[time-nuts] FE-5680A performance

Tom Van Baak tvb at LeapSecond.com
Thu Jan 5 19:27:46 UTC 2012


> All the wonders of the TSC boxes aside, there is one really annoying 
> thing you need to keep in mind when working with the phase data: the box 
> applies an arbitrary phase offset that reflects through to the phase 
> data output.  It appears as a linear frequency offset that needs to be 
> removed prior to doing frequency analysis.  That's not a big problem for 
> normal stability measurements, but in some cases it makes it a lot 
> harder to extract the data you want.

Right, it has no effect on ADEV plots (which are immune from fixed
phase or fixed frequency offsets).

I see this on the TSC 5120 but not the 5110. Same for you? When
in "single DDS" mode the 5110 phase is absolute. But when in the
"dual DDS" mode the raw phase output is scaled by "B/A frequency",
which is displayed at the lower right side of the screen. I don't recall
this value changing over time so I'm curious about your fcounter
comments.

> John, could you expand on your comment about the ethernet connection?  I 
> have seen random instances where the TSC seems to lose its TCP/IP 
> settings, but so far it's never happened during a measurement run, so 
> has been annoying but manageable.

I totally agree. I had no LAN trouble for many years but when I upgraded
(2005?) to a later TSC firmware my old PC would sometimes loose the
net connection. The RS-232 output is always ok. I tried multiple PC's
and multiple (Windows) OS's and megabytes of TCP traces but never
solved the problem. It's rare but when you make unattended runs it's
very sad to lose days of data due to a TCP connection closing. This is
one reason why (for modest data rates) I still prefer vintage RS232 or
GPIB connections compared to "modern" USB or LAN connections.

/tvb




More information about the time-nuts mailing list