The Greenwich Time Signal
British Broadcasting Company
In February, 1922, the Post Office licensed Marconi to operate two experimental broadcast stations. Once was "2MT" in Writtle, near Chelmsford (from which were broadcast the immortal words "Two Emma Toc - Writtle Testing") and the other was 2LO, operating from Marconi House in the Strand, London.
Within months, a number of companies were vying with each other to get a licence from the Post Office. But rather than issue many licences, the Post Office decided that it would charge the group with forming either a single company, or a group of companies, to be responsible for British broadcasting.
The group (which included Marconi, Western Electric and General Electric) decided to form a single company, and on November 14th, 1922, the British Broadcasting Company, as the commercial consortium was known, started regular public transmissions from the Institute of Electrical Engineers building on Savoy Hill, in London. This was despite the fact that it wasn't registered as a company (that happened a month later), and that it was transmitting without a licence, which wasn't issued until 18th January, 1923.
first time checks
However. right from the beginning, the British Broadcasting Company did have regular time signals before the news at 7pm and 9pm. These were not exactly high-tech; they consisted of the announcer playing the Westminster Chimes on a piano, and later on a set of tubular bells specially installed in the studio for the purpose. However, they quickly became a very popular feature of the BBC's transmissions
In order to improve the accuracy of the time checks, the BBC contracted The Synchronome Company to provide master clocks and slave dials for most of its studios centres. A few studio centres (Aberdeen, Bournemouth and Manchester) used special impulse clocks, made by Gent & Company.
The Synchronome system provided a tick which allowed the announcer to count himself down, and hit a gong on the hour. It isn't recorded whether these "ticks" were audible, or whether the announcer counted aloud.
pips are born
Frank Hope-Jones was a well-known amateur radio enthusiast and horologist of the day, and on 21st April, 1923, he was giving a talk on the subject. As his talk finished, he counted the last five seconds up to 10pm aloud. After the broadcast, he put forward a suggestion that perhaps the BBC could have a more accurate time signal using audible "pips".
John Reith, then General Manager of the BBC, and Frank Dyson, the Astronomer Royal, discussed the idea of "broadcasting Greenwich Standard Time", and in December, 1923, an agreement was reached to modify two clocks, used by the Royal Observatory to generate time signals for other users. The cost to the BBC was £20 each, and they were modified "to operate a signal at each 30 minutes to meet your requirements".
The equipment was designed so that a chronometer's escapement wheel at the Observatory controlled a switch, which in turn controlled the output of a 1kHz oscillator. This generated six short pips, starting at five seconds to the hour and ending on the hour, which were then sent down a GPO line to the BBC. They were first transmitted at 9.30pm on 5th February 1924, introduced by Sir Frank Dyson, the Astronomer Royal.
BBC gets the bill
Consequently, they proposed that the BBC pay them £520 a year, and concluded that the BBC "will realise that no guarantee attaches to the service".
The BBC was not best pleased at this suggestion. Vice-Admiral Sir Charles Carpendale (Deputy Director General at the time) responded with an internal memo to the Director of Programmes, stating that he wanted "to fight this charge for all I am worth", and asked for a list of the things that the BBC did for the Admiralty at no charge. The result was a somewhat tongue-in-cheek letter to the Secretary of the Admiralty.
In it, Carpendale enumerates all the time that the BBC devotes to "such subjects as Navy Week at The Three Ports, the departure and arrival of HM Ships, weather forecasts, gale warnings, and the announcement of vacancies for the entry of naval cadets into the RN College, Dartmouth". And he points out that, in the past, no charge had been made and that the BBC would be happy to discuss such charges as may seem appropriate!
Of course, declared Carpendale, it would take some time to prepare the necessary material and that the BBC would be in a position to "apprise Their Lordships of the results before April 1st, 1938, namely the anniversary of the date on which [their original letter was received]".
Their Lordships' response expressed surprise at the suggestion of what could be considerable charges. However, they were relieved to learn that the BBC's Charter actually required "at the Corporation's own expense, send from all or any of the Broadcasting Stations any announcement of other matter which such Department may require to be broadcast". Indeed, such a requirement still basically exists in clause 8 of the BBC's Agreement.
As for the £520 a year bill from the Admiralty for the pips - it doesn't look like it got paid.
Pips on the Move
Back at Broadcasting House, a special amplifier was used that detected the holes in the tone - where there was a hole, locally generated 1kHz tone was broadcast.
There were two reasons for this. Firstly, the tone for the pips was actually generated in Broadcasting House and so would be "pure" and not suffer any interference that might be on the line.
