Fullscreen HTML5Med Res (??MB)HTML4

Vintage Radio 1940 1940 AWA AWA “Fisk” “Fisk” Radiola Radiola model model 501 501 By Associate Professor Graham Parslow The Radiola model 501 is the console version of a series of similar circuit designs by AWA. It’s featured in a simple timber cabinet with a 12-inch Rola speaker. It measures about 86.5cm tall and weighs around 16.3kg. Console radios were used as display centrepieces through the 1920s and 1930s. Some elaborate examples of the carpenter’s craft in making large cabinets now look hideously overornamented. We have progressed to an age where we accept the minimalist styling made popular by Scandinavian designers in the 1950s and beyond. Paradoxically, this now elevates the simple design of the model 501, compared to how it would have been considered in 1940, sitting beside other elaborate and more expensive consoles in a shop. The simple dial and escutcheon and the lack of wave-changing also contribute to the economy of this model. However, the sound is equally as magnificent and sumptuous as the expensive models, because that quality is largely determined by a 12-inch (300mm) speaker mounted on a reasonable baffle. In the late 1930s, AWA took the same chassis and components and packaged them as several different models: the model 301 (a radiogram), models 84 & 194 (mantels; the 194 is incorrectly listed as a console radio in some service manuals), and models 193 & 501 (consoles). This is a clever use of resources; just as car manufacturers don’t design a bespoke engine or transmission for every model, why should a radio maker design unique circuitry for each set, just because its cabinet is a different shape? Radio evolution in the 30s The set has a width of 58cm and depth of 30cm, with the front-facing veneer likely stained white oak, and the sides Queensland walnut. siliconchip.com.au Australia’s electronics magazine The 1930s was a decade of remarkable evolution in radio design and presentation. At the beginning of the decade, floor-standing console models looked like cabinets with ornate elaborations featuring sculpted wooden feet, which would have suited a lounge chair. Most radios in the early 1930s were based on tuned radio frequency (TRF) circuits that used multiple tuned October 2020  85 stages to achieve selectivity between stations. The valves in those sets had four or more pins at the base and a range of matching sockets. That would soon change as the eight-pin octal base became a standard that would prevail for twenty years, before 7-pin and 9-pin miniature valves took over. Many heritage valves were simply repackaged with an octal socket, like the 5Y3 valve in the radio featured here. The 5Y3 was designated type 80 when produced with a four-pin socket. The patent problems preventing Australian manufacturers producing superhet radios were resolved in 1934. The all-octal-valve model 501 encapsulates the change from the dominance of TRF radios to mature superhet technology in only six years. Circuit details A paper label glued inside the lefthand panel (reproduced overleaf) shows the complement of valves and the location of components on top of the chassis. The same label was attached to all models sharing this chassis. The circuit is a fairly basic superhet design with a mixer/oscillator stage based around a 6A8G pentagrid converter valve, one 455kHz IF gain stage using a 6U7G variable-mu pentode, a detector/audio preamplifier stage based on a 6B6G dual diode-triode and a Class-A audio output stage using the 6F6G power pentode. The set has delayed AGC, and an HT voltage of 265V once warmed up. The circuit diagram for the model 501 was drawn to suit four differ- The chassis is mounted in the typical location for console radios, with the 12inch Rola loudspeaker below it (marked type AS7 in the service manual). 86 Silicon Chip Australia’s electronics magazine ent models. The circuit also serves the model 301, but that was drawn separately to include the gram-radio switching circuitry. The main drawing is for models 84 and 194, with alternative wiring to represent console models 193 and 501. Starting at the aerial, it can be seen that the smaller cabinets incorporated a loop antenna that served as the tuning inductance for station selection. The loop antennas worked well in strong signal areas. The 501 has a conventional aerial coil with the secondary acting as the tuning inductor. The tuning capacitor connection to the grid of the 6A8 and the first IF transformer connection to the grid of the 6U7 are both top cap connections. This minimises stray capacitance that can cause unstable operation. All of the top caps are at low voltage (usually 3V or less), so accidental contact is not dangerous, but ill-advised as a general principle, because top-cap anodes on output valves can be lethal! The local oscillator is L5/L6 with the tuned section L6 providing a frequency that is 455kHz offset from the station frequency. L5 functions to provide positive feedback to sustain oscillation, a method developed by Edwin Armstrong, the acknowledged inventor of the superhet principle. After IF amplification by the 6U7 valve, D2 of the 6B6 detector provides a negative AGC voltage. This is directed to the control grids of the 6A8 and 6U7 via R5. The negative potential across R13 ensures that the AGC is delayed until stronger stations are tuned. D1 in the 6B6 feeds the detected audio signal to 500kW potentiometer R6, which then goes to the 6B6 triode section via 100nF capacitor C16. A more advanced circuit might have featured negative feedback from the output and bass-enhancing circuitry acting at this audio preamplification stage. The 6F6 output pentode is a solid performer, easily putting out 3W at the voltages used in this radio. The 6F6 is unlike most other common output valves in requiring a high grid bias voltage, specified as -17V for this radio. Economy of components is achieved by tapping the 6F6 grid bias off the HT line using a 300W resistance (R12 plus R13), inserted