Fullscreen HTML5Med Res (??MB)HTML4

Vintage Radio The Monarch “All-American Five” Wedge Radio This “All American Five” design appeared in the late 1930s as demand for cheap domestic radios took off. Accepting five valves as necessary for a well-performing superhet radio, the “AA5” design aimed to simplify the circuit as much as possible. By Ian Batty T he most obvious first step was to eliminate the power transformer. That would make the radio lighter and smaller. Being made for the common US 110~117V AC supply, designers chose to run the valve heaters in series across the mains supply. Astor’s Mickey OZ1 (up to Serial No 460) adopted such a design from one intended for the US mains supply. As the 12V series had not been released at that date, the OZ used valves with 300mA heaters (6A7, 6D6, 6B6, 43, 25Z5). The 43 and 25Z5 worked with 25V heaters (to give a 50V drop in series), 96 Silicon Chip but the remaining three only added some 19V. At around 69V in total, operating from a 110V supply would demand a series resistor to drop around 40V – wasteful, but probably not unreasonable. Australian releases ran on 240V and needed a series resistor to soak up a massive 170V. So Astor just popped in a 580W dropping resistor, with a power dissipation of over 50W! Some US manufacturers, needing to add voltage drops to meet their 110V mains, took the ingenious idea of making the mains cord resistive. Australia's electronics magazine While it did work, it meant that to replace the cord, either a cord with identical resistance was required, or the fitting of an actual resistor inside the chassis. It also earned these cords the nickname ‘curtain burner’ – hardly ideal! One solution to the problem was to use valves with 150mA heaters and double the heater voltage to compensate. Many of the 6xxn series (6SA7, 6SK7 etc) were released in 12V versions by 1939. This change simply required a different heater wire resistance: the rest of the valve was identical. siliconchip.com.au B7G miniatures (6BE6, 6BA6, 6AV6 etc) were also re-engineered. While 12V signal valves could give sufficient emission (as the heater power was still around 1.9W), this would not give sufficient emission for 12V/150mA output valves or rectifiers. But since the heater string needed to add up to the mains voltage, why not design output valves and rectifier heaters for higher voltages? This idea resulted in the 35V 35W4 with 5.25W of heater power and the 50C5 with 7.5W (in comparison, the 6X4 had 3.8W and the 6AQ5 2.8W). The extra emission would help give better performance at the low anode voltages found in these radios. Philco’s PT44 used Loctal 7-series valves with 6.3V, 150mA heaters for the converter, IF amplifier and audio preamplification, but glass octal valves for the audio output (50L6GT) and rectifier (35Z3). The total heater voltage only added up to 103V, so the dial lamp, shunted by a resistor, made up the rest. Putting a dial lamp in series with the heaters sounds like economy, but a blown dial light would stop the radio dead. Valves such as the 35W4 rectifier were designed with a heater tap that created a suitable supply for the dial lamp and used different resistance values for the two ‘halves’ of the total heater circuit. While this worked, a blown dial lamp would allow excessive voltage across its heater section, leading to heater burnouts. If you are working on any All American Five, check whether it has a dial light in the rectifier’s heater circuit and, if so, that the lamp has the correct rating and is working. Having no mains transformer meant half-wave rectification, with a resulting low HT supply; only 95V in the PT44. Signal valves would work satisfactorily at such low supply, with the 6BE6/12BE6 specified for 100V operation with little reduction in performance. But the lowest HT specified for the venerable 6V6 and its miniature equivalents was 180V. Valve designers, needing to produce high heater voltage types such as the 50L6, took the opportunity to redesign siliconchip.com.au The Masonite rear panel has plenty of ventilation and a stuck-on circuit diagram; not something you see these days, sadly. the electrode structure, allowing the 50L6 to be fully specified with an HT requirement of only 110V. While this only offered some 2W of output, they were used in economy mantel sets, where this lower output power would be acceptable. Valves such as the 25Z5/Z6 were designed with two completely isolated rectifier diodes and were used as voltage doublers in some sets. This would easily give the more usual HT of 250V, but the extra complexity was against the design concept of the AA5, and was rarely used. The direct-from-mains transformerless design meant that such sets would run from either AC or DC supplies, and were often branded as AC/ DC sets. They would, confusingly, sometimes not work on a DC supply until the mains plug was removed and flipped over. That is, until the positive side of the mains connected to the rectifier’s anode! DC operation often gave worse performance, as the filter circuit was not being charged to the peak value of the AC mains, around 150V, giving about 125V at the filter output for approximately 50mA of HT current. The rectifier’s forward voltage drop was only about 5V at the expected current of 50mA. Starting with, say, a 110V supply, the set would only be getting some 105V of HT on DC mains. Australia's electronics magazine The Monarch AA5 The Monacor 5-1H shown on Radiomuseum bears serial number 319387. Mine has no number, but if the serial numbers for this basic design started at one, there must have been around 400,000 made! It’s a minimalist set. The combination of the low US mains voltage and 150mA heater currents allowed the mains transformer to be eliminated. However, making it work with an HT as low as 100V would have been a challenge. Either the designers would need to put in effort to deliver acceptable performance, or buyers would need to accept this was a ‘kitchen radio’ and not expect outstanding performance. It is compact – I have any number of transistor radios that considerably exceed its volume of a bit over 1600cc (1.6L), and its weight of just under 1kg. The chassis weighs just 420g! Its transformerless design makes it economical in use, consuming only 23W when running. The chassis underside photo is not distorted; the chassis front is angled to match the central depression in the cabinet face. Circuit description The circuit is shown in Fig.1; it’s a conventional five-valve superhet using a pentagrid converter and simple automatic gain control (AGC). January 2025  97 Fig.1: the monarch “Wedge” circuit with suggested test points and expected voltages. 98 Silicon Chip Australia's electronics magazine 12BE6 converter valve V1’s local oscillator section uses the common Hartley circuit, with cathode-to-gridone feedback and R1/C5 providing bias for the local oscillator (LO) circuit. The LO tuning section (C7) uses cut plates, giving a different capacitanceto-­rotation characteristic from that of the antenna section (C4) and removing the need for a padder capacitor. The moving plates are identical for both sections, so it’s the stationary LO plates that have the cut profile. The low HT voltage allows the converter screen to connect directly to HT, rather than via the dropping resistor used in most radios. The converter runs without cathode bias, but the high value of series resistor R3 allows some contact potential effect. Added to around -0.4V from the AGC circuit, this sends the converter’s signal grid to about -1.1V. With no external antenna/ground connection, this set relies on the effectiveness of its ferrite rod for signal pickup. This proved to be quite short compared to other sets, as shown in the photo of the chassis taken from above. The rod is original, and part has not broken off, as you might think. Unusually, the converter feeds to a simple LC IF circuit (L3), then capacitively couples (via 30pF C8) to the IF amplifier grid. IF amplifier valve V2, a 12BD6, has similar characteristics to the better-known 6BA6/12BA6. Like the converter, the IF amp runs without cathode bias, but the combination of the AGC line’s -0.4V and contact potential bias across 1MW resistor R2 sends the 12BD6 signal grid to around -1.1V (like the converter’s). The IF amp feeds a conventional IF transformer (IFT1) with a tuned, untapped primary and secondary and ferrite core adjustments. The signal from IFT1 feeds to both diodes in V3, the demodulator/first audio valve, a 12AV6. Capacitor C9 does the IF signal filtering, and the audio signal is developed across 500kW volume control potentiometer VR1. The AGC control voltage is picked off and sent back to the IF amp and converter via 2MW resistor R3, with the audio signal filtered out by 50nF capacitor C2. The audio signal from the volume control is fed to the 12AV6 triode’s grid via 5nF capacitor C10. Contact potential bias for the V3 triode develops siliconchip.com.au across 5MW resistor R5. The amplified audio output from its anode is fed to output valve V4 via 5nF capacitor C12, and any remaining IF signal is filtered out by 250pF capacitor C11. The 50C5 output valve’s grid returns to the chassis via R6, and the stage is cathode-biased by R7. Unusually, there is no cathode bypass capacitor. The circuit for a similar Monacor set shows output transformer T1 with a primary impedance of only 2.5kW, further confirmation of the special characteristics of the 50C5 and its low-voltage applications. ‘Full HT’ types, such as the 6V6/6AQ5, commonly require load impedances in the 5~6kW range. The mains supply connects directly to the anode of the 35W4 half-wave rectifier (V5) and to the series heater chain. This chain has the rectifier first in line, then the output valve. In common with battery-powered sets, the 12AV6 audio amplifier is the last in the chain, so that one side of its heater connects to ground, minimising any induced mains hum. The 35W4 rectifier supplies a halfwave rectified output to the first filter capacitor, C15 (30μF). This point directly supplies the output valve anode. Although the supply is not fully filtered, output pentodes are not very sensitive to power supply hum. This connection has the advantage of taking off the largest current The top view of the chassis shows the very short ferrite rod antenna, which gives mediocre performance consumption before the series filter resistor. To place the anode after the filter would increase the filter’s voltage drop by three or four times. Although the filter resistor R8 has a high resistance of 1kW, the current drain from the converter, IF amp and first audio is modest, so the filter only results in an HT drop of about 30V. The circuit diagram’s voltage callouts show the effect of AGC: on strong signals, the reducing current draw from the signal part of the converter, plus the IF amplifier, allow the RF/IF/ Audio HT to rise by around 20V. Such a voltage rise on strong signals is common, it’s just not often reported. Point-to-point wiring is used in the underside of the “Monarch Wedge” chassis, which truly is wedgeshaped. siliconchip.com.au Australia's electronics magazine The