Fullscreen HTML5Med Res (??MB)HTML4

Vintage Radio 1938 VE301Wn Dyn Volksemfänger: the People’s Receiver By Ian Batty Was it only ever a propaganda radio? You will have to decide for yourself. I addressed the historical and political context of radio that came after this one, the DKE38 Kleinempfänger, in the July 2017 issue (siliconchip. au/Article/10728). My reading casts doubt on the common belief that the VE301’s design was purely the result of political pressure. Otto Griessing designed the VE301 at the company Dr Georg Seibt AG. This followed a request by propaganda minister Joseph Goebbels to design a reasonably-priced but high-­ quality broadcast receiver. The cabinet was designed by Cologne’s (Köln’s) Professor of Artistic and Technical Design, Walter Maria Kersting and his students. Costs had to be kept down, but even so, the VE301 cost roughly two weeks’ 100 Silicon Chip salary. Edwin Armstrong’s superhet patent, owned by RCA, was only released for use by other manufacturers in 1930. But superhets required up to eight valves, so they were more expensive to build than simpler regenerative sets. Also, the very complexity that gave the superhet its superior performance was not widely understood and would not be easily supported by existing local repair shops. A previous Armstrong patent, the regenerative receiver, had been widely used for almost a decade and was well understood. It was the design of choice for many experimenters, young and old. With a single radio valve costing several days’ wages, a minimal threevalve set was the obvious choice. Australia's electronics magazine The VE301 was released at the Internationale Funkausstellung (International Radio Exhibition, Berlin) in August 1933. At only 76 Reichsmarks, it was half the price of any competitor. Over 100,000 sold in the first two days. VE301 initial release The VE301 was clearly a result of that 1933 request by Goebbels, but the official ban on foreign broadcasts was not issued until September 1939. While it’s true that the Nazi government progressively forced more and more draconian restrictions on the German people, casting the VE301’s limited reception range as purely the result of its being a propaganda radio is historically inaccurate. That’s reinforced by the absence of siliconchip.com.au A close-up of the slide-rule dial. Note that German and Austrian cities are both listed. the Reichsadler (imperial eagle) on initial VE301 releases, by print articles of the day advising on the construction of antennas, and by a thriving accessories industry. There were stick-on dial charts listing stations all over Europe: London, Oslo, Paris, Prague, Warsaw, Toulouse, Budapest, Stockholm, and Rome among them. And there were add-on dial mechanisms listing international stations. Radiomuseum is a good place to find examples of these (website: www.radiomuseum.org). The Antique Radios website also has an extended discussion – see the references below. The set is built on a steel chassis and the need for mass production did not force compromises on the mechanical design or the quality of electrical components. There are even shallow stampings in the chassis to show valve positions. Different versions The VE301 was issued in various models: AC-only, AC/DC, DC-only and battery. Many battery versions came in timber cases, while the mains versions were in tall Bakelite cabinets. The initial issue featured no overt Nazi symbolism, though it did have a “speaking eagle” below the uncalibrated semicircular dial. The VE301 “German People’s Radio” was to be ‘a radio in every house’. It needed to be cheap enough for people to buy, simple to operate and use technology that technicians and tinkerers could maintain. The initial VE301Wn used a triode in the RF amplifier/demodulator stage, a moving-iron loudspeaker with no speaker transformer and a 3kΩ filter resistor. Altogether, the design made the cheapest possible choices. A minimalist design It was minimal, but was it cheap and nasty? The initial release used a triode RF amplifier/demodulator and a high-impedance moving-iron loudspeaker – both the cheapest possible choices. Component quality was at least equal to comparable radios. My set had some capacitors replaced by a previous repairer, but I only found one resistor sufficiently out of tolerance to need replacement. The VE301 Dyn, released in 1938, upgraded the design to an AF7 pentode RF amp/demodulator and an electrodynamic speaker with a speaker transformer. The VE301Wn Dyn design, which is what I have, replaced the initial 0-100 semicircular dial with a lettered slide-rule dial and dial cord mechanism. My dial lists cities in Germany and Austria (as you’d expect after the annexation) and, more significantly, cities in what are