Wrist-worn frequency standard

[David Rasmussen]

I note with a smile that Citizen has made what it touts as the most accurate watch in the world: +/- 1 second per year. It's heart is a solar cell driven oscillator using an AT cut crystal at 2^23 Hz (8.388608 MHz). That is much different than the standard 32.768 kHz tuning fork crystals used in most quartz watch movements, and is very much akin to the 10 MHz frequency standards technology that we lovingly designed into all sorts of products over the history of my experience with HP > Agilent > Keysight. The watch oscillator is not ovenized (though it does sit on a wrist that is biologically temperature controlled :-) but the watch circuitry does measure temperature and applies compensation to the oscillator based on known TC for the quartz crystal. 

Most people won't care about this stand-alone time/frequency standard you can wear on your wrist. You can get a cheaper Apple watch that is always as accurate as GPS clocks. But it makes the engineer in me smile. We were never guilty of over-engineering something. No. Not ever.

David Rasmussen

[Bill Harris]

Do they spec phase noise?

--
You received this message because you are subscribed to the Google Groups "exlks" group.
To unsubscribe from this group and stop receiving emails from it, send an email to exlks+un...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/exlks/ce948d8c-b082-4962-96c4-9c21f201e1bcn%40googlegroups.com.

[David Rasmussen]

No, but in any case I’d like to see you measure it, just looking at a second hand :-)

[Neel Malik]

Looks like not spec'ed.

https://www.citizenwatch-global.com/the-citizen/lineup/cal0100/index.html

Get Outlook for Android

---

From: ex...@googlegroups.com <ex...@googlegroups.com> on behalf of David Rasmussen <davee...@gmail.com>
Sent: Monday, June 26, 2023 2:51:05 PM
To: bill_...@facilitatedsystems.com <bill_...@facilitatedsystems.com>
Cc: exlks <ex...@googlegroups.com>
Subject: Re: Wrist-worn frequency standard

 

You don't often get email from davee...@gmail.com. Learn why this is important

To view this discussion on the web visit https://groups.google.com/d/msgid/exlks/CAHJGFWfS%3DMVuqfQmDrA7u-dotryT5fR7q%2BexcJ2dk%2BYqoOgJOQ%40mail.gmail.com.

[smoothfox]

But do they go to the edge of a pier and throw it into the Sound for a stress test?

 

Cheryl Marks

 

Sent from Mail for Windows

 

From: 'Neel Malik' via exlks
Sent: Monday, June 26, 2023 2:52 PM
To: David Rasmussen; bill_...@facilitatedsystems.com
Cc: exlks
Subject: Re: Wrist-worn frequency standard

 

Looks like not spec'ed.

https://www.citizenwatch-global.com/the-citizen/lineup/cal0100/index.html

 

Get Outlook for Android

To view this discussion on the web visit https://groups.google.com/d/msgid/exlks/PH0PR00MB1152B164C95061F0E3DAD1EBAE26A%40PH0PR00MB1152.namprd00.prod.outlook.com.

 

[Glenn Engel]

For my CrewTimer mobile apps I need accurate time and my first implementation many years ago worked on the assumption that the tablets with 3G/4G would sync with the mobile network and have consistent accurate time.  After a few experiments I discovered that while they might sync to the carrier, they could be 'off' from actual time by 2-3 seconds either way.  While the actual 3G/4G transport relies on accurate timing between the device and the towers, I suspect that the OS got a time update on some interval and in between was left to drift with the temperature and when left off, might drift at yet a different rate.

In the end I had to create my own SW time sync using public NTP servers which gets me to several milliseconds accuracy.

I suspect the apple watch may suffer the same inaccuracy as iPads.  However, since smart watches all sync with their companion phone they will always be accurate to 'several seconds a year' as well as 'several seconds' at any point in time.

G

To view this discussion on the web visit https://groups.google.com/d/msgid/exlks/CAOBquGrTDq_607ZG2o6qP_e1zywFWXXkUxWp2cvMwhFxcRj0pw%40mail.gmail.com.

[Daniel Wisehart]

After I left LSID I went to Labs where I got to work with John Eidson, who, among others, helped create the time sync protocol use by Vantera, which later became the Precision Time Protocol, IEEE 1588, of which John was the Chairperson.  You probably know that already.  Years later I contacted John about some extensions to PTP done by the group at CERN called White Rabbit.  John not only knew about it--which he called a piece of brilliant engineering--he had talked to them and maybe even advised them on their work.

In my work with trading systems, we use White Rabbit to give us reliable, ~100ps accurate relative timestamps across WANs utilizing synchronous Ethernet.  Since the master clock is GPS derived, our absolute accuracy is no better than other GPS devices, but for relative times up to a day in length, these are amazing tools.

Daniel