On 02/07/20 17:00, time-nuts-request@lists.febo.com wrote: > Experiments in NTP performance (Watson Ladd) Hi, I may have something to contribute to this, Built an experimental system last year. Two or three surplus spares or repair grade and refurbished time servers on a separate subnet, interrogated by a Mini ITx PC box running FreeBSD / ntpd. 1 PPS support via the serial port dcd line. Registered with ntp.org, but not online yet as still confirming system is rock solid. On a ups and has been running now for months. Typical output from ntpq -p is: root@ntp-host:/etc # ntpq -p remote refid st t when poll reach delay offset jitter ===================================================================== oPPS(0) .PPS. 0 l 2 8 377 0.000 0.001 0.001 *192.9.200.168 .GPS. 1 u 45 64 377 0.189 -0.010 0.024 +192.9.200.169 .GPS. 1 u 19 64 377 0.378 -0.012 0.037 -ntp0.uk.uu.net .GPS. 1 u 42 64 377 16.940 0.352 0.129 Two local sources and one external, for a sanity check... Regards, Chris