now Poland and Russia. Editor’s note: At the end of WW2, the Soviet Union annexed East Prussia while much of Pomerania and Silesia became part of Poland. Two Reichsadler symbols flank the dial, and all original components (including the inside of the cabinet) bear that symbol. siliconchip.com.au Opening the rear of the VE301Wn Dyn radio reveals the chassis and electrodynamic speaker (rather than a metal reed type used in the versions from 19337). This later model of Volksemfänger also added an audio output transformer. Australia's electronics magazine February 2023  101 Over twelve years and a variety of models, nine million VE301s were made. 42 manufacturers were involved in pushing out the radios and accessories for the German population. Radiomuseum lists 290 VE301 variants and accessories, so this article cannot cover all possible variations. You can draw basic distinctions from the full type number. VE301B (batterie) is a three-valve battery version, -G (gleichstrom) is the DC version while -W (wechselstrom) is AC only. GW versions were AC/DC, with a barretter (similar to a ballast resistor) in the series heater circuit. Wn (Wechselstrom neu) initially denoted the AF7 RF amplifier/demodulator and revised antenna circuit, but later “W” versions dropped the “n”. Dyn versions use an electrodynamic speaker and the AF7 RF/demodulator. There are inconsistencies, and Radiomuseum is the best single authority. Circuit description I’ve redrawn the whole circuit in Fig.1. My VE301Wn Dyn begins with the dual-wave antenna circuit of L1 to L4. L1 is tapped to allow matching with short or long antennas, with C1 extending the matching capability. L1 is mounted on a swing arm. This allows the user to vary the antenna coupling, substituting for the usual potentiometer-style volume control. Tuned winding L2, in series with feedback winding L4, is used on the 150-350kHz long wave position. L3 shunts L2 for medium wave, reducing the tuned-circuit inductance to cover the range 500-1500kHz. The grid leak resistor-capacitor R1/ C4 combination is in a single casing and sits under the AF7’s shielded grid cap. Its high resistance allows the AF7 grid to drift to a bias of about -0.7V. The screen grid is supplied via R3, bypassed for audio and RF by C7. The anode circuit supplies RF feedback to the antenna circuit via variable capacitor C3. The anode also provides audio, via C6, to output valve V2. The RF amplifier/demodulator stage is decoupled from the main HT supply by resistor R4 and capacitor C5. The output stage, based around V2, uses back-bias developed across R9 and supplied via decoupling components R6/C8 and grid resistor R5. As the RES164 is directly heated by the 4V AC filament supply, R6 is used to balance the average filament voltage to ground, thus reducing mains hum. V2’s screen, unusually, is fed via dropping resistor R7, bypassed by C10. This agrees with the RES164 screen voltage specification of 75V. V2’s anode connects to output transformer T1, then to electrodynamic speaker LS1. T1’s primary is bypassed using anti-resonance capacitor C9. Mains transformer T2 has 220V and 4V AC secondaries. Rectifier V3, an RGN1064, has its filament supplied from an extension of the HV secondary. It’s an unusual configuration but avoids the need for high-voltage insulation between the HV and filament windings. This does commit the design to halfwave rectification and the resulting need for better supply filtering, but the low HT drain of only 24mA eases the task. It is unusual to see a valve rectifier’s anode connected to supply ground (via back-bias resistor R9). Still, the circuit works perfectly well, with the rectified DC supplied from the other end of T2’s HV winding. Supply filtering is by the combination of filter capacitors C11/C12 and the field winding of electrodynamic Fig.1: the redrawn circuit diagram for the VE301Wn Dyn radio. As there were many different versions of this radio produced, many circuits found online will have small changes compared to this one. 102 Silicon Chip Australia's electronics magazine siliconchip.com.au Most of the rubber-covered wiring on the set was in good condition, but some sections had lost insulation and were promptly replaced. The AF7 wears a shielded “top hat” over its grid cap connection. The grid leak resistor and capacitor are housed underneath the shield. speaker LS1. Finally, R8 adds to the HT current drain of the two valves, ensuring enough magnetising current for the loudspeaker’s field coil. Regeneration Edwin Armstrong showed that controlled regeneration could greatly improve receiver performance. As detailed below, full regeneration in the VE301 Wn increases sensitivity by some 40 times. It’s easy to understand that regeneration increases gain, and