PSN-L Email List Message

Subject: Re: Seismograph Timing
From: Bob Smith bobsmith5@........
Date: Wed, 29 Nov 2000 07:31:42 -0500


Hello PSNers --

I "lurk" here as a result of a latent interest in seismology
though I am not personally active in the field.

I have, however, maintained a long standing professional and
personal interest in the field of precision timekeeping.  I
can appreciate the importance of precision time references
to seismology.

The seemingly simple requirement of maintaining a precise
time reference (clock) leads to a field that can become
extremely complex when precision is taken to its limits.

For your purposes I can suggest the following (in order of
increasing precision):

1.  Use your PC clock to time stamp or otherwise mark your
plots.  PC clocks are notorious for wandering all over the
place however.  In the absence of regular discipline they
can be expected to accumulate errors in the order of 10s of
seconds within a few days to a few weeks.  There is
sometimes discussion that improving the accuracy of the Xtal
in the PC will improve the accuracy of the clock.  This is
doubtful, however, since much of the inaccuracy is traceable
to failure of the OS to reliably service the timer
interrupts.

The accuracy of the PC's clock can probably be improved by
regular (say daily) manual adjustments using the radio
transmissions of WWV (as an example) as a reference.

There are also I/O expansion cards available that carry a
high accuracy time base which is often referenced to an oven
stabilized (OCXO) or temperature compensated (TCXO) crystal
oscillator.  These can hook into the system BIOS so that
calls for system time will be derived from the add-in
reference.

2. The PC clock can also be disciplined in the manner
suggested by Mr. Hancock using a program that communicates
with the Network Time Protocol (NTP) or SNTP servers
maintained by numerous sites throughout the world such as
Tick and Tock at the USNO.  A particularly nice package for
this purpose is 4DTIME which is offered as a freeware
package at 

	http://www.thinkman.com

This is true "freeware" not a "demo" or "evaluation"
version.  The only "payment" that Rob requests is an email
describing your application and suggestions for improvement.
If you are so lucky as to enjoy a full time internet
connection such as I do through my cable modem, this
software can be set to wake up periodically and do an
automatic PC clock up date while running minimized.  Using
this technique you can probably maintain PC clock accuracy
to better than 0.1 second.

3. Next would be using a WWV or WWVB radio time standard
receiver.  One of my favorites is the old (now obsolete)
Heathkit GC-1000 Most Accurate Clock.  This is a scanning HF
receiver that samples the WWV transmissions on 5, 10, and 15
MHz and decodes the time signals which are used to
discipline a local 6 MHz oscillator.  I have the optional
RS-232 output port installed on mine.  This emits the local
time base once per second.
Using a reasonable antenna so that it gets regular updates,
this clock will maintain time accuracy in the order of a few
10s of milliseconds.  These things are becomming quite rare,
so if you are lucky enough to run across one at a ham radio
fest or yard sale, *grab* *it*!!.

4. Finally, through the efforts of Dr. Thomas A. Clark,
W3IWI, a well known radio amateur and consummate tinkerer
and a number of associates, truly precision GPS based time
keeping has been brought to the level of amateurs.  Dr.
Clark has developed hardware/software solutions which bring
the very high precision of the GPS satellite time bases to
the common PC.  Using his methods, sub-microsecond accuracy
time bases become available to the amateur seismologist. 
Ultimately, time base accuracy in the order of a few 10s of
nanoseconds is possible.

For further information I suggest that you subscribe to the
TAPR (Tuscon Amateur Packet Radio) TACGPS email list.  Go
to  http://www.tapr.org  and follow the links to the Special
Interest Groups (SIG) mail lists subscription page.

Also, visit   http://www.synergy-gps.com/   Synergy
specializes in marketing the Motorola family of Oncore GPS
receiver modules which are great for timekeeping
applications.  In particular, the Oncore UT+ GPS receiver
has been optimized for precision timekeeping purposes and is
the one capable of serving up time to a precision of about
10 nanoseconds.

5. If that doesn't keep your seismometer happy, go buy an
atomic standard.  These occasionally show up on ebay!  
http://www.ebay.com

	Good luck and good time, Bob Smith

	

"Robert L. Hancock" wrote:
> 
> I have noticed periodic exchanges on the problems and solutions to
> obtaining the correct time for timing and keeping timed a seismograph..
> I plan on building one in the forseeable future, and this along with
> other subjects has been of interest to me.  Recently, I ran across a
> page from the US Naval Observatory, our official timekeeper, about
> computer programs that can be downloaded and used to maintain timing
> with computers  I have installed one on my computer and was surprised at
> how much drift there was in my cpu clock.
> 
>         http://www.eecis.udel.edu/~ntp/software.html
> 
> The main page for the US Naval Observatory is
> 
>         http://www.usno.navy.mil/
> 
> I am not certain how useful this will be to anyone, but thought I would
> mention it as I do remember having seen this source discussed.  Most
> discussions seemed to be focused on either receiving WWV or a similar
> type signal, or using the timing obtained from the GPS satelites.  This
> site has a large variety of programs utilizing many different operating
> systems.  Some are shareware and free while others require registration
> and $.
> 
> Hopefully this will help someone..........
> 
> Cheers,
> 
> Bob Hancock
> Randolph, NJ
> 
> __________________________________________________________
> 
> Public Seismic Network Mailing List (PSN-L)
> 
> To leave this list email PSN-L-REQUEST@.............. with
> the body of the message (first line only): unsubscribe
> See http://www.seismicnet.com/maillist.html for more information.

-- 

---------  Avoid computer viruses  --  Practice safe hex 
-------------
 * * Specializing in small, cost effective embedded control
systems * *
Robert L. (Bob) Smith			Smith Machine Works, Inc.
internet   bobsmith5@.............. Lumlay Road
landline   804/745-1065	                Richmond, Virginia
23236+1004
__________________________________________________________

Public Seismic Network Mailing List (PSN-L)


[ Top ] [ Back ] [ Home Page ]

Larry Cochrane <cochrane@..............>