[EMAIL PROTECTED] wrote: > > > Hi Bruce, > > My math is essentially their reference [4, p.26], and I did state that I > don't know the exact interpolator, trigger, and frequency-dependent noise > functions of the Wavecrest. > > DTS-2075 is superior to the SR-620 and 53132A in time interval measurement, > especially for single-shot measurements (parameters from manufacturers data > sheets): > > DTS-2075: > > +-25ps single shot accuracy, 3-sigma > 800 femtoseconds resolution with 1 femtosecond LSB display > > SR-620: > > Relative errors: > <±(50 ps typ. [100 ps max.] + > Timebase Error × Interval) > > Absolute errors: > > <±(500 ps typ. [1 ns max.] + > Timebase Error × Interval + > Trigger Error) > > 25ps single shot resolution (31 times worse than DTS) > > 53132A: > > 150ps resolution, at 100ps uncertainty (RMS) (187 times worse resolution > than DTS units) > > bye, > Said >
Said Your claim of 1E-13 resolution in 1 sec is not sustainable. The counter has no better resolution in measuring the Allan Variance of a source than any other counter with a single shot resolution of 25ps. If you read the paper I mentioned it should be clear that averaging doesn't improve the resolution when resolution when measuring the Allan variance of a source. Although it may have better differential and integral linearity specs. However they appear to give no specs for thes linearity. The claimed 800fs resolution after averaging is irrelevant for this type of measurement. The Wavecrest instrument only has a single shot resolution of 25ps not as you claim 800fs (this is only achieved after averaging) The above counters can only be compared for this purpose if their single shot resolution and linearity are known. Bruce _______________________________________________ time-nuts mailing list time-nuts@febo.com https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts