One of our members is looking for a radio piece so he can repair one of his radios. Contact him here or find him playing at

(First published in the BVWS Bulletin, Winter 2004)

See also Weymouth Vintage Wireless

There are pictures of these wireless sets here

The Weymouth and Portland area has been known for high-tech industry for the past hundred years. The increasing degree of sophistication required by the Royal Navy – which was based in Portland Harbour – and the Admiralty Research Establishments provided the impetus for a cluster of advanced technological companies to gather here. There was certainly some interaction with specialists in the district.

This is an attempt to investigate the local response to the country-wide wireless craze of the 1920s, when almost every locality had its share of small retailers/manufacturers. Most of these were on too small a scale and, being unable to compete with better organised national firms, had sunk into oblivion by the nineteen-thirties. Such a fate befell those under review here.

Weyrad, whose coils and kits live in the fond memory of many readers, are outside the scope of this article as they appeared much later.

Of particular interest are four three-valve wirelesses, two made by Smith’s, and one each by Marshall’s and Bennett’s. Marshalls also made a two-valve set. The names were mentioned in a page in the Dorset Evening Echo in 1996; this paper is strong on memorabilia and receives nostalgia from older residents.

It was while I was working on the circuits of two of the sets that I began to ask deeper questions than ‘how do they work?’, because they seemed to be quite straightforward TRFs. Further investigation however revealed that each company had an individual approach to tuning the sets though their output stages were perfectly orthodox. They appeared to have such an interesting tale to tell that I realised that there would also be a fair amount of historical research involved.

Researching documents and books

First stop was Weymouth library and Kellys Directories for the 1920s and 1930s. Kellys are the well established sources that contain personal and trade sections grouped under categories. A sequence of these can tell the life of a company from beginning to end

‘Smith, W and Son, electrical engineers, 7 Royal Arcade’ was listed in 1923, but they had already been a respected electrical company for twenty years. Their premises is now one of the tourist shops on the Esplanade. In 1929 they also described themselves as ‘makers of the Smith Majestic Musical Equipment’ but by 1936 had reverted to simply ‘electrical engineers’. That was the last entry for Smiths.

Also in the 1923 edition, for the first time, is ‘Marshall, Ernest, watch maker’ which was indeed his profession and why he opened a business in Portland. By 1927 he described himself as ‘Marshall, Ernest, electrical, mechanical and radio expert and watch repairer’. Entries for Marshalls continued through the 1930s.

The first appearance of V.H.Bennett the Department Store, in this category, was in 1927 when electrical engineering was listed with other retail items. It was included in ‘wireless dealers’ in 1932 but in 1936 ‘Bennett & Escott Ltd. Registered electrical installation contractors and wireless engineers’ appeared.

I had long suspected that they had devolved in their own right from the parent company. Verification came from an advert in the Southern Times dated September 6th 1924, announcing that Mr T.H.Escott ‘will personally superintend all repairs and give free estimates for new work. Special attention given to Wireless installations’.

Trawling through the Southern Times of the 1920s was a long, neck-aching chore peering at an unreliable micro-reader, as the paper has been photographed onto this format. I was hoping to discover when these wirelesses were introduced, as surely any self-respecting manufacturer/trader would want to inform people about their products. But no. I find it almost incomprehensible that there was such a lack of adverts, that the firms didn’t bother to advertise at a time when this new and accessible technology was sweeping the country, and trade was there for the taking, but that seems to be the case.

My search did turn up some interesting other evidence though: at the time of the renowned South Dorset Radio Society’s 40th year in 2001 I had investigated its early days, linking the SDRS with the Weymouth Short Wave Club of pre-war days. It was quite a surprise to find that there had been a South Dorset Radio Club in the 1920s – they had their annual dinner in June 1924 when the guest of honour was Capt. P Eckersley of the BBC. He was pleased to see so many aerials in Dorset, a sign of the growing popularity of wireless.

There was an advertisement placed by Fredk. Young, an engraver. He is known to have engraved radio panels for both Smith and Marshalls.

There was also a very slight hint that another company may have been involved: in June 1925 an advert by D.N.Menzies of Weymouth mentions sets with one, two or three valves at between £3 and £15, plus one called a ‘Super III’ at £17; tantalisingly, it doesn’t say who made them – but one is tempted to suppose home-made, to order, by Mr Menzies. Considering that the shop did not advertise again, nor is there any further record of its existence, perhaps there was a problem. About the same time a Dorset correspondent to Amateur Wireless was quoted by ‘Southern Times’ columnist Radio Rex:

Hideous Disturbance: It is usual to saddle the hapless listener with the whole of the responsibility for the oscillation nuisance, and one who has become acquainted with the methods of certain ‘manufacturing’ retailers feels constrained to ask whether they have not got something to answer for. A single valve circuit evolved by a retailer…. was a vile contraption, bearing no indication of its origin, of course. The components were crowded on to a panel measuring about 3in by 9in and the circuit was probably never intended for expert eyes. Closer inspection of what appeared to be an innocuous variometer revealed a very crude form of reaction at the aerial, and a seasoned experimenter could not tune the thing without creating a hideous disturbance.

Research of this nature provided a good deal of background material; the radios that I could now investigate at first hand, and the personal sources still available to me, enabled me to piece together the history and technicalities of the local wireless industry.

Marshalls’ of Portland

This is a well known Portland electrical store, and has existed for as long as anyone there can remember. I made contact with Frank Marshall. Frank – still active as G2XQ – was a mine of information about the Portland sets in particular and local radio matters in general. Indeed, it was Frank the hoarder who had unearthed that wonderful animated movie by W. Heath Robinson advertising Amplion speakers His elder brother was Harry, who with his father Ernest developed the Portland series of radios. Harry made some tape recordings of his memoirs which have been invaluable.

Now in his eighties, Frank remembers that as a small boy – in the ’20s – he helped to wind some of the coils that went into the sets his father and Harry had designed .

The circuit details of the Portland Three were described in Harry’s tapes. It seems probable that the set was a development of a design in Amateur Wireless by R.W.Hallows.

A detector valve was followed by two LF amplifiers – all triodes of course – (this configuration is described as 0-V-2); aerial matching was achieved by 7 tapped studs. The swinging reaction coil originally had 2 or 3 tappings with extra ones on the loading coil. Harry added another reaction coil in series with the swinging one which enabled good oscillation over a wide range including Long Wave – presumably in response to the popularity of the new Daventry station.
The controls were simple – once you tapped into the required band, you fine-tuned with a variable condenser, and used the reaction to set the volume. To general astonishment, it was very successful and achieved a good reputation locally. These sets often emitted ‘hideous disturbances’ to the annoyance of neighbours who would retaliate by setting their own sets in oscillation!
The cost was £16, plus loudspeaker which could be from 30/- to £25 (Brown Super-duper was recommended). They were made to order in batches of about six. This so undercut Mr Smith’s rival set that he wrote a sharp letter to Mr Marshall accusing him of infringing Marconi’s conditions of licence by emitting too much radiation due to reaction acting direct on the aerial coil. Marconi, with whom Marshall took up the case, said there was no such regulation.

Although the centre of interest was the Portland Three, Frank also recalls that there was a long-ish set, perhaps a portable. We have no evidence of this so can only hope that one will turn up – as long as it has escaped the jaws of the skips!

Frank did not know anything about a Portland two-valve set – until I heard of a facsimile of a 1926 advert from the Royal Manor Times, a now defunct Portland local paper. There, clear for all to see, is mentioned a Portland II, complete with loudspeaker for £11-10-0: this vindicated Ray Henville’s belief that two empty cabinets in his possession were for Portland IIs. Ray is a collector living in Blandford Forum, who has been both helpful and kind.

In the end galloping technology, screen-grid valves, dual-ganged condensers and cheaper, better mass-produced sets rendered Marshall’s enterprise uneconomic.

Dating the Portland sets

Harry in his memoirs is quite certain that the Smith upright set was on the market first and that it was the one with double doors.

The relaxation in 1924 of the regulation that the reaction circuit should not cause the aerial to radiate may explain Marconi’s reply following Smith’s complaint. The date of the relaxation along with the probable 1924-25 date for the Smith III could help fix the Portland III at not earlier than 1925, i.e. after the ending of the GPO registration scheme, which would account for the absence of a number or stamp. Every indication is that the P-3 was Marshall’s first foray into radio production. The advertised P-2 (October 1926) would have been an addition, or replacement.

A further clue could be in the few recognisable components. When were a Brandes VLF tuning condenser (0.0005uF) and a Gecophone BC 720 ratio 2/1 transformer available at the same time? And, when was the R.W.Hallows design published in Amateur Wireless?

The Smith Three-valve radios

Only two sets made by W. Smith & Son are known to exist. The earliest is a model in the style of a smoker’s cabinet, with full-length doors; there is also a sloping front model.

The Smoker’s Cabinet Model (the Upright)

If ever someone was in the right place at the right time, this is an amazing stroke of luck. The South Dorset Radio Society’s chairman a few years ago, Bill Young, filled time before an appointment by visiting a nearby antiques shop.

That evening he told me he had seen an old radio there, bearing the name ‘Smith’. My considerable eyebrows went into orbit with suppressed excitement and I think he understood that he should try to acquire it.
If Bill hadn’t visited the shop at that time, would we ever have known? Would some other customer ever have appreciated it? This proved to be a missing link. It is an upright model – similar in style to the Portland offerings, and by the same cabinet makers.

