[time-nuts] Need help phase locking with small offset

[Poul-Henning Kamp]

--------
Stewart Cobb writes:

> Any comments from the analog gurus?

Many years ago I implemented a "timecounter" in a Xilinx XC6200
FPGA and one of the things I learned is that FPGA's are really bad
at crosstalk.

The active circuit was basically a 26 bit binary counter running
at 100 MHz, with a bunch of latches triggered by external signals
through proper metastability mitigation.

But if I put a single flipflop in a corner of the XC6200 and clocked
that with a beat-frequency, I would clearly see that beat-frequency
in my timestamps.

Whatever you do, spend enough time to measure your noise-floor before
you trust your data...

-- 
Poul-Henning Kamp       | UNIX since Zilog Zeus 3.20
phk at FreeBSD.ORG         | TCP/IP since RFC 956
FreeBSD committer       | BSD since 4.3-tahoe    
Never attribute to malice what can adequately be explained by incompetence.