Fullscreen HTML5Med Res (??MB)HTML4

Vintage Radio 1940 1940 RCA RCA BP-10 BP-10 Personal Personal Radio Radio By Ian Batty The incredible shrinking portable radio: RCA’s next-generation all-B7G set, the BP-10. I bought this radio some time ago for reasons that I can’t quite recall. But after picking it out and doing a bit of research, I was glad that I had. The RCA BP-10 is pretty much the first outing for the all-glass, B7G “miniature” valve lineup that, with its B9A cousins, was to dominate valve production until transistors took over. While Compactron tubes by GE and subminiature designs further refined valve technology, the only true innovations that came later were metal-ceramic Nuvistors by RCA, and all-ceramic VHF/UHF types. RCA, established in 1919, had become a major market force by 1935. Their successful development and release of metal valves that year confirmed RCA as a serious research and development player. Successful, reliable and robust as they were, metal valves were similar in size to their conventional precursors. The fact that pin 1 was reserved for earthing the metal case prohibited the development of twin triodes and other multi-unit types. It’s strange to think that an “octal” valve should actually be a 7-pin device with a factorysupplied shield. Metal valves had mounted the element assembly onto pins in a glass base disc with some support from a metal base rim, pointing to the possibility of all-glass construction. All-glass construction was pioneered in the specialised Acorn series, designed for the VHF range. Their small size (just 18mm in diameter) and use of peripheral connections allowed the 954 pentode’s application “for wavelengths as short as 0.7 meters” – that’s around 430MHz. 70 Silicon Chip Ongoing development yielded triodes capable of oscillating past 1GHz. But the connection ring’s size, plus the limited number of possible connections, restricted Acorns to applications where no other design could be used. B7G valves Consumer-applicable construction materialised in the B7G series, first released in mid-November 1939. A full description of the lineup appears in RCA’s Radio Review of April 1940. B7G construction economised and improved valve construction, reliability and performance, equalling and bettering their mainstream octal predecessors. First, the element structure was designed to fit inside a T5½ (11/16th of an inch; around 17.5mm, with a maximum diameter of ~19mm) light bulb, which is just a little wider than an Acorn valve (eg, 955 acorn triode, 14mm envelope diameter). Designers were able, for example, to reproduce the gain of the octal 1N5 pentode in the B7G 1T4, and improve slightly on the 1A7’s conversion gain with the 1R5. Using an all-welded construction, where the valve assembly was welded directly on to the base pins, unreliable solder joints were eliminated, as was the octal valve’s infamous loosening of the envelope’s attachment to the base. Curiously, there seemed to be some confusion over the exhaust tip. Although I’ve never seen an example, provision was made for a base exhaust Australia’s electronics magazine that would have protruded down between the pins. Some advice exists that the central shield on the B7G base should never be filled with solder, as this would have prevented the insertion of base-exhausted B7Gs. The electrical path from any B7G electrode, via its base pin to the equipment’s circuit, is very short. This meant that B7G and their larger B9A cousins would operate at up to 860MHz in UHF TV tuners. Indeed, three early B7G releases were simply re-packaged Acorns. So we have improved reliability, compact size, and improved highfrequency operation. Need a batterypowered transmitter delivering over a watt at 100MHz? Look up the data for the 3B4. The BP-10 It’s curious that one of RCA’s competitors, Sonora, just beat RCA to market given that the four B7G valves were all invented by RCA. It appears that RCA had supplied samples to other manufacturers, realising that industrywide uptake would be a real commercial advantage. The BP-10’s first date of issue was early March 1940. For one of the most thorough descriptions of any set I’ve come across, see TinkerTom’s excelsiliconchip.com.au The RCA BP-10 is shown above slightly smaller than actual