Recently, I have been experimenting with synchronizing my stopwatch to UTC, using one Anthony G. Atkielski's WWV simulator for the PC, wwv3.exe, with very good results. It is available via the following page, listed as "WWV simulator (Windows) Version II BETA": http://www.atkielski.com/main/ Direct link to the program: http://www.atkielski.com/main/data/WWV3.exe Last November, Anthony described his programs, and provided some tips on accurately setting a PC's clock, which is the prerequisite to their successful use: http://satobs.org/seesat/Nov-2011/0216.html I have synchronized my PC's clock using Alexander Panchenko's SP TimeSync for many years, with good results - most recently version 2.4, available here: http://www.spdialer.com/download.html I set SP TimeSync to re-synch every 10 minutes, which on my PC has proven adequate to maintain synch to within about 0.01 s. I offer the following description of the method I use to synchronize my stopwatch using WWV, and a discussion of my experience with wwv3.exe as a substitute. My stopwatch can record up to 200 time splits. I synchronize it to UTC by starting the timer at an arbitrary time, and then taking a series of 15 to 20 splits at the audible ticks broadcast by WWV at the start of all but the 29th and 59th second. When the signal is strong and free of interference, I can establish a rhythm with the ticks, yielding splits that consistently agree to within a few hundredths of a second. I take the final split at a specific UTC time, usually 5 s past the minute. I do not take it at the minute, because WWV omits the time tick of the 59th second, and switches from a tick to loud tone at the minute, which could throw off my rhythm. Subtracting the final split time from the UTC time at which it was taken, yields the UTC time that the stopwatch was started, but since any given split may be off by up to a few hundredths of a second, I first seek to verify, and if necessary correct the final split time. I do so by taking the average of the aforementioned series of splits, which I judge to be the probable fraction of a second of my final split. For example, let us say that my final split was taken at 07:38:05 UTC, and reads +00:23:09.45. I visually inspect the prior series of splits to verify the accuracy of the fraction of second, and adjust it as necessary. For example, if the series averaged 0.42 s, then I would take my final split to be +00:23:09.42. Subtracting that from 07:38:05 UTC yields 07:14:55:58 UTC as T0 - the arbitrary time I started the timer, enabling me to compute the UTC time of all subsequent splits. This method has proven accurate, but my shortwave reception typically is poor, so I cannot rely on having a usable signal when I need it. The alternative has been to dial up the telephone version, but it frequently suffers from significant delays, sometimes long ones caused by routing via satellite. As a result, I found it necessary to dial in at least three times, in an attempt to ensure accuracy. Since my PC's clock is accurately synced to UTC with SP TimeSync 2.4, I experimented with synching my stopwatch by taking a series of splits against the changing seconds displayed by the PC, but I found it impossible to achieve adequate precision. In my experience, hand to ear coordination is much more accurate than hand to eye, which is why I was interested in Anthony's WWV simulator. My recent tests of wwv3.exe, in combination with SP TimeSync 2.4, consistently yielded synchronized time to within a few hundredths of a second of WWV's telephone signal. In several of the tests, WWV's telephone signal was late by at least 0.2 s - the delay was clearly audible when listening to both signals at the same time. In one test, I had to dial in to WWV three times before I received a signal that was not significantly delayed. On a PC with an accurately set clock, wwv3.exe appears to be a very good substitute for the actual WWV signal. I am going to continue to use wwv3.exe and WWV side by side, and I will report any problems I encounter with the former. Observers who wish to try wwv3.exe on a PC with an accurately set clock, should verify its accuracy against WWV or similar time source before relying upon it for observations. Ted Molczan _______________________________________________ Seesat-l mailing list http://mailman.satobs.org/mailman/listinfo/seesat-l
This archive was generated by hypermail 2b29 : Fri Sep 28 2012 - 12:41:59 UTC