On Fri, Oct 5, 2018 at 5:08 AM AC0XU (Jim) <[email protected]> wrote:
> [...] Since it am especially interested in measuring short-term stability, > I am going to try using it the method described in (who??): > > http://www.efos3.com/53230A/HPAK53230A-High-speed-TI.html Thats my website. The method can be a little fiddly if you don't understand the instrument pretty well, I would suggest getting a bog-standard TI measurement working first. In either case I'll be happy to assist if you need some pointers, ofcourse. > I understand that at some point I will be seeing the noise/stability floor > of the instrument. According to > > http://www.anderswallin.net/2015/04/keysight-53230a-noise-floor-test/ > > that noise floor may be around 2e-11 at 0.1 sec and 2e-12 at 1 sec. [...] > Well.. I trust I will be corrected if I am wrong, but I believe the frequency measurements were taken using the "resolution enhanced mode" - which does not yield the correct ADEV, as shown by the slope on the plots. See here http://www.anderswallin.net/tag/53230a/ for a followup on that post by Anders. A 20ps counter should give a noise floor around 2e-11 at 1 second gate time, with a slope of -1 on the log-plot; 2e-11 at 1 second, 2e-12 at 10 seconds, 2e-13 at 100 seconds and so forth, untill the noise is no longer white PM. Theres a lot more in i.e. Handbook of Frequency Stability Analysis ( https://tf.nist.gov/general/pdf/2220.pdf) and other places - http://www.leapsecond.com/time-nuts.htm is a good place to start, if you have not seen it. Ole _______________________________________________ time-nuts mailing list -- [email protected] To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com and follow the instructions there.
