[time-nuts] M12M Receiver Disiplining PRS10

[SAIDJACK at aol.com]

Hi there Lee,
 
I am doing that right now. Attached see an ADEV plot of the PRS10  
disciplined by an M12+.
 
My experience is that:
 
   1) the PRS10 unit I have generates a 1PPS output which is  about 35-40ns 
too early on average when compared to the 1PPS input from the GPS.  I think 
this offset can be canceled by sending commands into the PRS10
 
   2) I use a 7-hour averaging time, the PRS10 itself is very  good for time 
periods shorter than 7 hours except:
 
   3) The unit has a very large 1-2E-011 hump at around 20s time  intervalls, 
this was discussed in this forum about a month ago. See the PDF file  
attached.
 
   4) the PRS10 generates a very short, and weak output  pulse on it's 1PPS 
output (10us if I remember correctly). That needs to be  extended and buffered 
in my application.
 
For the ultimate in Rb performance, I think the best one can do is use a  
very good stability cleanup oscillator with a 100s or slower PLL response time  
to filter out the annoying noise generated by the PRS10 between 1s to 100s  
intervalls. This can be done with an MTI 260 series oscillator for example. In  
that case, it's possible to get ADEV stability of about <5E-012 above  one 
second.
 
hope this helps,
bye,
Said
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Adev_prs10.pdf
Type: application/pdf
Size: 25625 bytes
Desc: not available
URL: <http://febo.com/pipermail/time-nuts_lists.febo.com/attachments/20060924/b62642eb/attachment.pdf>