NavList:
A Community Devoted to the Preservation and Practice of Celestial Navigation and Other Methods of Traditional Wayfinding
Re: Watches as chronometers
From: Brad Morris
Date: 2013 Jun 11, 15:17 -0400
From: Brad Morris
Date: 2013 Jun 11, 15:17 -0400
I have a question for Gary about the trials.
What is the environment that the watches are stored in? Out in your garage or in the house? In other words, did you provide temperature control or just let the temp vary with the season (assuming unheated garage!).
Will you now re-adjust the rate based upon this extended performance? Off only a few seconds over more than 1000 days is only a correction on the order of milliseconds per day. However, having such a significant number of days makes that miniscule correction valid, IMO.
Brad
On Jun 11, 2013 3:03 PM, "Bill B" <billyrem42@earthlink.net> wrote:
On 6/10/2013 6:51 PM, Brad Morris wrote: > The insertion on a leap second is a discontinuity in the standard count > of increments. Let me exacerbate the issue. I will begin my rating > period 1 second before the insertion and end my rating period 1 second > after the insertion of the leap second. TWO seconds have elapsed, but > the standard says THREE seconds have elapsed. Would anyone think that > the rate of their chronometer is 0.5 seconds per second? Now why in the > world would the LENGTH of the rating period adjust your thinking? Well put, reducing the argument to the absurd. Playing devil's advocate, if we average that extra second out over a longer *rating* period, say until before the next leap second, it will diminish the per-day affect. But there is no way around the fact that failure to account for an extra second will introduce a rating error. Be it large or small, it is still an error. On the performance side, the entire leap second added after rating must be backed out of the prediction. Bill B