开云体育

Jeff,

when I was young—several decades ago—astrophysicists had error bars that were plus and minus powers of ten: ±10^1, ±10^2. Not for all measurements, but for some. Why? Because it was the best that could be done.

You and I need an appropriate degree of precision for what we are trying to achieve. You'll have seen the recent back-and-forth on the need for Kelvin leads to measure a 10?kΩ precision resistor. And, if you look up the CERN open hardware design for an 8? digit DVM one is suddenly in a world where pretty much nothing is adequate to the task at hand. All those “high quality” test leads which are pretty good for a 3? digit DVM are disasters, Kelvin connected or not, for measuring anything beyond, say, 5? digits due to thermoelectric effects at metal junctions (other deleterious effects may also apply.)

As an Engineer, one exploits technology for the task at hand, sometimes for lowest cost. As a closet physicist (& technologist) one wants to “do better”. Suddenly, you & I can cheaply measure frequency to 1 in 10^9 (or far better) with GNSS (GPS, GLONASS, &c.) technology that was simply unavailable—at low cost—a decade ago. However, modern EMF standards to 1 in 10^9 are only available to those with very deep pockets and a ready supply of liquid Helium (plus a clear understanding of what they are trying to do.) 1 in 10^3, stable & reliable, for pocket money will do very nicely for so many routine purposes, thank you very much.

As for earlier voltage standards, a century ago the Volt was defined according to a standard wet cell. That raises so many issues!

In short:

A: all of us are ignorant in one respect or another

B: precision should always be set according to the task at hand

C: this is—by definition—an open group, catering to all tastes, competencies, and budgets: yours, mine, & extreme metrologists included

73, Stay Safe,

Robin, G8DQX

I’ve learned a lot from this group, but I suspect I don’t really belong here because:

A: I’m pretty ignorant.

B: My definition of precision is clearly nowhere near as precise as the group is designed for.