Hi Magnus, I've never seen it in print either but I coined the name when I wrote my ADEV tools. I've found a profound difference in the overlapping form of ADEV in many cases and so I use it a lot now -- especially when analyzing the effect of tides on precision pendulum clocks, or other periodic effects, such as cycling A/C or diurnal effects.
However almost every technical paper in the past few decades uses the plain old textbook non-overlapping back-to-back ADEV so my software tools call the traditional calculation "ADEV" and I call the overlapped version "OADEV". An alternative was to use words "ADEV(normal)", ADEV(overlapped)" and "ADEV(modified)" but I chose the shorter ADEV, OADEV, and MDEV instead. "ADEV" and "MDEV" are already standard, and so "OADEV" seemed to fit. So far no one has been confused, but I can see your point. See the tool page, which includes source code: http://www.leapsecond.com/tools/adev1.htm /tvb ----- Original Message ----- From: "Magnus Danielson" <[email protected]> To: "Discussion of precise time and frequency measurement" <[email protected]> Sent: Wednesday, January 21, 2009 12:32 AM Subject: [time-nuts] ADEV vs. OADEV Hi! I have been quite surprised to see the abbreviation OADEV appear. I assume that this means Overlapped Allan Deviation, but this is confusing since the Allan Deviation estimates already is overlapping. However, I have seen that some use a non-overlapping estimator, but this type of estimator has an unwanted filtering effect and should not be used. If a distinction between these ADEV estimators should be used, then the standard (overlapping) estimator should continue to be called ADEV and the non-overlapping (back-to-back) could be called NOADEV ór whatever... I have done a fair amount of digging around many sources around ADEV and friends so I think I got it right. Unless someone can give a meaningful explanation and I really expect a good article detailing the difference and benefits... Cheers, Magnus _______________________________________________ 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.
