OK, got the papers. I'm still trying to figure out if a minimum resolution can be determined for the TI measurement OCXO_PPS/GPS_PPS. That is, if the GPS has a 25nS random wondering, is it worth having 25pS measurement resolution? At first yes, I can use the negative sawtooth but I'm not convinced. Then, I'm using a simple average over 100 or 1000 seconds but my OCXO moves menawhile: the simple average is not the ideal tool and I'm still investigating what magical statistic can I use.
On Tue, Oct 2, 2012 at 3:39 PM, Tom Van Baak <[email protected]> wrote: > Azelio, > > In frequency average mode the 53132A makes something like 200,000 > measurements per second, according to the manual. This is why the > instrument does so well as a smoothed frequency counter. The resolution > "gain" is some fraction of sqrt(200000), subject to the normal caveats like > the need for independent samples. That's why these counters have reduced > resolution when the input frequency is too close a multiple/fraction of the > timebase. > > Details on this technique (used also in the Pendulum CNT-9x counters): > > "New frequency counting principle improves resolution" > http://tycho.usno.navy.mil/ptti/ptti2005/paper67.pdf > > See also this fine presentation by Staffan Johansson: > http://www.npl.co.uk/upload/pdf/20060209_t-f_johansson_1.pdf > > Or, google for: continuous time-stamping frequency counter > > /tvb > > ----- Original Message ----- > From: "Azelio Boriani" <[email protected]> > To: "Discussion of precise time and frequency measurement" < > [email protected]> > Sent: Tuesday, October 02, 2012 5:30 AM > Subject: Re: [time-nuts] Best counter setting for ADEV? > > > > Hmm, the frequency mode sigma is the same as the TI mode sigma for > 200*tau. > > That is if I take the sigma at 200*tau from the TI mode I have the same > > figure of the frequency mode. Maybe the frequency mode 2 seconds sample > > time uses 100 averages per second. > > > > _______________________________________________ > time-nuts mailing list -- [email protected] > To unsubscribe, go to > https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts > and follow the instructions there. > _______________________________________________ time-nuts mailing list -- [email protected] To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
