[hpsdr] coherent dual mercury receivers

[Chris Albertson]

On Mon, Mar 14, 2011 at 9:00 AM, Joe Martin K5SO <k5so at valornet.com> wrote:
> ***** High Performance Software Defined Radio Discussion List *****
>
>
> Phil,
>
> Now that I'm thinking about it, I recall having some feedback earlier from a
> Garmin engineer about the absolute accuracy of the 1 pps GPS signal.  He
> told me that the accuracy of the 1 pps signal is no better than several uSec
> or even much worse, as I remember.   If your application requires higher
> accuracy/precision than that, which seems likely to me, you may need to
> consider using other means for precise timing.

Two things about that (1) Yes that is what a GARMIN person would say
but there are other GPS recievers that as under $50 and work at the
tens of nanosecond level.

(2) One does not use a 1PPS signal directly.  You use it to phase lock
a very stable ovenized crystal oscillator or even a rubinium osciator.
 Then over time the oscillator becomes very accurate because it in
effect represents an average of thousands of 1PPs pulses.   For about
$200 you can get to sub nanosecond range.   In recent years the cost
to do this has really just hit the floor because of all the cell tower
equipment that is being replaced and resold on ebay.

Test equipment too.   an HP 5370B Universal counter can measure
periods of to an accuracy of well under a nano second and that show up
on ebay for maybe $250.   These sold new for $25K.

The GPS disciplined local oscillator at each receiver and the hardware
to convert to the required frequency, GPS antenna, cables and all
might be about $300 if you can round up parts and do some DIY
building.  With such equipment you can expect about 1 part in 10E11
level of error.  With much effort can get another factor of 10.


-- 
=====
Chris Albertson
Redondo Beach, California

 1300126253.0