correction.