Here's my REF01 in a box. I set the trimpot a few years ago and just pulled it out of a drawer and powered it up.
I should also measure their 50 ohm input impedance stability. Somehow.
That must be the A grade. Back in the day they came in grades all the way up to G, iirc. (The cheapest was always referred to as "crap grade". Amused the sales reps no end.)
The other nice thing about the REF-01 (of which I have probably 10 in my lab drawer) is that it has a reasonably-accurate PTAT voltage output. I used to use it to temperature-compensate other things, back in the pre-MCU days.
Cheers
Phil Hobbs
Nice.
Temperature, time or both?
What's your expectation? I guess I'd not expect much better than
+/-100ppm/degree C, maybe worse.
My customer wants about 1000 PPM noise+stability on some fast pulses driving an e/o modulator, measured over 8 hours. So we need better for the measurement gear. His temperature spec is +- 0.2 C! We may have to put the DUT and the oscilloscope in a temp chamber. Nuisance.
We also have a jitter + drift spec of 5 ps RMS for our fairly complex box, also 8 hours. That may be harder to measure than amplitude. No oscilloscope that we can afford would have that sort of performance trigger-to-vertical, but we might get it vertical channel-to-channel, with some math. Even cables will be a problem.
It's an interesting math problem to slurp a list of points from a sampling scope
time1 vert1 vert2 time2 vert1 vert2
etc, and compute RMS jitter between channels. A sampler can take the vertical measurements simultaneously, but only one pair per trigger.
Time drift is easy to compute.
Jitter is easier to compute if it's a modern digitizing scope, but the old samplers are a whale cheaper per GHz. Like 2000:1 or something.
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.