feedback calculations can either derive gain from feedback if the feedback factor is known, or derive the feedback factor if the gain is known. But it might not be so obvious why bandwidth varies so greatly. We’re familiar with negative feedback’s effect on bandwidth – it generally increases it. Thus, it makes sense that positive feedback should reduce bandwidth. In the regenerative tuned circuit, feedback does this by reducing the effective circuit resistance, increasing tuned-circuit Q. Q can be calculated either as 2πf × L ÷ R or as (1 ÷ R) × √L ÷ C. The second formula is preferred as it indicates that a tuned circuit with a high L/C ratio will have a higher Q. Q can also be measured as fc ÷ ∆f, where fc is the resonant frequency, and ∆f is the bandwidth between half-power points. Measured bandwidths of ±400Hz (maximum regeneration) and ±4.15kHz (zero regeneration) at 1400kHz gave calculated Q factors of 750 and 66, respectively. Circuit resistances came out to 1.5W and 17W, respectively. The ratio of the measured Q values (approx 11.3:1) conforms to the calculated resistance ratio of 1:11.3. We can also determine the relationship between gain and bandwidth by assuming that the gain-bandwidth product is constant; increased gain gives reduced bandwidth. In summary, along with the general stability problem, the regenerative circuit suffers from reduced bandwidth with increasing feedback. The VE301 also suffers from dial calibration errors. Tuning to 600kHz with maximum regeneration, the set drifts about 40kHz high as regeneration is adjusted to zero. The close antenna-grid circuit coupling also affects dial calibration. Restoration I auctioned this set off at an HRSA auction to a fellow HRSA member. He was kind enough to let me borrow it to check out this classic radio. The first problem was the mains cord. It was not anchored, and the Active wire had broken clear of the mains switch connection. Fortunately, I observed my own advice to never This Volksemfänger was designed with multiple models to suit batteries and AC or DC mains. For example, this set has a wire on the power transformer to select between 110V, 150V and 220V mains operation. Some other manufacturers would provide 125V or 130V instead of 150V for the primary tap. Be sure to check the circuit diagram for your set. siliconchip.com.au Australia's electronics magazine February 2023  103 turn anything on until I’d checked the power supply. A cordgrip clamp fixed the problem – these grip the cord securely, insulate it from the chassis penetration and prevent twisting. The set had been worked on, with both electrolytics and many of the fixed capacitors replaced. Much of the rubber-covered wiring was still good after some eighty years, but I replaced the sections that had lost insulation. The original regeneration capacitor was missing and a potentiometer had replaced it in the feedback circuit. On testing, this arrangement worked well enough. I did alter the pot’s connections to give more predictable control. The resulting changes to the circuit are shown in Fig.2. This is an example of a set where you either demand complete authenticity and try to salvage components from a wreck, or accept some compromise and create a working radio. Valve V1 and rectifier V3 both tested weak, while V2 had been substituted with a Russian valve for which I’ve been unable to find data. As this had a “loctal” base, an adaptor to the European 5-pin base had been fitted. V3 had been bypassed with a 1N4004 silicon diode. The HT drain is low, so I accepted that the original RGN1064 would work well enough and removed the 1N4004. On test, with just the RGN1064, the main HT measured 234V – close enough! The AF7 and the substituted RES164 were more of a problem. I was able to buy a pair of NOS AF7s online and Using clip leads for testing has the advantage that you don’t have to reach into the chassis to make measurements and it’s harder to slip and short something! they tested perfectly. The RES164 is one of the “Miniwatt-class” of output pentodes such as the B443, with 4V/150mA filaments. I couldn’t get a suitable replacement in time, so the substitute stayed in place. The tuning capacitor’s outer plates have the ‘petalled’ form that allows exact tracking adjustment. These had been mangled, preventing the capacitor from fully rotating. Some inner plates were also distorted and shorting, but I easily straightened them up and restored correct operation. The hum-balancing pot in the output stage filament circuit was intermittent. It had also been twisted through 360° at some stage, badly bunching up Fig.2: the radio’s regeneration circuit was modified with a potentiometer replacing the original (failed) regeneration capacitor. This modification would have been easier than fixing or finding a substitute. 