[time-nuts] Expected 10 MHz offset from a GPSDO?
kb8tq at n1k.org
Sat Aug 20 11:14:58 EDT 2016
At the most basic level, ADEV is a standard deviation. When people read an instrument they usually want a "max error" sort of number. ADEV operates on delta between readings so a simple "three sigma" mental math conversion is not all you need. It *is* the right measure. It can take a bit of work to convert it to useful units .
> On Aug 20, 2016, at 7:38 AM, Charles Steinmetz <csteinmetz at yandex.com> wrote:
> Bryan wrote:
>> With a GPSDO as a general rule is there much variation of the 0.005 to 0.02 ppb range. Would there me much if only 2 satellites were visible but the unit was no in holdover vs say 6 satellites. I assume there is probably a lot of variables such as quality of GPSDO and OCXO, multi path interference, antenna location, etc, but worse case what would a user expect assuming the unit isn't in holdover.
> There are lots of ADEV plots of GPSDOs on the web, if you search for them. Tom has the attached plot posted on his web site -- it includes one decent (but not stellar) GPSDO, the HP Z3801. That is about what you can expect without heroic tuning efforts (on a GPSDO that supports user fiddling, such as the Trimble Thunderbolt).
> Since many counters gate for one second (or gate faster, then average), the ADEV at tau=1 second is a reasonable estimate of the real-world performance of a good counter driven by a GPSDO. Note that, since the ADEV at 1 second depends largely on the OCXO, the heroic tuning efforts mentioned above won't change the ADEV at tau=1 second hardly at all. That is attributable strictly to the OCXO (unless the GPSDO is very poorly designed). Building a GPSDO using the best OCXO available can improve things. For example, look at the "BVA" plot -- at 1 second, its ADEV is nearly a full decade better than the Z3801, better even than the hydrogen masers. (While I referred to the 3801 above as "decent but not stellar," that is based on its performance at tau from 100-100k seconds. Its OCXO -- and, thus, its ADEV -- is better than most others at 1 second.)
> Best regards,
> 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