[time-nuts] Are these PRS10's worth it ?

[Charles Steinmetz]

Bruce wrote:

>By the way, The TS-2500 is a GPS-referenced source; however, the 
>unit seems to compare its internal crystal oscillator with GPS, the 
>PRS-10, and other 10 MHz sources (that can be connected for 
>monitoring) and keeps track of their behaviors with a 
>microprocessor.  It appears the PRS-10 is not locked to GPS, but is 
>simply monitored and kept in reserve as a replacement frequency and 
>timing source should GPS service fail.  If the PRS-10 frequency 
>wanders out of the acceptable range, the TS-2500 shows a fault 
>signal, but does not attempt to adjust the PRS-10.  Apparently the 
>PRS-10's frequency is set at the Symmetricom factory and the unit is 
>on its own after that.

I have played with several TS-2700s (which are disciplined by CDMA 
cell transmitters rather than GPS), and they work the same way.  They 
track all of the timing data they are given, and output the "best 
fit" 10MHz they can calculate using their "BesTime" engines.  The 
TS-2700 keeps a running log of its estimate of the PRS10 frequency 
but, as you say, it does not adjust the PRS10.

>It is included in the TS-2500 box as baggage to be used only if the 
>GPS timing fails. As it is kind of expensive to burn an SRI PRS-10 
>simply as a standby

I'm not sure the TS-2700 (and, by implication, perhaps the 2500) ever 
just outputs the raw PRS10 frequency.  I think it always outputs a 
frequency calculated from the various timing inputs.  If it loses the 
CDMA signal and falls back to the PRS10, I think it uses the stored 
offset for the PRS10 to generate the output frequency.  The manual 
says, "If the CDMA signals are disrupted, the BesTime algorithm 
continues to predict CDMA timing information, which enhances system 
holdover performance."  I don't know if it uses the last-known 
offset, or tries to project a "current offset" based on the trend line.

Best regards,

Charles