size (230mm wide, 1.9kg), and was one of the first commercial radios to use B7G-type valves. The set’s power is controlled via the opening of its flip-top lid. This version is one of the later models which have an arm (upper left) to limit the angle of the lid. lent writing on Blogspot (http://bp-10. blogspot.com). The BP-10 was a runaway success, with some 210,000 produced between 1940 and 1942. Production ceased with the United States’ entry into WWII. It has been variously described as a “music box” (open the lid, and it plays), the first truly Art Deco radio, and a “camera construction” radio. The latter tag would capitalise on the ubiquity, usefulness and total portability of film cameras of the day. Part of RCA’s delay in the BP-10’s release was caused by the creation of a substantial marketing campaign. The BP-10 was seen in movies, photographed with movie stars of the day, mentioned by famous columnist and broadcaster Walter Winchell and advertised in pride of place by major department stores. And you could “personalise” your set. A spares list contains a set of engraved metal letters that buyers could attach to their prized possessions: mine belonged to “OM” – one wonders whether the family might one day read this article. The review set’s tuning capacitor code of 91742 hints at a construction date of September 17th, 1942. It uses valves with date code RE6 (NovemberDecember 1940) and the decorative RCA Victor brand. It’s a conventional valve set, using sockets mounted onto a pressed-andpunched steel chassis. Most wiring is point-to-point. B7G valves, at under 25% of the volume of even the most compact octals, would challenge designers to apply miniaturisation techniques elsewhere. The largest single components, the A and B batteries, were targeted. The 1.5V filament supply could come from a single 950 (“D” size) cell. B7G valves work just fine with high tensions of 60V+, so the logical choice was 67.5V – one-half of the old 135V HT battery. RCA’s original instructions quote some 3~5 hours of life for the LT cell against some 25~40 hours for the HT battery. Purchasers were advised of the discrepancy, and warned to always try replacing the LT cell before replacing the HT battery. Battery life is certainly a compromise compared to STC’s octalequipped 418, which had a battery life ten times longer (or more). The most unusual result of shrinking this set is the loudspeaker: it is oval-shaped with a permanent magnet that seems to be cut in half! The vacated space allows relaxed mounting for the two audio valves, although RCA service notes describe possible problems with the speaker’s magnetic field upsetting the output valve’s internal electron flow. Hmmm... Since you’d only glance at the internals when changing batteries, most might not notice the quality of construction. It’s good, and even though the RF/IF section is built within a metal trough, most components can be accessed for testing or replacement. Left: To showcase the small size of the B7G-type valves, here are how pentagrid converters changed over time. From left to right: 2A7, 6SA7, 1R5 (B7G) and 1E8. Right: the ‘strange’ 3-inch, 3W loudspeaker, which looks to use a permanent magnet that has been cut in half. siliconchip.com.au Australia’s electronics magazine November 2020  71 V4 Output V3 1st Audio 2nd IF V2 1st IF 1st IF V1 Converter Antenna Gang Oscillator Gang 1.5V “A” Battery 67.5 “B” Battery Oscillator Coil The rear view of the BP-10 chassis showcasing the miniature B7G-type valves. A bit of a ‘spy radio’ Louis Muelstee, in his superb four-volume series “Wireless For the Warrior”, features the BP-10 in his supplement to Volume 4. Muelstee states: “BP-10 receivers were issued to the French Resistance pending the mass production of MCR-1 ‘biscuit tin’ receiver. In 1943, 150 units were delivered in France during a clandestine landing (operation ‘Orion’), to Commander Paul Riviere alias ‘Marquis’. A BP-10 receiver which belonged to him can be seen in the ‘Museum of the Order of the Liberation’ in Paris.” Circuit description Four-valve portable designs were well refined by the late 1930s, and the BP-10 yields few surprises. The signal from the loop antenna connects directly to the 1R5 converter’s signal grid, and the loop is tuned by one half