Dating a Radio

 There are many features which can help to date a radio, including the types of valves (or transistors), type of circuitry, dial markings and even cabinet style. If you know where to look, an exact date of manufacture may often be found and for specific models their date of introduction will be available. Assuming one does not have specific information to hand however, the following details may prove useful.



click the valve to see a selection of valve bases

 Valves may be categorised by their type numbers and their bases. Sets in the UK, Europe and the US tended to use different types of valve but, for simplicity, I have concentrated on UK sets and main trends only:-
Early receivers used the 4 pin B4 base up to 1932 generally, a little later for some portable sets, but much later for rectifiers.
B5 was introduced to cope with the requirements for additional electrodes and was used up to 1932 or 1933.
Then followed the old 7-pin B7 base (and the rarer B9 in 1934, used for frequency changers) which was common over the period from about 1932 to 1937.
A few types of chassis employed the American UX style, with bigger heater pins, from 1934 to the mid 40s (including early models of the HRO).
The Side Contact (Ct8 or 8CS) appeared occasionally from 1934 to as late as 1946.
The most common bakelite valve base is International Octal, which appeared circa 1937 and, for specialised purposes, is still used today in expensive Hi-Fi amplifiers.
Mazda Octal, similar to IO, was used in specific makes such as Murphy, from 1938 to the mid 1940s.
All-glass types were introduced just before the war; B9G, from 1939 to 1950, was typified by the EF50.
The B8B (loctal) was more popular in the US from 1938 but was usually encountered in the UK from 1946 to 1952. Many of these valves had 7 volt heaters and were specifically designed for car radios, wired in pairs, where a charging battery delivered around 14 volts. The metal rim with its pip, which located in the valveholder, skirt, prevented the valve coming loose in its car radio application.
B7G was very common just after the war and was used commercially from 1947 to the end of the valve era. Some portable sets made after the war used IO valves probably on the grounds of price, but very quickly changed to the B7G series until the advent of transistors.
B9A (noval) was developed a little later than B7G and was very common from 1951 to the end of the valve era and may still be found, typically as a 12AX7, in high priced Hi-Fi.

The all-glass B8A valve with a side pip was used in later valve radios and followed on from IO types in the period starting about 1947 and was probably favoured because it didn't come loose, being held in place by a springy holder gripping the glass pip. B7G and B9A types required a screening can to hold them secure.
Generally speaking a commercial set (as opposed to a military or professional set) will not date before 1947 if it has all-glass valves. Bakelite based valves were used in all pre-war sets but some manufacturers continued to use them generally up to about 1953 although certain rectifiers and audio output valves were used in receivers very much later.



 Transistors were initially germanium (Ge), and the more common types are characterised by a leading "A" or "O" in their nomenclature. Although it was invented in 1948, the transistor was very expensive and therefore only commonly used from about 1956/7. The first transistor set was the Regency TR1 made in the USA in 1955 and the first UK set was the PAM 510, made by Pye in 1956. Germanium transistors in RF stages usually required neutralising to combat positive feedback due to high interelectrode capacitance, and early transistor IF stages are prone to oscillation if incorrectly tuned or if components having different characteristics are substituted. The shape of a transistor will often indicate its original design date. Early types, not surprisingly, look old and may be "top hat" devices, made by GEC, or the fat four-lead Mullard types such as OC170.

The silicon transistor (Si) introduced in the 1960s, was basically immune to the undesirable temperature effects, such as thermal runaway and critical biasing characteristics of the Ge types but were much more expensive than their Ge counterparts. Germanium transistors were still used, for cost reasons therefore, up to about 1970. After a period when Ge spares were difficult to find, these seem to be once more available in limited types. Roberts used a combination of Ge and Si in that year (1970) but, for example, Grundig were still using Ge in 1975, nearly 20 years after their introduction. The manufacturers presumably chose the cheapest device available consistent with ease of manufacture; higher cost Si types for easy RF alignment and cheap Ge types for low cost audio stages.

