Associated Companies in over 175 towns in Great Britain and in Barbados, Bermuda. Canada, Ceylon, Guyana,
Hong Kong, Jamaica, Malaya, Malta, Singapore, South Africa, Trinidad and West Africa etc.
I can't remember how it all began. My first memory of anything remotely electronic is of being ill with rheumatic fever, and being bedridden for many weeks, at an age of about ten. A colleague of my mother lent me a crystal set, for use in bed. I can remember taking the lid off this receiver, and noting that the internal wiring was done with square-section tinned copper wire, about 16 swg, and with right-angle bends in the wiring. I wish I still had it!
So then I had to try making my own. I bought a detector by mail order (it must have been an early germanium point-contact diode), but I could never get the thing to work. And yet it was made with 'modern' components.
Among my bits of electrical junk was an old telephone magneto, and I discovered that by connecting this to a pair of headphones, I could receive the Light Programme. It was many years later that I realised I had been listening to the Rediffusion audio network, whose cable (with an earth fault on the 'B' programme) ran under my bedroom window.
Next ( while still at home)
Being dissatisfied with crystal sets, I decided to try valves. I remember buying an HL2 triode from Super Radio on King Street in Nottingham, and getting a lead-acid battery for the 2 V filament supply. In order to keep this charged, I built an elegant charger unit, using a 60 W lamp as a current limiter, straight off the mains, which I must have discovered was d.c. Not very efficient, and highly dangerous, with naked connections everywhere.
At first I used batteries for the h.t. supply, but later, while at school, I built an 'eliminator', which was just an enormous potential divider. I can still remember the sweetish smell of those 1 watt carbon composition resistors when they overheated (as they usually did).
The big problem at home was the d.c. mains supply. This (200 V) was about right for modestly-powered valve amplifiers, but the heaters had to be in a series string, and so I could not use really beefy valves: I was limited to KT33Cs. However, I built several amplifiers in those days, usually with barretters as heater droppers (big bottles with iron filaments in an atmosphere of hydrogen, which were supposed to stabilise the heater current).
The state-of-the-art equipment was a 10 W amplifier with KT33Cs, an Acos crystal pickup on a Collaro 3-speed deck, and a Whiteley WB12 speaker in a reflex enclosure in 'contemporary' style. (It was only much later, after I had moved away from home and was enjoying the new-found delights of a.c. mains, that I learned that the supply at home had been positive-earth, so that all the metal chassis units I had built had been 200 V live to earth; what a good job that we had good carpets.)
Whilst still at home, I was asked to make a hearing aid for my grandmother, who could not then qualify for a National Health unit. This used three 1.4V valves (I think two 1T4s and a 1S4). It was built in a flat cigar tin, about 6" x 4" x 1", and most of the space was taken up by the 60 V h.t. and 1.5 V l.t. batteries. However, it worked well (apart from the clothes rub) and Grandma was very proud to say that I had made it for her.
Our electronics lecturer at Nottingham University had been in one of the first technical teams to visit Germany at the end of WW2, and had brought back an early Magnetophon tape recorder. This triggered an interest in magnetic recording, and although much of the work was done at the University, I did try at home to improvise a primitive tape deck. I had made a brass 'capstan' (I still have it), which was 1.9" diameter, to give 7.5 inches per sec when run at 78 r.p.m., fastened to a blank lacquer disc. (This was long before the 'Gramdeck' was commercially available.) I had to make my own head, of course, from a single lamination from an old audio transformer bent into a ring. I knew enough to leave a front gap, but not a back gap; the results were awful.
One day back in nineteen-fifty-something our electronics lecturer prefaced his talk (on klystrons) by quoting an article he'd just read in the Bell System Tech Journal. A group of engineers claimed that by adding a second point contact to a germanium diode, they could make it amplify. 'Ridiculous!', said he.
In the Army
Although I spent two years training fellow National Servicemen how to service radio receivers and transmitters, the only constructional project I can remember was to build a receiver for Lesley. This was from a mail-order kit, and used the standard superhet line-up of 6K8, 6K7, 6J6, 6V6 and 6X4. No, it can't have done, because it was a.c./d.c. and the heaters were in a series string, with a line cord as dropper. So they must have been 25L6, etc. It was housed in a small plastic cabinet, and would certainly not have passed BS415.
With Rediffusion in Nottingham
Once married and settled in an a.c. flat, home construction began apace. In those (mono) days, everyone's ambition was to have a 'Williamson' audio amplifier, and so I was able to build one at last. I couldn't afford to buy super deluxe transformers, and the acoustic hum from the cheap mains transformer was such that the main amplifier had to be housed in the eaves (we were in a top-floor flat). But the push-pull PX4s gave good service, even though I was using a horn-loaded speaker which didn't need much power.
It was at that time that 'Wireless World' published a design for a horn speaker enclosure, which was designed for the Lowther PM6 driver. The rear (bass) horn was built from sheet tinplate, soldered together and covered with strips of perforated zinc to act as a key for the skimming of concrete which gave it its strength and absence of resonances. The treble horn was a delicate construction of plaster of paris. This concrete horn lurked in the corner of our lounge for some years, and was a real talking point when visitors first saw it, until the advent of stereo meant that I needed two speakers, and I boggled at making and housing another one. So I built two 'Acousta' wooden folded horns instead, and gave away the concrete horn to a friend. (Legend had it that he used it as a chicken brooder.)
