Early Radio-Page 2

Development raced away

 Once a little of all the mysteries had been investigated and Marconi had showed that worldwide communication was feasible, radio development raced away. New detectors, proper tuning coils, proper aerials, and the triode valve amplifier all helped accelerate the process. Soon, just before WWI, a radio system would be developed using a proper carrier instead of a spark then amplitude modulation using speech, and finally broadcasting.

Falling by the wayside were mechanical detectors, and embryonic spark radio receivers and transmitters, with their large components and festoons of wiring, looking more like spider's webs, and something more at home in a Physics Laboratory, than advanced communications electronic equipment.

Taking their place, in the home, were crystal sets and valve receivers built in beautiful mahogany or oak cases and using black ebonite panels.

Shown here is an Ediswan crystal set from 1923. Despite its extreme simplicity it still cost a whapping £4:7:6d plus BBC Tax of 7:6d.

It covered 200m to 500m, and with its statutory 100 foot aerial, had a range of 25 miles.

 The first valve receivers (or "straight receivers") had low gain but soon developed "reaction", first discovered by Armstrong around 1914, which enabled them to drive a loudspeaker instead of the usual pairs of headphones. The reaction control coupled some of the output back to the input and in so doing made the receiver inherently unstable, but it had been found that just before the set burst into oscillation, it's gain had risen to a very high degree and because of this it was able to provide enough power for a loudspeaker.

To use an early reaction set was not always a simple matter as the larger types needed maybe three or more controls to be adjusted simultaneously for best results. It was not uncommon therefore for at least one set in the neighbourhood to be oscillating merrily, usually unbeknown to its user.




The Radionette V2 of 1923 (left) with a Magnavox loudspeaker (centre) from a year or two earlier.

With its 2 valves it's range was guaranteed to be 40 miles or 25 miles on loudspeaker.

If you were so inclined it could drive no less than 12 pairs of headphones so could provide entertainment for the largest family!

Price, less speaker was £10 guineas including BBC tax but without valves.

If you couldn't find suitable black market valves then Mullard types, together with an HT battery, insulators, aerial wire and headphones, could be supplied for an additional £4. A suitable filament accumulator would be around 15/-

As far as the loudspeaker was concerned you could check the stock situation by telephone! If you needed to ask the price you couldn't afford it!

This collection of bits and pieces (above right) made by KB was the ultimate receiver available at the dawn of broadcasting, around 1921.

Different units could be wired together to make your dream wireless.

Where would you put it? Once you'd wired it up it would be pretty difficult to move around.

 The radio magazines of the day were full of advice concerning "reaction" and tried at length to educate the reader to get the best from their set without oscillating.

The problem was that an oscillating wireless was a pretty efficient transmitter and resulted in a strong local carrier being super-imposed on the broadcast band.

This carrier would beat with the associated incoming broadcast station, to which the set was tuned, and produce a howling sound in any receiver tuned to the same station within a range of up to several hundred metres or more.


Enter the Superhet

 In the mid-20s, many years after 1918 when Armstrong had first demonstrated his superhet receiver (the name being abbreviated from "Supersonic Heterodyne"), straight receivers which had all the HF amplification at the signal frequency slowly became obsolete.... however, as always, what was practical slowed down the process so straight receivers were still being generally sold for as many as 8 years later.

Why did the superhet take so long to appear in Radio shops? When Armstrong invented the technique, valves were pretty useless for amplifying high frequencies. they were OK for wavelengths of say 1000 metres but gain dropped off in the medium waveband. Having a fixed amplifier at say 50kHz (corresponding to 6000 metres) was the answer. Lots of gain could be achieved at this frequency but at the expense of additional valves, for example a heterodyne oscillator.

Two things conspired against the introduction of the superhet though. First was better valves which could amplify the higher frequencies much better than before. Second was the patent situation. In those days a patent was granted to ensure that the inventor made some money from his endeavours. The life of a patent was 16 years but the intervention of WWI often was considered justification for the courts to grant an extension of 4 years. Why should manufacturers switch to superhet designs when valve performance was so much better? And of course royalty payments to Marconi were already raising the price to the general public without the extra which would have to be paid to Armstrong.