Later the silicon transistor, its number usually starting with a "B", at least in Europe, took over universally when its price dropped, because it was a more stable and resilient device, particularly being not as critical to use in RF amplifiers. Early types have metal cans but the more modern types have plastic cases. If you have a Japanese transistor radio you will find its transistors are typically marked C1302 or A1173 (the lower the number the older the transistor). The markings are shortened from 2SC1302 and 2SA1173. The "2S" is normally dropped and the leading letter "A" or "B" means the device is a PNP and "C" or "D" means NPN. The (missing) figure 2 represents the number of legs less one. A 3SK104 would be a FET with 4 legs. Here, K is N-channel (and J is P-channel). The "S" of course stands for Silicon. American transistors often use "2N". Here the lower the number the older the device but, unlike Japanese coding, the code is totally anonymous and provides no clue as to its type. Beware of small European transistors which sometimes adopt a code such as "C107". This isn't "2SC107", it really means "BC107" but there wasn't room on the case to fit it all in and still be large enough to read! On that topic "surface mount" transistors are the very devil to identify as abbreviations are rarely unique (and sometimes not even used).
Integrated circuits started to become popular in the 70s and some portable radios made use of them, for example the first Roberts RIC series, introduced in 1969, used the Mullard TAD100 chip. Ic's often have a date code stamped on them which gives a clue to the radio date.

Dial Markings


An early dial with no station markings

Early sets had a tuning dial marked with something like 0 to 100 and a card was sometimes fitted to a set on which the owner would enter a station name with its dial reading. This will provide a clue if this still exists with the original information written on it.
Later sets used a dial calibrated in wavelength often with separate long and medium scales but had no station names.


Station Names on Dials


 Station names began to be added about 1932. A host of continental stations are often included but I have done no investigation as yet on these. In terms of UK stations, these were introduced progressively over the pre-war years. As a guide the following list is included, from which a receiver's birth date may be deduced. Note that manufacturers would be given advance warning of transmitter completion dates and would commence manufacture prior to the opening date and so, for good measure, a year may be knocked off the following dates:-
For receivers with stations marked on its dial, take the latest date of the marked stations and subtract a year. For a handwritten list you may be able to use callsign information or wavelength.
Aberdeen (2BD), Nov 1922, 361m
Belfast (2BE), Sep 1924, 306m
Bournemouth (6BM), Oct 1923, 326m
Birmingham (5IT), Nov 1922, 384m
Cardiff (5WA), Feb 1923, 353m
Chelmsford (5XX), Jul 1924, 1600m
Chelmsford (5SW), Nov 1927, 255m
Daventry (5XX ex-Chelmsford), Jul 1925, 1554m
Daventry (5GB), Aug 21, 491m
Daventry National (5XX), Oct 1929, 1554
Droitwich, Oct 1934, 1554m
Dundee (2DE), Nov 1924, 294m
Forces Program, Jan 1940, 342m
Glasgow (5SC), Mar 1923, 405m
Home Service, Sept 1939, 394m...replaced National & Regional Programmes
Hull (6KH), Aug 1924, 294m
Leeds-Bradford (2LS), Jul 1924, 250m
Light Program, Jul 1945, 251m
Liverpool (6LV), Jun 1924, 297m
London (2LO), Nov 1922, 361m
London National, Mar 1930, 261m (Brookmans Park)
London Regional, Mar 1930, 356m (Brookmans Park)
Manchester (2ZY), Nov 1922, 384m
Midland Regional (5GB), Oct 1929, 479m
Newcastle (5NO), Dec 1922, 312m
North National, Jul 1931, 301m
North Regional, Jul 1931, 479m (Moorside Edge)
Nottingham (5NG), Sep 1924, 275m
Plymouth (5PY), Mar 1924, 350m
Scottish National, May 1932, 288m
Scottish Regional, May 1932, 376m
Sheffield (6FL), Nov 1923, 272m
Stoke on Trent (6ST), Nov 1924, 294m
Swansea (5SX), Dec 1924, 294m
Third Program, Sep 1946, 460m
West National, 1933, 256m (Washford Cross)
West Regional, 1933, 285m (Washford Cross)

