NTP Reference Clock Statistics
I run four NTP Stratum 1 servers with attached reference clocks. Below this introduction are graphs that show performance statistics about them, as well as some other internal and external NTP servers that I monitor.
tick, tock, toe, and cesium are dedicated NTP servers, built from Soekris net4501 single-board computers with 133MHz AMD "Elan" processors (486 class). That may not seem like much horsepower, but these little guys have a secret weapon: a very high resolution timer that can be used to get timekeeping accuracy far better than any normal PC.
In addition, these Soekrises (Soekri?) have been modified to replace the normal crystal oscillator with a synthesizer driven by an external high stability source which makes them almost immune to temperature variations and other errors.
Here are details about the four servers:
- tick syncs to the PPS and serial data stream from an HP Z3801A GPS Disciplined Oscillator. Its 10 MHz clock also comes from the Z3801A. (Refid "GPSZ")
tock is connected to a Spectracom 8170 WWVB receiver. It syncs
to both the serial data stream and the PPS output. Its clock comes from
the house standard (normally an HP 5065A Rubidium clock). (Refid "WWVB")
While tock has the advantage of being directly traceable to NIST time through WWVB, it is actually the least stable of the ensemble in the short and medium term because ionospheric conditions affect the WWVB signals as they propagate from Boulder, Colorado to my house. It often wanders by a few milliseconds, and when the signal momentarily disappears completely, it can spike 20 or 30 milliseconds before recovering.
- toe is connected to a CNS-II Clock, which is mainly a Motorola M12+2 GPS receiver with a hardware sawtooth correction circuit that reduces the nanosecond-level jitter in the PPS output. It syncs to both the serial data stream and the PPS output, and its clock source is the house standard. (Refid "GPS")
cesium gets its PPS from an HP 5061A Cesium Beam
Frequency Standard (a real "atomic clock"). Because the 5061A provides
only a raw PPS and not a full timecode, cesium relies on one of
the other NTP servers to tell it the time to the nearest second; the atomic
PPS then provides a precise second marker. (Refid "ATOM")
Because cesium does not rely on a radio receiver, it has low jitter and is not dependent on external factors. But that also means that it's not getting any correction from NIST or USNO time, and as a result it will drift over time. Currently, I'm measuring that drift as about 20 billionths of a second per day (20 ns), so it's not a major concern.
I use cesium as the reference system that monitors the other servers. Its 10 MHz clock comes from the house frequency standard.
The charts below show the offset of each of the internal servers, as well as some internal and external servers, referenced to cesium. All the internal machines are on their own 100 Mb Ethernet switch, which connects to the house GigE network, so network traffic and delays are low, and these systems represent probably the best an NTP network can do using standard networking hardware. There's also a chart, for what it's worth, showing what cesium thinks is the offset of the HP 5061A clock that drives it.
The "current day" charts show performance for the current day since 0 UTC (updated hourly). The next set of charts shows the preceding seven full days, plus the current day. The final charts show 30 days of data. (Here are the tools I used to generate these plots.)