entire circuit is isolated from the chassis metalwork. I have used the ‘ground’ symbol for power and signal returns. Capacitor C1 connects the isolated ground to the chassis at radio frequencies. Restoration The original figure-eight mains cord had shed its insulation just as it emerged from the chassis, shorting it out. It’s a stark reminder to never just plug in a set in unknown condition! Fortunately, the mains cord was secured by a two-part cord anchor, so it was easy to replace the original figure-eight with a new section and secure it against movement. Given the set’s age, I was a bit apprehensive about the valves. Happily, all five tested good after a bit of time on the tester. This is a common as oxide-coated cathode formulations include barium, a highly reactive element. Barium is so reactive that barium powder scattered on a benchtop will spontaneously burst into flames! During manufacture, the applied coating contains the metallic oxides as inert carbonates. After the envelope is evacuated, induction heating and heater activation achieve two outcomes: any entrained gases in the structure ‘boil out’ and are drawn out with the evacuation, and cathode carbonates reduce to oxides. Normal operating temperatures January 2025  99 maintain the oxide compounds, but, on cooling, the highly reactive oxides tend to absorb any residual gases not already ‘cleaned up’ and oxidise to more complex compounds. Such absorption compromises the cathode’s emission and degrades performance – it’s known as cathode poisoning. Rather than a random occurrence, it’s common with valves that have been left unused for extended periods. Fortunately, all that’s needed in most cases is a few minutes of normal operation, and the cathode coating will reduce back to simple oxides. ‘Rejuvenation’, a period of over-running the heater, can accelerate the process. The valve sockets all required a good contact clean. I like to leave the radio off, applying my BWD 216 0~400V power supply to the HT line to test for electrolytic capacitor leakage. There was more than acceptable, but I left power applied, and the current fell as the two filter capacitors reformed. It all seemed to be working OK, and only needed an IF and antenna/LO alignment. The IF was a bit off, but I was able to calibrate it without difficulty. Remember that it’s important to do this for a low output, maybe 10mW, and to reduce the input signal to keep the output low as the set comes into full alignment. This is because, with higher level input signals, the AGC action ‘mushes out’ the tuning response, making the optimal peak difficult to adjust to. Like many sets, there’s no effective antenna alignment at the 600kHz end of the band, so it’s a matter of tuning to 600kHz, then tweaking the LO coil’s slug and looking for an improvement in sensitivity. The top end had trimmers on both antenna and LO sections. I simply used the LO trimmer to align to 1600kHz for full dial rotation, then dropped back to 1400kHz for the antenna trimmer. As usual, I did both ends a few times, as there is some interaction between adjustments. cathode resistor. Popping in a 470μF bypass cap brought the output stage gain up by a factor of two, doubling the sensitivity at every point, including RF sensitivity. Unmodified, its sensitivity (for a 50mW output) was 1.6mV/m at 600kHz and 550μV/m at 1400kHz. The signal+noise-to-noise ratio was 20dB or better in both cases. The IF bandwidth was ±2.2kHz at -3dB; at -60dB, it was ±39kHz. The audio response from antenna to speaker was 210-2500Hz, with a 2dB peak around 1kHz. From volume control to speaker, the audio response was 240Hz to 10kHz. Total harmonic distortion (THD) at 50mW output was 8%. The maximum output was 0.9W at 10% THD. The signal sensitivity was a bit underwhelming, and I suspect that the main culprit is the very short ferrite rod antenna, combined with the lack of cathode bypassing on the output valve. My final test demands good reception from Warrnambool’s ABC station, 3WV, at 594kHz. It was present, but only at full volume, and noisy. This compact marvel is, indeed, just what it appears to be: an economy ‘city and suburbs’ radio. Special handling Although the chassis metalwork is isolated from the chassis, this transformerless set presents an electrocution hazard. Any work with power applied must be done using an isolation transformer. Be aware that variacs and other autotransformers do not give electrical isolation. Would I buy another? I already have this example, but I’m interested in the idea of mass-­ produced minimalist radios. To me, it’s a continuation of the VE301 Volksempfänger (February 2023; siliconchip.au/Article/15671), DKE38 Kleinempfänger (July 2017; siliconchip.au/Article/10728) and their English counterpart, the “Wartime Civilian Receiver Utility Set”. Given that I only need the English unit to make a complete set, I might just check eBay once in a while. Further reading The set appears as the Monarch Wedge on Radiomuseum (siliconchip. au/link/ac1y). There’s also a 220V version, which looked identical at first glance. How did they soak up the extra mains voltage? A series resistor? No. An old and rarely used trick – pop in a series capacitor with the required reactance! Just 2.1μF will do. See siliconchip.au/ SC link/ac1z How good is it? For what it’s meant to be, pretty good. I did notice the lack of a bypass capacitor across the output valve’s 100 Silicon Chip Repairing the mains cord and cleaning the valve sockets had the radio back to operating condition. Australia's electronics magazine siliconchip.com.au