So superhet development was effectively put on hold for 16 years. Until the interest in short waves and long distance reception pushed requirements of cheap valves beyond their limit; the superhet patent lapsed and one knob tuning was too great a benefit to be overlooked in a very competitive set market.

A superhet had two major advantages for the user.

First it had one knob tuning. The design allowed for sufficient gain to be developed without reaction being necessary so the whole broadcast band could be searched whilst the set was operating at its maximum efficiency.

Second, because there was no reaction your neighbours wouldn't suffer from howling.

True the set had an oscillator, necessary to produce the intermediate frequency whence the main amplification was carried out, but this was conveniently placed some hundred or so KHz away from the station being received and was anyway not directly connected to the aerial as in the case of the straight receiver.

Before the mid 30s the TRF receiver was obsolete, despite the last ones being given exotic names to hoodwink would be purchasers; although for some odd reason they made a small comeback in kit receivers advertised in the 50s.


 What about aerials?

An aerial, generally speaking works best when it is about half a wavelength long.

A long wave aerial would need therefore to be 750 metres in length.

One cut for 30,000 metres, which was the longest wavelength used in the 20s, would need to be 15,000 metres (or about 10 miles) long and even for medium waves you would need say 200 metres of wire.

To prevent the land becoming entangled in a sea of aerial wires the government imposed a maximum aerial length of 100 feet.

Two common ways were used to overcome this restriction.

First, if you were fortunate enough to have a telephone, one could connect your wireless to your telephone line.

Second, you could buy an "aerial eliminator" which was merely a hefty capacitor which connected the set's aerial socket to the mains wiring of the house.

In both cases these methods, although potentially providing many miles of aerial wire, tended to pick up electrical noise which degraded the program you were listening to with bursts of static or crackling.


What could you hear?

 In April 1923, a typical list of stations broadcasting, together with callsigns and wavelengths is given below


369 meters


400 metres


385 metres


425 meters


415 metres


353 metres

Croydon Airport

900 metres


2,600 metres


2,800 metres

The Hague

1,085 metres


1,100 metres

Radio-Electrique, Paris

1,565 metres

School of Posts & Telegraphs, Paris

450 metres

Bar Lightship, Liverpool

900 metres


900 metres

Le Bourget

900 metres


900 metres

The BBC transmitted programs for an hour each morning and for around 5 hours in the evening.

At other times one had to be content with even shorter schedules from foreign stations, ship to shore traffic or the various high power transmissions to aircraft from the international airports.


 Although the mathematics of valve radio were available to anyone that wished to carry out the necessary calculations, development in the 20s was very much trial and error.

Not only were the parameters of radio circuits largely guessed at, the operating frequencies of transmitters were very much governed by circumstances than by science.

In the early days of broadcasting two wavebands were being used.. "Long" and "Short".

Not what one would understand today however.

Long waves were something like 1,000 metres to 30,000 metres and short waves were 400 meters to 1,000 metres.

Don't be surprised if an old radio set to "short waves" peters out before Radio One is discovered.. and you'll have to wait until the mid 30s to guarantee finding a true short wave band.

As the number of broadcast stations increased, and sets became more sensitive, the problem of interference reared its head. Transmissions were very wobbly and many were not crystal controlled, not to mention wideband spark transmissions that were still encountered and which were to be still legal as emergency sets on ships for many years.

Day to day variations in transmission frequencies of particular stations were very large and, as better transmitting valves were produced, powers were beng steadily increased.

Bearing in mind that the whole of the known broadcasting spectrum occupied a mere 750 KHz there was clearly a limit to the number of stations that could use the spectrum without problems. If one allowed 10KHz per station then 75 could be fitted in. Allowing some space in between and the number reduced maybe to half this number.

The first thing to happen was that radio amateurs, who had been allocated the useless frequencies below the broadcast bands, were pushed out to make space for broadcasters to occupy extra wavelengths, initially down to 300 meters and then to 200 metres.

Surprisingly, but as the long-suffering radio hams had already found, the lower wavelengths were actually rather good.

Less power was needed to provide a local service, and much to everyone's surprise, after dark the bands truly opened up and one could listen to broadcasters all over Europe and even from the US.

The drive for more space and to utilise shorter and shorter wavelengths continued.

The proper short wave band was discovered and the old short waves were re-named "Medium Waves".

