[time-nuts] Relationship of frequency to stability
jmfranke at cox.net
Wed Jun 7 11:33:14 EDT 2017
There is no direct relationship between frequency and stability. For any given frequency, the stability is determined to a large extent by the technology used for the oscillator. For instance, at 1 kHz, a quartz crystal is normally more stable than a tuning fork which is normally better than a LC circuit which is normally better than a RC circuit. And even amongst a set of oscillators the stability is determined by the stability of the individual components, the design of the oscillator, and environment factors. There have been pendulum and tuning fork clocks better than some quartz crystal clocks. Quartz crystal clocks eventually became more stable than free pendulum clocks which were limited by gravitational variations. Some quartz oscillators are better than ammonia atomic clocks. I have tuning forks at 1 kHz that are more stable than some crystal oscillators and microwave klystrons. Stability is more related to the quality of the parts, the design of the oscillators the base technology and the control of environmental factors than to the frequency. Examples can be found where an oscillator based on older technology outperforms a newer potentially more stable technology through attention to detail in the construction and operation.
Hope this helps!
> On June 7, 2017 at 10:25 AM John Sloan <jsloan at diag.com> wrote:
> SHORT VERSION: I’ve been trying to figure out what the relationship is between frequency and stability - if there is one - that is, why oscillators with higher frequencies tend to be more stable.
> LONG VERSION: I got into this by building a home-brew NTP server using a cesium-disciplined oscillator, specifically a Jackson Labs GPS-disciplined oscillator board that incorporates a Microsemi Chip Scale Atomic Clock. I started thinking about the 9,192,631,770Hz oscillator in the CSAC, the quartz oscillator in my $10 Casio wristwatch that is perhaps 32768 Hz, the 8 Hz balance wheel in my Rolex GMT Master II, the 6 Hz balance wheel in my Hamilton Jazzmaster, and the maybe 5 Hz of some of my less expensive mechanical wristwatches. In my personal experience, there is a correlation. I kinda figure this has to do somehow with the Q calculation, but it’s just not happening for me, math-wise. For example, arguments about relative error seem to cancel out because of the greater number of ticks per second. I’m putting a talk together and would like to rationalize this somehow. Googling hasn’t provided any insights so far. Links to references would be great. Thanks for any pointers!
> J. L. Sloan Digital Aggregates Corp.
> +1 303 940 9064 (O) 3440 Youngfield St. #209
> +1 303 489 5178 (M) Wheat Ridge CO 80033 USA
> jsloan at diag.com http://www.diag.com <http://www.diag.com/>
> time-nuts mailing list -- time-nuts at febo.com
> To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
> and follow the instructions there.
More information about the time-nuts