104 Silicon Chip Australia's electronics magazine the heater circuit wiring and other connections. The pot has a fragile resistance deposition that cannot withstand extensive rubbing from the moving contact. If you find one of these pots in good condition, resist the temptation to adjust it unless really necessary. A similar pot is found in the DKE38, but that one is used to set the bias, not for hum reduction as some online sites mistakenly assert. I substituted the VE301’s faulty pot with a miniature version. How good is it? It’s a three-valver with just two signal valves. The AF7 pentode operates as a conventional regenerative gridleak demodulator, while the RES164 is the output valve. On test, I was able to get the standard 50mW output for an input of 400µV at 600kHz and 1400kHz, 1200µV at 155kHz and 200µV at 350kHz. These were achieved just below the point of regeneration and are the maximum possible figures. For the reasons described earlier, these sensitivities gave extreme -3dB bandwidth restrictions of ±250Hz at 600kHz (really!), and ±400Hz at 1400kHz. A practical regeneration setting, needing about 2mV input for 50mW out, gave bandwidths of about ±1kHz at 600kHz and ±1.8kHz at 1400kHz. Signal-to-noise ratios exceeded 20dB for all measurements. With full regeneration, -40dB skirt selectivity was ±25kHz at 600kHz and ±86kHz at 1400kHz. siliconchip.com.au Experience with the DKE38 led me to expect a significant change in the AF7’s anode/screen voltages. They did rise a bit, but less than expected. The result is that the VE301 gives a pretty constant output for most signals, as though it had an AGC function. Regenerative gain Going back to my maximum sensitivity of 400µV for 50mW out and then cutting regeneration completely, I needed around 15mV to get 50mW output again. That implies that regeneration supplies an extra stage gain approaching 40. The DKE38 Kleinempfänger’s best sensitivity of around 600µV suggests that regenerative demodulators give comparable RF performance just before the point of oscillation. Any major improvement would need extra RF gain before the demodulator, or extra audio gain after it. Buying another I picked up another VE301Wn (AC operation, moving-iron speaker, triode RF amplifier/demodulator) online for about half what most were asking. It was described as “working”. It will be interesting to see what some folks think “working” really means. Special handling The VE301Wn Dyn radio is a fairly simple set, with sparse few parts located on the top and underside of the chassis. Many of the capacitors had been previously replaced; only one resistor ended up needing to be changed. On air, with about 10 metres of antenna thrown over the carport, 774 ABC Melbourne rocked in with minimum regeneration and the antenna coupling backed off. With some adjustments, I could easily pick up all metropolitan stations. Given the optimum sensitivity of 400µV, would I be able to pull in 3WV at 594kHz? Yes and no. Tuning to that frequency, I still had a strong signal from Radio National at 621kHz. Adding a signal generator on its CW setting, I could hear a heterodyne at 594kHz, but could not make out the broadcast. The VE301’s skirt sensitivity was siliconchip.com.au just not good enough to sufficiently reject the 621kHz signal in favour of 3WV at 594kHz. Among the many accessories marketed for the VE301 were several passive antenna tuners/preselectors and a powered RF preamplifier. Either of these would have improved the separation of closely-spaced stations. What about strong signals? Starting with 400µV, giving 50mW output, I cranked up the signal generator. The output reached the VE301’s maximum of 0.5W at 2mV of signal and did not rise much as I got to 100mV on the signal generator. Australia's electronics magazine P-base valves such as the AF7 seat into the socket by sliding down past leaf springs. When seated, the valve’s Bakelite base does not extend very far upwards past the chassis/socket rim, so it’s tempting to remove a valve by grasping the envelope. Don’t do this. Take the time to grasp the top rim of the valve base. You may need to rock the valve side-to-side to get it moving, and be careful – it might suddenly release with the risk of smashing the envelope against other parts of the radio, the case, or other things on your bench. I should register a new acronym, TNMTAM (they’re not making them anymore) and just use that from now on. Acknowledgements Thank you to Herbert Detlefsen of the HRSA for the loan of this historic radio. For more on this series of radios, see the extended discussion: siliconchip. au/link/abgf For the relevant Radiomuseum page, visit siliconchip.au/link/abgg SC February 2023  105