The wavelengths listed above are approximate as frequency rather than wavelength was the parameter by which transmitters were tuned. In the 1926 Geneva Plan (and the later Brussels Plan of 1928) separation of medium wave stations was agreed as 10kHz but at Prague in 1929 this was reduced to 9kHZ in order to squeeze in more stations (a total of 106 channels between 200m and 550m). In the US however station separations to the original Plan were left in place. In 1931, because of interference problems, Europe was about to change its Long Wave channel spacing to introduce 11kHz spacing.

Interestingly, during WWII the BBC swapped transmissions around in order to confuse enemy bombers. This was often carried out by changing only one broadcast, switching to an alternative transmitter prior to an air raid.

Station Names on Dials

from 1950

 The Copenhagen Plan was introduced by the BBC in Mar 1950 as follows:-

Home Service

434m or 692kHz (North West, N.Wales, Yorks, Notts, Derby and Lincs)

371m or 809kHz (Scotland)

341m or 881kHz (Wales)

330m or 908kHz (London, SE and Home Counties)

285m or 1,052kHz (West, I.O.W., South Coast),

276m or 1,088kHz (Midlands & Norwich)

261m or 1,151kHz (N.I., N.E. & Scottish Border)

206m or 1,457kHz (Somerset, S.Gloucs, S.Hants, S.Wilts)

202m 1,484kHz ..started in 1952 (Barrow, Ramsgate and Montrose)

Light Program

1500m or 200kHz

247m or 1,214kHz

Third Program

464m or 647kHz

194m or 1,546kHz

American receivers usually employed dial markings based on frequency(kHz) rather than the usual wavelength (metres) of British manufacturers. It is slightly confusing as they dropped a digit i.e. 200m would be 150 rather than 1500, and 500m would be 60 rather than 600. American receivers for their indigenous market normally didn't have a Long Wave, as over there, Medium Waves were the norm though US receivers made for export usually did have Long Wave capability albeit not marked in metres or should I say "meters".

VHF receiver dials

 If one wants to date later receivers, particularly incorporating the VHF FM band, it will be noted that the VHF band got progressively bigger as time went on and as Police and other services were moved to alternative bands. The first VHF broadcast transmitter in the UK, Wrotham, opened in May 1955 and in that year the first receivers were made, tuning 87.5 to 100 MHz. One can use the listing covering makes and models covered by the Radio & Television Index elsewhere on this web-site. Be warned that although some manufacturers changed their models every year, other models were made over a period of several years. Later the VHF band was increased to 104MHz then finally to 108MHz.


Dates on components

 Buried in the innards of an old receiver may be precise details of its manufacture. Of course it may have taken a month or so to get a capacitor from one factory to another so add some time for this process. In a mains set the smoothing capacitors often have the date stamped on them and judicious use of a torch and small mirror may reveal this information. Also check any small electrolytics such as at the cathode of a valve as this may also have its date of manufacture inscribed on its case.
Some loudspeakers have the date of manufacture on the back or side of the magnet housing although (if you are lucky because the speaker will be immaculate inside) you might have to untie a black cloth bag to see the details.
Some manufacturers use date stamps on the chassis or on a label fixed to the inside of the cabinet. Upend a radio (make sure there's nothing loose inside first) and see if there's a date on the underside of the cabinet. Look for codes ending in the last two digits of the year of manufacture. Repaired sets might have later dates on parts.