of the gang. AGC is applied in series with the loop’s winding. The converter’s local oscillator (LO) uses the screen grids (internally-connected grids 2 and 4) as the oscillator anode. This differs from other designs that ‘collect’ the two screens and the main anode to function as the anode in the oscillator circuit. Padder V1 1st IFT V2 The tuning gang, unusually, uses non-symmetrical sections and a padder. The antenna section’s range is 10~325pF while the oscillator section is only 10~225pF. Such asymmetry would usually eliminate the need for a padder, but the BP-10’s oscillator section obviously had too high a maximum capacitance, as 270pF capacitor C4 was added in series with the oscillator gang. The 1R5 screens connect, via oscillator coil L3’s primary, to the screen of the 1T4 IF amplifier for supply. Since the IF amplifier is part of the AGC circuit, I’d expect the 1T4’s screen current to fall on strong signals, allowing its screen voltage to rise. This would also allow the 1R5’s screens to rise, thus increasing the supply voltage to the LO section – usually a recipe for frequency instability. That aside, the LO circuit is what you’d expect: an untuned primary with a tuned secondary and a high-value oscillator grid resistor. The RCA circuit lists oscillator grid voltages at both ends of the tuning range, but don’t be surprised if you’re unable to get the correct measurements. It’s common for such low-power circuits to stop working when a test probe is placed on the grid due to meter loading. The 1R5 converter’s anode feeds the signal to the first IF transformer, which has a tuned, untapped primary and 2nd IFT V3 Volume Control Antenna Gang Oscillator Gang The underside view of the BP-10 chassis. 72 Silicon Chip V4 Australia’s electronics magazine siliconchip.com.au secondary. Its output signal feeds the 1T4 IF amplifier. As noted above, this shares its screen supply with the LO circuit, provided via 15kW resistor R2 and bypassed by 20nF capacitor C10. The IF amplifier also receives AGC, supplied in series with the first IF primary. The AGC line is bypassed to ground for RF and IF by 50nF capacitor C7. Amplified IF is applied, via the second IF transformer, to the diode within the demodulator/audio preamp 1S5 valve. It’s a sharp-cutoff pentode with a diode designed for this application, offering an audio gain up to 66 times. Demodulated audio, filtered by 100pF capacitor C13, passes via 47kW resistor R5 to the 1MW volume control potentiometer, R6. Audio from R6’s wiper goes via 1nF coupling capacitor C14 to the control grid of the 1S5. This stage gets contact potential bias via high-value 10MW grid resistor R4. This allows the grid to drift weakly negative due to the space charge “cloud” of electrons surrounding the valve’s heated filament (see June 2020, p39 for details). The signal’s DC component is fed back, as AGC, via 3.3MW resistor R3 to the IF and converter stages. The 1S5 uses high-value screen and anode resistors: 4.7MW (R8) for the screen and 1MW (R7) for the anode, with the screen bypassed for audio by 50nF capacitor C15. R7 and R8 only permit low electrode currents (reducing the valve’s mutual conductance), but the potential loss of gain is made up by their high resistance values. Expect a gain of some 35+ times. The output from the 1S5 goes, via 1nF capacitor C19, to the signal grid of the 1S4 output pentode. This original valve, with its 100mA filament drain, could not economically be put in series with the other three valves to allow 6V operation, as their filaments only demanded 50mA. The 1S4 was quickly superseded by the near-identical 3S4 that possessed a tapped filament. This could be powered from 1.4V, drawing 100mA, or 2.8V, drawing the more common value of 50mA. The 1S4/3S4 amplifier requires a -7V bias for Class-A operation. This is supplied by 820W resistor R9, in series with the battery’s negative terminal to ground, so passing the set’s entire HT current. It’s a simple solution, but it does “steal” some 7V from the HT. The output valve’s 3.3MW grid resistor, R10, connects from the grid to the negative terminal of the HT battery, supplying the required -7V bias. Cleaning it up The front cover was in excellent condition inside and out, and the set was mercifully free from battery