NTP Reference Clock Statistics
Note: I originally set these NTP servers up many years ago and when we moved from Georgia back to Dayton in 2017 they did not get reinstalled. Four years later I'm finally getting around to that. In the interim, a couple of things have changed, including a change to the WWVB standard time broadcast format that broke my WWVB receiver. Because of that, I've rearranged the server configuration and what's shown below is a bit of work-in-progress as I get things running again. At the moment I'm getting the hardware working and it may be a little while before the statistics software that generates the plots below is brought back on line.
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 hydrogen 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.
I've written up details on how to convert the net4501 computers for high performance timekeeping.
Here are details about the four servers:
- tick syncs to the PPS and serial data stream from an HP Z3805 GPS Disciplined Oscillator. Its 10 MHz clock also comes from the Z3805. (Refid "GPSZ")
- tock is connected to a CNS-II Clock, which is mainly a u-blox M8T 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 10 MHzclock source comes from the 10 MHz oscillator in the clock. (Refid "GPS")
- toe gets its PPS and 5 MHz from an HP 5061A Cesium Beam Frequency Standard (a real "atomic clock"). Since the second and the Hertz are defined in terms of cycles of the cesium atomic resonance, if a cesium clock is working properly, by definition it's correct within some degree of uncertainty. This 5061A is an older unit and it has more uncertainty than either a GPSDO or the maser, but it's still more than good enough for this application. Because the 5061A does not provide a timecode, toe learns what second it is from one of the other servers. (Refid "ATOM")
- hydrogen gets its PPS and 5 MHz clock from a VCH-1008 passive hydrogen maser which is by far the most stable frequency source in the lab. However, unlike a cesium clock the maser does have a small amount of frequency drift with time. To compensate for that its frequency and PPS are steered very slowly (on a scale of many days)to follow GPS. It also gets its timecode from one of the other servers, though this may change someday as the maser can be queried for current date and time and that could be made to look like a timcode to the NTP software. It's Just a Small Matter of Software(tm). (Refid "MASR")
The charts below show the offset of each of the internal servers, as well as some internal and external servers, referenced to hydrogen. 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 hydrogen thinks is the offset of the maser 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.)
One of my externally-visible stratum 2 servers, meow.febo.com is part of the pool.ntp.org project. Statistics about meow are here.