[time-nuts] Stability vs. Accuracy

wje wje at quackers.net
Tue Jul 22 07:03:56 EDT 2008


Simply put, stability is a measure of how much something varies from a 
constant value.
A stable value doesn't mean an accurate value; accuracy is a measure of 
how close to some standard value a device emits (in this case) or 
measures, for a meter.

So, you could have something that's perfectly stable but not accurate. 
You can also have something that's accurate but not stable.

In theory, they can be perfectly (ok, nearly perfectly) accurate because 
their lack of stability is averaged out over a long period of time, and 
the average value is an accurate representation of the time kept by the 
satellites, which themselves provide an accurate representation of 
'true', i.e., NIST time.

In practice they wander around a bit, and that wandering is the 
stability measurement you see. What you can tell from the stability 
figures is that the unit is that close to perfect accuracy most of the 
time. The stability figure is really a measure of the statistical 
probability that at any given time the accuracy is within that bound.

Why do they wander? Many factors contribute, ranging from the stability 
and accuracy of the local VCXO and its control loop to atmospheric 
propagation variations to variations in the satellite clocks themselves, 
etc.

Bill Ezell
----------
They said 'Windows or better'
so I used Linux.



Richard Dabney wrote:
> I'm not a scientist or engineer but have a question to those of you who are.
>
> The many recent posts regarding the GPSDOs and comparisons between the various ones  have been comparing stability. How about accuracy compared with the national frequency standard?
> Are stability and accuracy the same? Stability to E-13.5 with the Thunderbolt. How close to
> perfect time and frequency?............  Thanks...Dick W5UFZ    FMT-nut
>
>
>
>
>       
> _______________________________________________
> time-nuts mailing list -- time-nuts at febo.com
> To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
> and follow the instructions there.
>
>
>   



More information about the time-nuts mailing list