Secondly, by detecting the presence of the tone from Herstmonceux, the BBC would know if the line had failed and could automatically switch across to the other line for the pips. But there would be a short delay before the equipment detected the line failure, which could result in a spurious pip being broadcast. To avoid this, the output of the pip generator was internally isolated until a few seconds before the pips were due. Despite this precaution there have been occasions when a line failure in those few seconds did cause a spurious pip to be broadcast!
Atomic Time and "natural" time are not the same. Atomic time is relentlessly accurate, but "natural" time, on which GMT was based, depends on the rotation of the earth and is subject to some variation as the earth "wobbles" slightly. Until 1971, International Atomic Time had been corrected by making small adjustments every now and again to bring the two time scales into line.
It was eventually determined that Atomic Time would only be changed by whole numbers of seconds and, since the earth was actually running a little slower than Atomic Time, this would require the introduction of an extra second into Atomic Time every now and again to allow it to catch up. These so-called leap seconds are only introduced either at midnight on 30th June or 31st December, and whether we need one or not is determined by the IERS (International Earth Rotation Society)
But this gave the pips a problem - sometimes there would need to be seven, and not six. In which case, how could anyone know they were listening to the last one? So, from midnight on 31st December 1971, the Greenwich Time Signal's last pip was lengthened from 1/10th of a second to 1/2 a second.
end of an era
So, on 5th February 1990 (the 66th anniversary of the first pips), Herstmonceux was replaced by equipment in the bowels of Broadcasting House.
A great deal of effort went into making the pips sound just the same as they had done before the move - they continued to be five pips of 1/10th of second with a final sixth pip of 1/2 a second (the actual time reference is the start of the final pip).
The BBC's time equipment has a very accurate clock which is, in turn, kept accurate by comparing itself to other world time references, such as that transmitted by the Global Positioning Satellite navigation system.
This move included the relocation of the pips equipment, and since 2006 a custom-built pips system (not that unlike the system that had been in use) was brought into operation.
Laws of Physics
In reality, this is highly academic and largely irrelevant. Sound travels very much more slowly than radio waves and at 10 feet from the radio, the pips will be delayed almost 1/100th of a second, much greater than any delay in the transmitted signal.
But the advent of digital broadcasting has changed all this. Turning the audio into a digital signal takes time, distributing the digital signal to the transmitters takes time, and turning the digital signal back into audio takes time - and these times are significant, particularly if satellites are being used.
As an example, the audio from a DAB radio is about 1.5 seconds late, but this can vary significantly depending on the design of the receiver you're using. And if you're listening on the Internet through the BBC's Web site, the delay can be 10 seconds or more. So the idea of an accurate set of pips is slowly going out the window, and there's very little that the broadcasters can do to compensate for the laws of physics.
There was also one occasion when the pips appeared about 4 seconds early, and another when they appeared to come out 10 seconds early (although this last occasion had nothing to do with the pips being early, but rather all the clocks were 10 seconds fast!)
Before the move to Broadcasting House, "spurious" single bursts of tone were occasionally heard, and since the pips are now selected to the network manually (at one time they were fully automatic), once in a while an unscheduled set of pips is broadcast because of operator error.
On another occasion the BBC got its knuckles rapped (unfairly, in my view) for broadcasting a leap pip an hour early! One June in the early 90s, a leap pip should have been broadcast at midnight GMT, that is at 1am BST. The BBC had been advised that the leap pip should be broadcast at midnight, without any qualification as to whether it should have been GMT or BST - calls were made to two authorities to check which it should be. Both sources inadvertently confirmed the time as midnight BST, and that's when the leap pip was inserted - an hour early!
Then, at 8am on 17th September, 2008, they turned up 6 seconds late - and there were seven of them. For some reason the pip-generating equipment had decided to go it alone, and wasn't locked properly to the time reference. Nobody knows why this happened, but the problem was solved by resetting the equipment.
But perhaps the greatest problem with the pips (at least in terms of the degree of publicity it got) occurred in the mid-90s. The late Dr Dolmetsch was a recorder player and composer - and he had written a piece for recorder ensemble and pips. He noticed that the pips had fallen in pitch, and created something of a national outcry when he wrote to a national newspaper on the subject.
As it happens, the pitch of the pips was only ever approximate - the tone was derived from a very simple oscillator circuit, which was prone to drifting a little in frequency as it ages.
However, since the move in 2006, the audio generation is rather more sophisticated, and a drifiting pitch is far less likely.