[hpsdr] Horton LO

[Ray Anderson]

Bob N4HY wrote:

>It will be exceedingly important for phase coherent processing that we 
>use the same clock to derive other clocks.   So for example,  Horton 
>needs to hear the minimum Who by having this same clock have a divider 
>that will produce a 200 kHz or N*200 Khz clock for the AKM5394A.  This 
>will make  the entire system phase coherent with a dynamic range, and 
>other parameters never before achieved to my knowledge.  We are 
>basically talking a serious revolution with this.  I would suggest 
>strongly in addition that we make provision for driving Mercury from
the 
>125 MHz clock (giving up10 MHz of bandwidth) to, once again,  get a 
>phase coherent system.  NISSSHH.

>The Phil's have hit a home run here.

>Bob

As Bob pointed out, distributing a common reference clock to JANUS, OZY,
CASMIR, MERCURY, HORTON or wherever from a low jitter accurate source
will enable the configuration of a phase coherent system.

If HPSDR goes this route we need to consider if we want to distribute
the low jitter reference via the ATLAS buss or via dedicated coax
jumpers between boards. Further, as the High frequency clock (125 MHz ?)
will probably be LVDS, particular attention needs to be paid on how the
clock distribution buss is terminated. Creating a global clock
distribution tree is more involved than establishing a simple
point-to-point network. We'll need to provide a clean clock to all the
consumers without excess ringing, reflections and such lest we destroy
the low jitter potential established at the clock generator.


-Ray	WB6TPU





 1150490770.0