In The July 1998 issue of QST Brooks Shera, W5OJM, wrote an article titled, "A GPS-Based Frequency Standard".  It uses the GPS satellite constellation to improve the accuracy of voltage controlled oscillators.


The picture above is of a copy of his GPS controller that I built from his article.  It is sitting on top of the DSP-10 I built from Bob Larkin, W7PUA's article.  The frequency standard for the DSP-10 is an HP 10811 High Precision Oscillator sitting in an HP 5328 500 MHz Universal Counter that is in a rack cabinet out of the picture to the left.

The -092 displayed on the digital display is a relative indication of the GPS-controlled control voltage being applied to the 10811 at the time the picture was taken.  The meter to the left of the digital display is indicating the phase difference between the GPS timing signal and the 10811.  The needle stays in the yellow zone, where it is in the picture, while lock is maintained.

The little metal box sitting on top of the GPS controller is a buffer/amplifier by DJ8ES that I built to provide several outputs from the 10811.  It feeds a TTL signal to the GPS controller, an analog signal to the DSP-10, and has 2 spare analog outputs and spare 5 and 1 MHz TTL outputs.  It was built with SMT technology, like the DSP-10.  The GPS controller uses thru-the-hole construction techniques.

Brooks has a great web page and  with lots of info regarding his controller.  The company listed in the QST article still makes the circuit boards.  Mine came so fast I didn't have all the parts ready yet when it arrived!

This device really works.  It should keep the frequency accurate to 1 part in  10 to the 11th power.  My rough calculation here indicates that the variation over a day here was less than 0.003 Hz for the reference signal at 10 MHz.  I calculated this using a record of phase differences I recorded from the ASCII output of the controller, as I don't have anything here capable of measuring that degree of accuracy. Needless to say, when I compare the frequency to a separate oscillator controlled by my Rb standard but running at 144 MHz, I can detect no drift (with a resolution of 1 Hz at 144 MHz) over the period of a week.  This is truly a remarkable piece of work by Brooks Shera.  And it has allowed me to put away my Rb standard (with its finite 'filament time') and to save the Rb standard for occasional calibration tests.

I will be using the GPS controller for some of the low power EME experiments I have been doing with Bob Larkin.  I've devoted several other pages of this website to Bob and his DSP-10 design, and have links to his site from those pages.

Copyright 1997-2007 COPYRIGHT Roger Rehr W3SZ. All Rights Reserved

Brought to you by the folks at W3SZ