Superhet or TRF

 Early sets used a "straight" technique, that is all tuned circuits were at the incoming signal frequency. As more stages were added it became increasing difficult to track the stages so that maximum performance was achieved across the whole dial. In order to improve gain, positive feedback was introduced. The amount of feedback was set by a "reaction" control and could be adjusted from zero additional gain to the point where the set burst into oscillation or "howled". At this point one could receive CW but AM broadcasts were muted and the accompanying whistle could only be eliminated by setting the frequency to the dead centre of the broadcast. When a set was oscillating all receivers in the neighbourhood could pick up the carrier so being generated by the offending set and, if any other set was tuned to the same broadcast would whistle in sympathy. However, at a point immediately before oscillation took place the incoming signal would be much augmented, becoming a lot louder. Unfortunately the nearer to the critical point one set the reaction control the more limited would become the perceived bandwidth of the signal. Fidelity would be lost and the broadcast would be deprived of higher audio frequency content. The reaction control setting varied depending on the setting of the tuning dial and this required two handed operation if one wished to look for stations over the band. Also, as the incoming frequency rose higher and higher, the less the amplification available from the valves, making weak signal short-wave reception difficult. Very strong stations would often break through and be omni-present because selectivity in the TRF set could be poor.

The use of reaction was so annoying that the government stepped in and insisted that every new set offered for sale in the UK must be tested and passed for use. This testing included, not just the overall performance of the set, but the undesirable effects of a reaction control.

Even before the 1920s Armstrong developed the superheterodyne receiver. This technique shifted the bulk of amplification to a fixed intermediate frequency (IF) by mixing it with the output from a local oscillator, and of course had a number of significant advantages. The IF stages could be adjusted for maximum gain which was independent of the incoming signal and did not need to be readjusted as the receiver was tuned. The IF chosen was at a low frequency enabling a good gain figure to be achieved from the valves and of course interference to neighbours receivers was eliminated. Early superhets used frequencies of around 100kHz whilst later sets used around 465kHz. There was one major disadvantage however. At a given dial setting the receiver would be able to pick up more than just the wavelength indicated on the dial. The set would also receive an image signal which was twice the IF away from the wanted signal. It would also be possible to receive various other signals which were related to harmonics or multiples of the local oscillator as well as straight-through signals if these were strong enough. To avoid problems such as these the superhet must have a measure of selectivity before the mixing process and the degree of this selectivity necessary depends on the incoming signal strengths. For short wave reception where the relative strengths of wanted and unwanted signals can be extremely high it may be necessary, in order to achieve suitable selectivity, to raise the IF from 465kHz to 1.6MHz, 10.7MHz or higher.

The date at which TRF sets disappeared and the superhet took over is a bit woolly. A rough guide is 1931. Before that date a set would probably be a TRF set and after that date a superhet. The date can be plus or minus say 3 years. In 1929 the superhet was referred to as the "Rolls Royce of receivers. The receivers chosen IF is a pointer to a superhets age. Again the changeover date is blurred. One might say that all sets in 1938 used around 465kHz, as did most sets in 1937 and some sets in 1936. See my pages on IF's for pre-war receivers elsewhere on this site. Why did it take so long for superhets to gain favour? Lots of reasons including the fact that the principle was patented.


Mains versus battery

 Before about 1930 most sets were battery powered, using an accumulator for the valve filaments and typically a 90 or 120volt HT battery. Usually there was a grid bias battery with tappings to 9 volts. Its use would enable the set to consume less power by reducing anode currents through the expedient of introducing a negative bias to their control grids.
Mains powered sets started to become popular around 1929. By that date the country still had a mixture of AC and DC mains. Mains voltage was not standardised and varied from say 100 to 250. Some mains sets used a range of valves with higher than normal heater voltages, typically 20 and these were suitable, in series connection, for use with DC mains, however most early mains sets used a mains transformer which would be useless if you had a DC mains supply. The early mains valves had 4 volt heaters but these were gradually replaced by 6 volt types from about 1937. Some mains valves even had directly heated filaments which made them useful only in large signal circuits. A thing called a "humdinger" appeared for reducing hum and was basically an adjustable centre tape for grounding a directly heated AC filament. Battery powered sets were still in common use in the years before the war and it was common for bicycle shops to repair radios because that is where one took the filament accumulator to be charged.


comments welcome

Return to entrance>>