Ed, I make sequential measurements as well and create the 3-hat from the calculated ADEV values. It's ok to do it that way. I have heard that even using simultaneous measurements you can get negatives, but I have not verified this. It usually means it's too close to call or you data runs are not as clean as they could be.
Make sure to look carefully at your phase and frequency plots before you blindly compute ADEV. Or do multiple runs each and pick a set that's well behaved. Or cross your fingers and use the absolute value trick as you have. You'll notice that's what I use in http://leapsecond.com/tools/3hat.c /tvb ----- Original Message ----- From: "Ed Palmer" <ed_pal...@sasktel.net> To: <time-nuts@febo.com> Sent: Wednesday, May 01, 2013 1:09 PM Subject: Re: [time-nuts] three cornered comparison tools >I recently made some measurements between 3 oscillators. It wasn't a > true 'Three-Cornered Hat' measurement because the measurements were made > sequentially. When I do the three-cornered hat calculation for the > hopefully 'better' oscillator, I end up trying to take the square root > of a negative number. Is that a red flag that the data is invalid? If > I ignore the minus sign, the results seem reasonable and I can > successfully calculate from the result for each oscillator back to the > measured results. > > Ed _______________________________________________ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