Whilst Marconi was building his huge short-wave transmitting stations to span the world, experimenters were struggling to find enough gain from their old valves, and battling with hand-capacity effects when attempting to tune in hitherto undreampt of exotic stations in the new 40 metre band.

The public was bombarded with dozens of DIY magazines, many published weekly, to assuage the demand for information about the new technology. This fuelled the demand for wireless sets and in turn led to a huge growth in broadcasts which were fast outstripping the space available. Eventually International Regulatory bodies were instituted.

By the end of the 30s a number of conferences had been established setting out strict guidelines within which broadcasters were obliged to operate. In particular, station frquency management was imposed directing the multitude of broadcasters to transmit on specific wavelengths designed to maximise their numbers and at the same time reduce interference between them.

Several plans were formulated over the years and it is interesting that these differed fundamentally; for example between the Europe and the US. The plans had to be pragmatic and to take account of existing transmitters and previous plans whether these had been formulated by trial and error or by calculation. To this day the channel spacing across the medium waveband is "scientifically sound" in Europe but "practical" in the US.

The last significant change was highlighted when the Light Program on 1500 metres was moved from precisely 200KHz (where incidentally it was classed as a Secondary Frequency Standard) to 198KHz when European Long Wave channel spacing was finally settled. No such problem in the US as Long Waves had never caught on. Strangely the US has always employed KHz and never Metres whilst in the UK it was the reverse.


What was the cost of a Wireless Set?

 One of the things, of which we can now have little concept, was the cost of owning a wireless. Not only was the BBC license relatively expensive but, in addition, fees were payable to the BBC and to Mr.Marconi who had cornered the broadcasting patents. Where he hadn't invented a specific technique, he purchased the rights from the original inventor. For example, one pretty important feature of radio, that is "tuning", he purchased from an Engineer at Liverpool University. That must have been a significant coup!

Why was the crystal set with its fiddly "cats whisker", very weak output and no real means of excluding unwanted stations, so popular when valve sets were available in abundance?

For a start, a royalty was levied on each valve in a set and to this end kits were often sold without valves in order to cut the cost.

Why? Well various cartels (that is.. groups of people who had cornered the market and artificially setting prices well beyond that reflecting true value) were in operation to ensure the public paid through the nose for everything associated with radio. As with most restrictive practices of this nature, there was of course a black market of "under the counter sales" and through this, foreign components including valves were available at much reduced prices.

What sort of price did a law-abiding citizen have to pay for his wireless in terms we can understand today?

Well if you equate a good 1924 radio sold for £15 with a good 2002 computer sold for £1500 the comparison in terms of wages holds good. The wireless was not just a plaything... it represented a pretty substantial investment for the average family.

A large proportion of the population could not afford to buy a wireless outright so either relied on a crystal set (still not cheap) bought a kit of parts, so they could build their own wireless, or perhaps paid for their purchase weekly on the "never-never".

This Cossor design was one of the more sophisticated kits.

The model 347 was designed for AC mains use only and used a lot of in-house Cossor parts. This enabled the Company to keep the price at the lowest level by cutting out the middleman.

The kit was sold for a little under £9 and included a factory built power supply. This reduced, but not totally prevented, the danger of electrocution of the constructor and would certainly reduce the number of returns from incorrect power supply wiring.

How long would construction take?

The pamphlet says "A pleasant evening's work".

Well I suppose there wasn't the distraction of a TV in the background... and the radio hadn't been built yet!
The end of steam radio?

 In the 21st century, Morse code transmission, the last remnant of the early days, is almost defunct and analogue tuning is almost looked on as quaint.

Taking over is digital radio which is no longer detectable on an ordinary receiver.

Radio Luxembourg (was) transmitting, not just on 208 metres over a path of 400 miles and megawatts of power but over a satellite with a wavelength of nearer 20.8 millimetres over a path measured in tens of thousands of miles and with a power of just a few tens of watts.

Military systems use "coded modulation" and "spread spectrum" to which an ordinary receiver is completely deaf, transmissions appearing only as a miniscule increase in background noise.

How long will it be before broadcasting as we know it comes to an end and we can no longer twiddle the dial of a short wave receiver and listen to those fading and echoey distant stations from the other side of the world?
See Histories