By this time the audio inputs were being provided by a Collaro tape deck, with valve preamplifier, and I forget what pickup: but I'm sure I had moved on from the Acos. The exciting arrival was the start of the v.h.f./f.m. service, and I (with lots of colleagues) began building the 'Wireless World' design by Amos and Johnstone (soon re-christened the Amos and Andy). It used an ECC85 cascode r.f. amplifier, an ECC81 as oscillator and mixer, EF80 and EF85 i.f. amps, and a ratio detector. This was the basis of f.m. receiver development for many years, and mine went through many metamorphoses. The big improvement was to use switched preset tuning, and the next logical step was crystal control. I still have my STC 4434 triple-crystal unit, for Home, Light and Third, for the Holme Moss transmitter. Getting this thing to oscillate in the correct series overtone mode was not easy, and led to much frustration. I recall one colleague winning an award for a new stable oscillator circuit: he was using it in parallel mode, and relying on harmonic mixing!
Transistors came into general use while I was still in Nottingham. Our Chief Engineer went to a conference run by Texas Instruments to encourage the use of transistors, and came away with two 'freebie' state-of-the-art power devices (germanium, of course). But he wouldn't let me borrow them for test purposes.
At that time the only transistors readily available to ordinary mortals were the top-hat 'red-spot' transistors, and ‘yellow-green spots’ if you wanted a frequency response above a few tens of kilocycles per second, as it was then. However, we all made a.m. radios, although we couldn't see far enough ahead to imagine using transistors at v.h.f.
But good-quality power amplifiers were still the province of valves, and Mullard published their famous '5-10' design for home construction. (Five valves, ten watts.) This used an EF86, an ECC83, two EL84s and an EZ80 rectifier. These amplifiers were all the rage, and as I was just getting interested in the production of printed circuits, the 5-10 became my first major enterprise. I learned a lot. I photographed the original artwork myself, and hadn't got the camera and copyboard quite parallel. So mine was the first trapezoidal 5-10.
I needed a large amount of copper-clad laminate, which was then not readily available on the home market. So I placed an order with De la Rue for a square metre, which arrived in an enormous packing case. I'd got the thickness wrong, too, so my eventual pcbs were thick and rigid enough to support the mains and output transformers as well.
There were no easy photosensitive resists; you either bought Kodak KPCR at a ridiculous price, or brewed your own. I tried a double-coating method published in 'Wireless World', which used bichromated albumen as the top layer on a sub-layer of cellulose lacquer. Unfortunately 'Wireless World' got the albumen formula wrong, and my local pharmacist gave up when he tried to dissolve 135 g of ammonium dichromate in 500 ml of egg albumen solution. Next month 'Wireless World' published a correction: it should have read 13.5 g. After UV exposure, this was developed in water, and then the sub-layer was developed in Carbowax.
After a few experiments getting negative and positive the right way round (I still have a test piece with a negative valve-holder layout), I went into production. About ten colleagues at Rediffusion built these amplifiers, very successfully.
There was a thriving constructors' group at The Rediffusion Central Training School, and most lunch-times were spent in building something electronic. Favourites were, of course, f.m. receivers, audio amplifiers, and home-made testgear. (I built a transistor beta-meter in the early days, when transistors were not as tightly specified as now, and there was much sorting through handfuls of bargain rejects.)
My first real transistor radio was a superhet, using OC44/45s, OC71s, and push-pull OC72s. I made my own wooden case, covered with black leatherette, and used it for many years.
Again, most f.m. receivers still used valves. Two designs which I recall gave much pleasure were, first, a revolutionary idea published in 'Wireless World' which was surely the precursor of all modern f.m. detectors. It was extremely compact, using four ECC81 double-triodes. One was a cascode r.f. amplifier, one was a crystal oscillator and harmonic mixer, and three of the remaining four triodes were RC-coupled i.f. amplifiers, working at a few hundred kilohertz. The last triode was an audio cathode-follower. The demodulator was a pulse-counter; the first time we'd encountered anything other than a ratio detector or a Foster-Seeley. This design was enormously challenging; the first requirement was to get three crystals which gave Home, Light and Third. We used ex-Army FT243 demountable crystals, which we bought low, and then etched in ammonium fluoride solution until they were near enough on the right frequency. Or had gone too far, in which case we threw them away and began again with another.
The second exciting project arose from our search for the ultimate. We ended up designing our own (with Gordon Yarwood and Scott Redfern). This final valve design used an ECC85 cascode, ECC81 crystal oscillator/mixer, EF80/EF85 i.f.s, and an American 6BN6 gated-beam discriminator, which gave superb limiting and capture ratio. This was so good that I made a really elegant chassis for it, with soldered corners.
Transistor f.m. receivers at that time were very straightforward, and used OC171s for i.f.s. I recall building one for Lesley's parents in Lincolnshire, and neglecting to include any input protection for the very expensive Texas r.f. amplifier. After a thunderstorm, it ceased to work, and the local radio dealer was given the job of sorting it out. He managed to do so, but it took him a long time.
Other Audio Amplifiers
After the mono Williamson and the stereo Mullard 5-10s, the next move was to make a Chinese copy of a Leak TL12. This used all germanium p-n-p transistors, including AD140s, and was my first real hi-fi transistor unit. Included in the case were a switch-tuned f.m. tuner and a separate stereo decoder, and the whole thing gave good service for many years.
Meanwhile, other amplifiers were being built at The Manor. Several of us built an early 'Wireless World' design using transformer-driven OC22s, and the driver transformers had to be home-made. Most of us just wound them by hand, but Bob Hull decided to make a coil-winding machine to simplify the task. Needless to say, making this took him many times longer than winding the things by hand, and I can still remember the clackety noise made by this machine when in full production.
Someone even made a Sinclair Class D amplifier.
Nostalgia Time by Jake Loddington Page 1
Move on to Page 2
© Rediffusion.info 2010
Rediffusion Central Services Limited
Training and Publications Division,
The Manor, Moorland Road, Poulton-le-Fylde, Blackpool. FY6 7HA
Tel: 0253 882278