[hpsdr] crystal Oscillators

[Chris Albertson]

--- michael taylor <mctylr at gmail.com> wrote:

> 
> I believe in essence the Gibraltar design is based around building
> our
> own MCXO (microprocessor controller crystal oscillator) which given
> the usage of an ovenized SC-cut crystal rather than a standard HC-49
> AT-cut crystal should be better than most MCXO modules. My impression
> is that most MCXO appear to be aimed at improving performance in
> comparison to analog TCXO while remaining cheaper than an OCXO. I
> could be easily wrong here.

This gives a good way to trade off performance for cost too.  For
example
an end user could decide to spend the money for an ovenized crystal
or not.  Same for the GPS.  He could go with one he already has or even
use some other time source such as  NTP.  Of buy a GPS with a good PPS
output.  At the low end the system could be nearly free, under $20 and
at the upper end it could go to four figures.  

What could be done here, that is new is a design that is flexible. 
Allowing
several options for both the oscillator and the clock input.  A design
that
could span one order of magnitude in price and several in accuracy 
would
be impressive.

The basic design is very simple.  It's just a variable oscillator and a
counter
to measure the  oscillator and generate an error signal that get's sent
back to the oscillator.  A servo loop.   The details will matter and
I'll bet
many end-users would like a system that they can tinker with and
experiment
with various cntrol algorithm ideas.  So think you want the controller
to
be written in something that can be modified, like say "C".


Chris Albertson
  Home:   310-376-1029  chrisalbertson90278 at yahoo.com
  Office: 310-336-5189  Christopher.J.Albertson at aero.org


      ____________________________________________________________________________________
Shape Yahoo! in your own image.  Join our Network Research Panel today!   http://surveylink.yahoo.com/gmrs/yahoo_panel_invite.asp?a=7 



 1182046430.0