The circuit is a 1-V-1 using triodes. It was “too complex, but clever” (Harry), with abundant controls for dad to play with.
Medium or long wave transmissions could be selected by switching in or out tappings on a coil on the HF valve’s grid and pairs of coils in the anode circuit. All the tuning and reaction was accomplished with variable condensers (home-made, along with some other components), which had to be set independently and accurately. An audio transformer (Triumph B-G) coupled the output triode and ‘phones or speaker were connected as usual. The dials were made by a firm called Ormond. Also, a centrally-located meter allowed voltage checks on HT and LT.
The set was said to be heavily damped and stable but had poor amplification because in order to achieve selectivity the valves had to work at times with grids very positive.
Trying to work out a circuit diagram by tracing the wiring was tricky, because some deterioration had taken place in this set, and to this end a digital camera was worth its weight in gold! (The technique is: print a light colour picture of the chassis wiring, use an analogue meter (not digital) to confirm the wire runs on the set, and highlight the same on the picture with coloured pens.)
Not surprisingly, the set was expensive at £40, and the Amplion Dragon speaker was 5gns.

Dating the Smiths upright

The BVWS list of GPO registrations shows: ‘2093 Smith (Weymouth) III. 3-valve sloping panel’.

I believe that it was Ray who submitted this information to BVWS. At that time the only Smiths III he was aware of was the sloper in his possession. The new find, the upright, came as a surprise. It bears the GPO/PMG stamp which (assuming Mr Smith was a law-abiding trader) appears to date it to 1922-1924.

The GPO registration 2093 would have been granted to a two-valve set (hence the 2..) in late 1922 or early 1923. Timing is suggested by the registration number 2001 given to a Marconi V2 in November 1922. The Smith two-valver then gained a valve – probably the HF valve – but kept its GPO number and PMG/BBC stamp. If there was a two-valve set, it has regrettably gone for ever.

Marconi’s reply to the altercation mentioned earlier between Smith and Marshall over the matter of oscillation suggests that the reaction ban had been lifted, which dates both sets to at least late 1924.

A switch is labelled push for 5XX and Paris and pull for local and other stations. 5XX was the experimental station in Chelmsford which began in 1924 and transferred to Daventry in mid-1925 operating on long wave as a broadcasting station. The switch itself is a three-way rotary, with an off position and it looks very original. A blanked-off hole into the battery compartment may have been intended for an on-off switch. Clearly, the fascia panel was made before the modified switching arrangement, but how long before? And which 5XX was received in Weymouth – Chelmsford or Daventry? Why did the panel still bear the GPO number and the PMG stamp, as 5XX wasn’t commissioned until well after the GPO scheme was history. The figure ’25’ appears below the Smith scroll: it is more likely to be a serial number than a date.

It is probable that this set was made in 1925 using an old fascia panel from the stockpile; the ‘advanced’ technology enabled band selection by switch, rather than the usual coil-swapping.

The Sloping Front model

The three valves here are a screened-grid (PM12A), a triode detector (PM1HL) and an output pentode (PM22A) with a side cap.

The whole circuit is mounted on a wooden baseboard and an angled fascia, being easily demountable for servicing. It seems very well made, the square-section wiring being bent and cut and soldered to a good standard. An oddity is the Pye Differential Condenser, which is mounted awkwardly, as if it was an afterthought or a development of a now defunct set, as clearly the cabinet was not designed for it originally.

Smith again used fixed coils and variable condensers for reaction and tuning, with simplified switching selecting MW and LW via pairs of coils. Some components are labelled: a condenser ‘Polymet New York’ and another one ‘Hydraworks Berlin’ while the audio transformer is ‘Bear Brand No. 1000’.

What is its date?

This set contains only the registration number 2093 and no BBC/PMG stamp, though there is a serial number 1059. It must have been made after the stamp scheme was dropped. Hence, the entry in the BVWS list itself might need to be altered. The use of multi-element valves puts it to 1928 at the earliest, a notion that is reinforced by the provision of a pick-up input to the detector grid. I believe this set was made in 1928 or 1929; possibly the former as by 1929 Mr Smith advertised himself as ‘maker of the Smith Majestic Musical Equipment’ yet there is no sign of this on the set itself.


This was an attractive looking item. The cabinet – 16″ x 12″ x 12″ – with a hinged top lid looks similar to some advertised in 1927 magazines; it may well have been bought in, and there is no similarity to the cabinets made for Marshalls or Smiths. Could the circuit itself have been bought in? It is not clear whether it was locally designed but if so it was probably by T.H.Escott.

Its valve configuration was 1-V-1 using triode valves PM1, P220, HL210. Filament rheostats controlled valves 1, and 2 + 3. The wooden baseboard plus fascia slides out for easy servicing. The most noticeable feature is a tri-coil pack . These coils are in the aerial circuit, switched for band coverage, with the central coil tilting between vertical and horizontal, being energised from V2’s plate and coupling inductively into V3’s grid via an audio transformer. The switch also selects one of two coils (below the baseboard) in V1’s plate circuit. The five coils are labelled 50, 58-60, 150, Atlas Patent 75 plug-in coil, Finston lo-loss No.75.

We wondered if this was a regenerative set. It appears, however, that the moveable coil was part of an inductive neutralisation circuit which prevented the circuit from oscillating. Normally, the inter-electrode capacity in the valves gave rise to unwanted oscillations – the ‘Miller Effect’. It was possible to neutralise this by sending equal signals but in the opposite phase through part of the circuit. If the moveable coil is connected the ‘right’ way round, neutralisation is accomplished.

Fine tuning is effected by variable condensers one of which is impressed ‘Regd 71928’.

The audio transformer is the reliable Ferranti AF3.

Dating the VHB III

This set is full of paradoxes. Dating is difficult, but 1926 seems a fair estimate. As it has no GPO or other markings it presumably post-dated the registration scheme – always bearing in mind that there were several ‘pirate’ builders in business who didn’t belong to the BBC.

But it isn’t as straightforward: the tri-coil pack is similar to those illustrated in the Harmsworth Wireless Encyclopaedia which appeared around1923. When did these go out of use? This neutralisation circuit was also popular in the early twenties, and in those days technology and design progressed by leaps and bounds. As the vast majority of radios bought in the early twenties were crystal sets, and V.H.Bennett’s electrical division was announced in 1924, this set is likely to have been designed some time in 1925 and manufactured in 1925-1926.

T.H.Escott’s son Brian showed me a receiver his father had made in the early 1920s. I had hoped this would be a prototype, but it looked very home-made, and had three bright-emitter valves. There was no resemblance to the VHB III.

The fate of old wirelesses!

Some sets, made by illustrious manufacturers, are sought after by collectors. The less famous – products of the many local firms countrywide – have little commercial value. They are, however, priceless in the context of their place in the local economies. They were a response to the immense interest in the wireless craze of the time.

I would hope that wherever you are reading this, where some firm nearby but long ago had supplied their sets to their customers, there may be a few collectors aware of such wirelesses and pro-actively seeking them out. Don’t let them find their way into skips. Acquire, cherish, pass on – to a museum if appropriate but ensuring that they will be looked after.

Those I have described are the only known surviving examples: it would be helpful to know if any others exist.

Perhaps, under the auspices of a BVWS initiative, some of the surviving local sets now scattered far and wide around the country, might find their way back to their home district. Alternatively, a register of such sets could be established, so that even if the present owners aren’t aware of their origin, there might be somebody out there who can identify them.

Now for a personal ‘want’. An HSP, made in the ‘thirties by H.S.Phillips of Weston-super-Mare, any offers?

John Rose, M0BQO

There are pictures of these wireless sets here


 (First published in the July 2006 edition of Catswhisker)

A few years ago, members may recall, I wrote a piece in Catswhisker about the three small-time firms who made radios in the early 1920s (February 2003). They were Marshall of Portland, Smith of Weymouth, and V. H. Bennett of Weymouth.

A Smiths two-valve radio appeared on the internet during 2005. Unfortunately I didn’t get to know about it until too late, by which time it had been sold at the National Vintage Communications Fair in October 2005. The trader had sold it, along with a lot of other old sets, possibly to an Italian dealer and it is this lead that I’m trying to follow up, more in hope than in anticipation.

Now, it looks as if this set, whose origin the trader hadn’t heard of (‘Never heard of Smiths – probably a small company somewhere’) is in many ways similar to the three-valve set with a sloping front that we are familiar with, except for the tuning controls. Enlarging the picture shows a scroll on the fascia just like other Smiths panels, the name ‘Smith’ has to appear somewhere so why not in the scroll. The shape of the cabinet is the same (not forgetting that ready-made cabinets were available then – but a local cabinet-maker worked for Smiths and Marshall).

The reason for the excitement is the fact that it was a two-valve set. On the panel of the Smith sets we know the GPO registration number 2093 appears. This was granted to manufacturers by the Post Office when their sets intended for sale had passed certain tests, such as radiation control. A number that started with ‘2’ indicated a 2-valve set but makers were allowed to keep their number for subsequent developments in their range.

So, was this mystery set really made by Smiths of Weymouth? Does it indeed have two valves? If ‘yes’ then it could be the missing link – the prototype or one of the first production run to carry the GPO number. We can’t be sure until it has been traced and photographed, and questions answered.

Unfortunately interest in ancient wirelesses and support through magazines and clubs doesn’t appear to be as strong in Italy as it is in UK, so I fear news of my quest may not reach a sympathetic ear – but one lives in hope.

I’ll try to keep you up-to-date with it.