> Hi Tom: > > The plot at http://www.thinksrs.com/products/PRS10.htm showing GPS and > Rb indicates that GPS for times less than say 500 seconds has more noise > than the Rb source. So a voltage ramped 1000B should be very clean > below 500 seconds.
I measured a 1000B a few weeks ago. See: http://www.leapsecond.com/museum/datum-1000b/ I agree about the 500 second figure. A normal GPS receiver is noisy below 500 s. The PRS10 also has rather poor short-term performance. So a 1000B will be better than either of them by one or two orders of magnitude. Still, I'm not convinced about your voltage ramp. How will you know what voltage to generate at any particular moment? A long-term ageing rate doesn't necessarily have anything to do with short-term or mid-term performance. Sometimes an oscillator will be a picture perfect example of linear or logarithmic drift, but not always. See also: Integral Crystal Oscillator Testing http://metrologyforum.tm.agilent.com/xtal-policy.shtml Lastly, see the section "Short-term Noise Contrast" at the very end of http://www.leapsecond.com/museum/datum-1000b/ and you'll get a very graphic view of how quiet a 1000B is and how noisey a PRS10 is. /tvb _______________________________________________ time-nuts mailing list time-nuts@febo.com https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts