Boston Linux Meeting Wednesday, October 20, 2010 Hardware Hacking: Atomic Clock Building

Bruce Labitt bruce.labitt at myfairpoint.net
Thu Oct 14 20:08:24 EDT 2010


  On 10/14/2010 5:22 PM, Michael Bilow wrote:
> On 2010-10-14 17:07, Bruce Labitt wrote:
>>    On 10/13/2010 9:25 AM, Jerry Feldman wrote:
>>> When: October 20, 2010 7PM (6:30PM for Q&A)
>>> Topic: Hardware Hacking: Atomic Clock Building
>>> Moderators: Federico Lucifredi, Product Manager, Novell
>>> Location: MIT Building E51, Room 325
>>>
>>> Federico builds an atomic clock out of a pocket-sized 
>>> Sheevaplug device
>>>
>> Any one go to this?  What was used as the clock?  How does
>> it get to be stratum-1?  That is reserved for a decent
>> clock, IIRC?  As in a primary reference, like caesium?
>>
>> Anyone got a link on the primary ref clock?  I checked out
>> the sheeva plug and its thermally mismanaged variants.  The
>> concept is tres cool.  The thermal issues, not so good.  Any
>> list members have first hand experience with wall-wart
>> computers?
>>
>
> Technically, a Stratum-1 NTP clock means that it is 
> traceable to a NIST reference by means other than NTP. The 
> accuracy of the clock is stated in NTP messages. As a 
> result, a Stratum-N clock could be more accurate in 
> absolute terms than a Stratum-1 clock (although obviously 
> not the one to which it is itself traceable).
>
> For example, it's quite common for a Stratum-2 NTP peer to 
> be within 25ms of absolute accuracy if it is referenced to 
> a very good Stratum-1 NTP peer. If someone set up a lousy 
> Stratum-1 NTP peer capable of only 100ms of absolute 
> accuracy, this would be perfectly legal although fairly 
> useless.
>
> -- Mike
>
>
According to the announcement:

"Using the lowest amount of custom hardware and pouring Perl and Shell
Script over everything as the glue binding it all, we create a
minimalistic device delivering a perfectly tuned network time source:
your very own stratum-1 ntp server, turning a pocket-sized Sheevaplug
device into your personal atomic clock."

and http://en.wikipedia.org/wiki/NTP_server#Clock_strata

snipped for your convenience:

Stratum 0
    These are devices such as atomic (caesium, rubidium)
    clocks <http://en.wikipedia.org/wiki/Atomic_clock>, GPS
    clocks <http://en.wikipedia.org/wiki/GPS_clock> or other
    radio clocks <http://en.wikipedia.org/wiki/Radio_clock>.
    Stratum-0 devices are traditionally not attached to the
    network; instead they are locally connected to computers
    (e.g., via an RS-232
    <http://en.wikipedia.org/wiki/RS-232> connection using a
    Pulse per second
    <http://en.wikipedia.org/wiki/Pulse_per_second> signal).
Stratum 1
    These are computers attached to Stratum 0 devices.
    Normally they act as servers for timing requests from
    Stratum 2 servers via NTP. These computers are also
    referred to as time servers
    <http://en.wikipedia.org/wiki/Time_server>.

I would think this implies the Sheevaplug is connected to a Stratum-0 source.  What was the
source that Federico used?  Anyone know?


-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mail.gnhlug.org/mailman/private/gnhlug-discuss/attachments/20101014/ab72ad03/attachment-0001.html 


More information about the gnhlug-discuss mailing list