[time-nuts] Synergy Drop in GPS evaluation

[rbenward at verizon.net]

Hi All,

I have taken my preliminary set of data on the Synergy drop-in GPS.  I've
got data and images on Flickr.  A single GPS antenna with a splitter was
used with three GPSDOs, a Trimble board level GPSDO, a Z3905A and a Z3801A.
I used radar plots, as Excel histograms didn't really show the data in a way
that seemed useful.  Neither did a plain line chart.  The radar circle seems
a good way to visually show spread for a 10 samples data set.

 

My preliminary impressions:

*	There is not a lot of difference between the Synergy GPS and the old
Motorola.  Note the Synergy is in the Z3805A with a UCT 8663 OCXO and the
Moto is in the Z3801A with a bare 10811.  The next block of data will have
the Synergy combined with the 10811, whether the pair is in the Z3805A or
the Z3801A.  I may also use another 8663 with the Moto in the Z3801A. TBD.
*	The Synergy GPS consistently shows six satellites on the CN chart,
the moto, at the same time may go down to three.
*	The Synergy is consistently missing the Az-ALT of many satellites,
even while showing them as tracked and on the CN chart. 
*	On the surface, the Synergy consistently had more satellites and
higher levels.
*	The CN levels are consistently higher on the Synergy, but I'm not
sure if the scale is arbitrary or they all use the same scale parameters
*	The block diagram shows a Rubidium oscillator as the reference, as
they have the short-term stability reputation.  After completing a set with
the Rubidium, I took the same with the PM6690 on internal.  The data shows
the internal reference of the Fluke PM6690 is a little better than the
Rubidium.  This surprised me.
*	The Trimble data. Although the data shows the P-P and STD deviation
to be significantly larger, the Adev deviation was comparable to the others.
Pg 3.
*	On the Time Stability measures, Allan Deviation, Z3805A & Z3801A.
These consistently leapfrogged each other. Below 10E11, take some more data,
and now it's below 10E12, more data and it's back to 10E11.  The lead often
went back and forth, as did the EFC stats.

 

Right now, the data are JPGs on Flickr.  Check the descriptions associated
with each picture.  I will be documenting and putting all the pictures into
a .doc or .pdf .  for now, pls refer to pg-1 thru4.  The pictures of the
Z38xx windows were captures on three occasions, denoted by an _1, _2, & _3
in front of the file name. Sorry, Flickr is not that great for re-arranging
the order.  Email me if you have questions.

 

https://www.flickr.com/photos/203157635@N08/

Bob