Note: click here for full MSF pages on this CSC web site MSF pages

MSF - 60KHz time signal ...
and some memories of the NPL (1988) Dr Jonathan Hare

In 1988 I spent a happy industrial year working at the Time and Frequency section of The National Physical Laboratory in Teddington (London) as part of my Physics degree course [1]. It was fun to be a Time Lord for a year!

I was fortunate to have Dennis Sutcliffe and Brian Swabey to guide me during this year. Dennis was in the final year of a very successful career at the NPL which included working with Louis Essen [2] on the pioneering research on the first atomic clocks.

My duties were to help out with stability measurements of the various atomic clock standards both at the site (London), the MSF time signal site (Rugby) and via also GPS with other standards labs around the world (USA, France etc).

During the year I also built MSF receivers [1], time comparison equipment to use a TV transmission as a reference signal and also made a stable portable stop watch with well known temperature coefficient for standard and calibration purposes in industry.

I regularly used the rooms in the basement in Bushy House. By the time I arrived these room had mainly become store rooms packed full of interesting equipment. Previous researchers had laid out table tops and table tops of electronic components and equipment so it was easy to find. To me the potential of all this stuff was just fantastic, it was an Aladdin's cave ... just waiting for me to experiment with (Brian Swabey and Dennis Suitcliff were very kind)!

The basement of Bushy house also had three temperature controlled rooms at -5C, 0 and +30C. Each had thick 6 inch doors to keep them stable and they were used to test out equipment at various temperatures over extended periods. I heard later that, as a cash cutting exercise to save money, the coolers and heaters were turned off so that they were not on all the time. The result was that the doors swelled up and they were no longer usable.

Short history of MSF
The MSF time code transmission (MSF, [3]) started in 1950's with the aim of providing a time signal / reference of known accuracy / stability for the UK and near Europe. By 1956 there was a permanent VLF signal established on 60KHz providing a slow and fast code signal in CW (more later). In 1998 the fast code (sent at the end of every minuet) was removed from the MSF signal, and now only the slow code remains.

LF radio time signals
include [4]:

MSF 60kHz (17 kW ERP, caesium standard) all UK
WWVB 60 kHz (50 kW ERP, Colorado, USA)
JJY 60 kHz (50 kW ERP, Japan)
Droitwich 198kHz (rubidium standard) England + Wales
DCF77 77.5 kHz (50 kW ERP, Germany)
Loran-C 100 kHz
Note: In 1988 the UK 2.5, 5 and 10MHz time signals were discontinued.

MSF at Rugby
During my year at the NPL I went up to Rugby with Brian Swabey (NPL Research Officer) on one of the regular visits to take up a portable caesium standard and use it to make direct measurements with the MSF standard at Rugby.

The drive into the site was very impressive; all around me I could see the vast antenna mast array with its many, many tall towers ca. 250m (800') having granite base insulators that looked the size of a small house! One story I heard was that each mast would have a service lift that travelled up inside the lattice tower to the top. On at least one tower the journey up was quite frightening as the lift would often get momentarily stuck part way up due to a twist that had developed over the years. At some point on the way up there would be a very uncomfortable quite stop (which probably felt a very long time) before the lift shunted up, and on its way up into the heavens!

The main site was a large building containing the MSF transmitter valve (ca. 4' high), antenna coupling coils (large litz wire vari-ometer) and also the 16kHz transmitter equipment. Going around as a visitor it felt like I had become an ant wandering around inside an old valve radio!

Why use valves? They tried using a solid state transmitter using an array of power transistor but they had problems with banks of transistors burning out, either from close strike lightning or other electrostatic build-up when the top of the masts were in the clouds (which was quite often). Valves are actually electrically very robust in high power applications so they are still used today.

In 2007 the station was transferred from Rugby (52° 22' N, 1° 11' W) to Anthorn (54° 55' N, 3° 15' W). The current station consists of a 27kW CW transmitter using a vertical (large top-hat) antenna of 17 kW ERP (vertically polarised).

The MSF signal is derived from two caesium atomic clocks (with a Rubidium back-up). Short term stability (phase distortion at reception point) is ca. 1ms of NPL (UTC, same a GMT). You need to integrate for at least several seconds to get rid of slow code 'bits' effect but if the MSF carrier is integrated over 24 hours stabilises of 2 x 10E-12 are possible. This is equivalent to a theoretical long term stability of an error of about one second in 15,000 years.

In addition to the UK MSF (2 bit code) on 60kHz the USA (1 bit code) and Japan (JJY, 1 bit code) also have time standard signals. I have no idea if these other signals can be received from the UK. It may be possible to separate them out using a directional antenna (nulled to UK MSF). It would be fun to try and detect the delayed signals from the more distant time signals underneath MSF. I recently made up a large 40 turn 4' loop so I am looking into it : )

The MSF UK transmission is a 100 % AM CW signal. The carrier drops to send (two) bits of information (only one bit at the start of the minuet where the carrier is off for 500ms). Best phase stability is usually at mid day and there are often (interesting [5]) large phase change at dawn and dusk.

For reception <1000km from the transmitter the signal is usually a ground wave signal while for >1000km its predominantly skywave. Between these distance it usually a vector sum of both. MSF signal strength should be ca. 100 μV/m over a 1000km range from the Tx site.

Note: MSF is off a few times a year for maintenance and service bulletins are provided by the NPL (see web site below) [6].

A small ferrite rod antenna provides easy pick-up but they have poor temp coefficient and are sensitive to noise. A large loop is best for a fixed device / receiver if you have the space.

Possible noise problems include the 4th harmonic of TV sync pulses close to 60kHz as well noise from switch mode power supplies (SMP) which are everywhere these days.

Notes / References
[1] My year at the NPL was part of the progressive Physics degree course at Surrey University. While there I had a lovely time building three MSF receivers; Tuned Radio Frequency (TRF), Direct Conversion (DC) and a Superhet designs.
[2] Louis Essen, see wiki page: Louis Essen
[3] click here for MSF web link URL
[4] For details of other LF radio times signals
[5] Interesting to me anyway! I was at a meeting at the NPL in 2007 and was shown firsthand the interesting phase change plots and changes in signal strength recorded on almost on everyday over the Wales - Teddington path.
[6] click here for MSF up date bulletins
[7] Wikipedia entry on MSF


Dr Jonathan Hare, The University of Sussex
Brighton, East Sussex. BN1 9QJ

home | diary | whats on | CSC summary | latest news