[time-nuts] Voltage standards

[Bruce Griffiths]

Bill Hawkins wrote:
> Group,
>
> The subject elicited a fine hardware discussion, as always.
>
> There was some casual talk about measuring fractions of
> microvolts, implying 10E-7 or -8 accuracy.
>
> There are lots of possible thermoelectric effects that may be
> in series with the source.
>
> Without describing how it's done, what is the lower measuring
> limit in nanovolts? That's the limit where the last digit
> moves randomly by more than one increment. Does the method
> involve cryogenics?
>
> I'd look it up, but I expect some of you are more current
> with the science than anything I could find with Google.
>
> Given that extreme accuracy and stability are expensive, what
> can an amateur do for less than, say, $500 or 400 Euros?
>
> Thanks for any enlightenment.
> Bill Hawkins
>
>   
Bill

The lower limit depends on the detector or DVM used, together with ones
skill in eliminating thermoelectric effects.
With a sufficiently low noise front end detecting changes of 10nV or
less at room temperature is possible if the sources being compared are
sufficiently stable.
An ordinary DVM wont suffice (at least without a low noise preamp).

An LTZ1000 standard is feasible for for around $100, however whether you
get a really good one or one that just mets its specs depends on luck.

Building a null detector with say 10 nanovolt sensitivity is also possible.
Picovolt sensitivity usually requires cryogenic cooling.

With luck you may even be able to obtain an Fluke 720 KVD.
Failing that it is possible to build a 20 bit monotonic DAC with ppm
stability without requiring precision resistors.

Cryogenic standards arent really feasible for home use as most require
liquid helium coolant.

However graphene has supposedly exhibited the quantum Hall effect at
room temperature.

Bruce