corrosion. The case, though, had lost much of its leatherette covering and the rear cover was corroded. After a clean-up and application of new vinyl, it looks a treat. Only three valves, all original RCA-branded, remained. Regrettably, two (the 1T4 IF amplifier and the 1S5 demodulator/audio preamplifier) tested low on gain, and so needed replacement. Electrically, the set offered several challenges. The LT battery current draw ranged anywhere from about 100mA to 200mA. This turned out to be due to corroded or dirty valve sockets; an application of spray cleaner fixed this. The HT current measured above 15mA. Leaky audio coupling capacitor C19 was putting a positive voltage on the 1S4 grid. Having replaced it, I expected the set to come good. siliconchip.com.au Australia’s electronics magazine November 2020  73 But no; the 1S5 screen voltage was low, and removing the valve only let it rise a bit, so screen bypass capacitor C15 was also leaky. Having replaced that too, I tested the set’s audio stage. I found the gain was low from the top of the volume control pot, but normal at the 1S5 grid. C14 was pretty much open-circuit. Replacing it resulted in screeching oscillation! So I decided (in a move I possibly should have made earlier) to replace all the 70+ year-old paper capacitors, along with 10µF electrolytic capacitor C17. I could now get some reception, but the tuning capacitor was hopelessly scratchy. The plates looked to be aligned OK. Luckily, a spray of contact cleaner on the ball bearings and the pressure/grounding spring at the other end restored it to correct operation. How good is it? The first of anything can be a bit ho-hum. Maybe it’s the problem of any first, but I found the BP-10 to be workable without being remarkable. For 50mW output, it needs around 1.5mV/m at 600kHz and 1mV/m at 1400kHz for signal-to-noise ratios exceeding 20dB. Its RF bandwidth is around ±3kHz at -3dB; at -60dB, it’s ±26kHz. The frequency response from the antenna to the speaker is 120~2700Hz. Trying to get maximum possible output resulted in a virtual square wave at only 70mW. At the more usual 50mW, total harmonic distortion (THD) was around 14%, and 5% at 10mW. The output is low compared to manufacturer’s figures, which have the 1S4 giving 180mW with around 60V HT. However, everything tested OK, and the set is loud enough for its intended use. It does benefit from correct loop orientation; the ability to reorientate it is useful for picking out distant stations while nulling strong city transmitters. And that converter screen changing with the AGC voltage, potentially compromising frequency stability? It shifts by less than 500Hz from no signal to a strong signal. Hats off to the designers on that point. Notes The original RCA circuit shows 67V at the 1S4 screen. Given the loss of some 7V across back-bias resistor R9, and the fact that the converter and IF amp anodes both show 60V means that 60V is the correct figure for the 1S4 screen when measured to chassis. The 67V readings would be taken to battery negative, but the notes do not make this clear. The RCA circuit usefully shows voltage gains for each stage. Be aware that the RF/IF gains are for modulated RF/IF signals and the audio gains are for audio. The “loss” shown for the second IF is at the intermediate frequency, and does not account for the additional loss in demodulating a 30%-modulated test signal. Consult the measurements in my circuit diagram for more details. Is it worth buying? Like all “firsts”, it’s well worth having. With some 200,000+ made, you’ll still find good examples online, some with original leather cases. Thank you to Graham Parslow of the HRSA for the loan of his STC 418 for the size comparison. Further reading ● For a thoroughly engaging and comprehensive description, visit http://bp-10.blogspot.com ● For Louis Muelstee’s description, see: siliconchip.com.au/link/ab3j and think about his entire “Wireless For The Warrior” series, my top reference for British and other military radios. ● RCA’s description of B7G technology: siliconchip.com.au/link/ab3k ● Techies, see: siliconchip.com.au/ SC link/ab3l The STC Melody 418 (left), at nearly 75cm tall, shown next to the RCA BP-10. 74 Silicon Chip Australia’s electronics magazine siliconchip.com.au