between the power transformer centre tap and earth. This eliminates an electrolytic capacsiliconchip.com.au This circuit diagram was scanned from the Australian Official Radio Service Manual Vol.4 (indexed under the Fisk name at the time). There are two notes just below the circuit which state: “L3,L4,C1 replaces loop L1,L2 on console models” and “external speaker connections included for console models”. Both of these obviously apply to the 501. itor that would otherwise be needed across a bias resistor in series with the 6F6 cathode. The tone control is the ultimate in economy, featuring a three-position switch that offers either no top cut, top cut via C21 (35nF), or less top cut with R11 (5kW) in series. In practice, this gives reasonable choice. The electrodynamic speaker has a 2kW field coil that generates the magnetic field and also acts as a filter choke for the HT line. The speaker is mounted in the lower section of the siliconchip.com.au cabinet, and is connected via a plugin four-conductor cable that delivers audio from the 6F6 anode (plate), two HT lines, plus an earth strap for safety and hum minimisation. Construction details Both the IF amplifier and audio preamplifier valves are in Earthed canister shields, serving to minimise hum and maintain stable performance. As mentioned earlier, the first three octal valves also have topcap control grids that allow for short Australia’s electronics magazine lengths of wire to their signal source when that source is mounted above the chassis. C22 is a large aluminium can capacitor mounted adjacent to the tuning capacitors, and is the only component not placed next to its area of function. Although it has no markings, the component list specifies it as 8µF 500VW. Electrolytic capacitors were large bulky components at this time, and C22 needed to cope with the high start-up voltage generated by the 5Y3 rectifier. October 2020  87 The AWA 501 chassis underside with a matching layout diagram shown below. These are from the service manual and can be found at www.kevinchant.com The inclusion of valve base pin labels is a welcome addition when checking sets. The centre of the chassis is reserved for a pressed dome with parallel ventilation slots, below which the power transformer is mounted. The downside of this arrangement is impaired heat dissipation and a cluttered underside relative to top-mounting the transformer. But the advantage of having the power transformer under the metal chassis is shielding of any 50Hz radiation that might create hum if the transformer was top-mounted. Restoration I bought this radio from a secondhand barn at Watsons Creek near the Yarra Valley, in Victoria. I remember the young salesman urging me on with “go ahead and buy it, you know you want to”. Somehow this also mollified my wife, so it came home with us. That was twelve years ago, when my primary interest was to restore the cabinet to glory. I did that by completely stripping the cabinet to bare timber and spraying it with satin polyurethane. To my eye, the character of the veneers gives great presence to this radio. The radio had its mains cord cut off, and no speaker was fitted, so I elected to leave it as it was. This may seem 88 Silicon Chip like sacrilege to some, but I installed a digital stereo AM-FM CD player with its speakers all mounted in the lower space of the cabinet. A 6V transformer to power the dial lamps made it look like the radio was functional when the transplanted hifi system was switched on. The radio then stood patiently in a corner of our back entertainment area, niggling away at me to do a proper restoration. COVID-19 restrictions brought the niggle to a climax. The first step was to remove all the valves. This revealed that the top cap of the 6A8 mixer was missing and an ingenious handyman had used tape to wedge the top cap connector around the glass nipple at the top. This could Australia’s electronics magazine only have provided capacitive coupling to the grid, because the grid wire was eroded back inside the glass envelope beyond the possibility of making a new connection (a conclusion reached after breaking the envelope). Luckily, I had a replacement 6A8 in stock. At first glance under the chassis, it looked like the restoration would be straightforward due to most of the original components still being in place. Only one capacitor, coupling audio from the 6B6 to the 6F6, had obviously been replaced. Several paper capacitors looked like the pitch sealant at the ends had dried and failed. I replaced all of these, except the audio coupling capacitor that had previously been upgraded. siliconchip.com.au The AWA 501 chassis with a layout diagram shown below. The capacitor at the tone control switch measured as a dead short, and the others exhibited various grades of leakage. It is a miracle that this radio could have struggled on with so many marginal components and faults that became evident later. R2, specified as 20kW 1W, was two 40kW resistors in parallel with the identity colours burnt off by sustained heat. Even so, the value was still correct. Nevertheless, I replaced them with two 10kW 2W resistors in series. I fitted a permanent magnet 12-inch Rola model M as the new speaker. This required fitting two 1kW 7W resistors in series in place of the 2kW electrodynamic speaker coil. This pair of resistors can be seen mounted on top of siliconchip.com.au the chassis, next to the dome of the power transformer. As a result of this substitution, I needed to fit a new speaker transformer, but there was no convenient mounting position available either above or below the chassis. I decided to mount it at the side of the chassis as this meant that the chassis metalwork would act as a shield against any hum radiating from the mains transformer. I removed 8µF electrolytic capacitor C22 and replaced it with a modern 47µF 450VW electrolytic capacitor. That high voltage rating is essential because the power transformer produces 2 x 370V AC for rectification by the 5Y3 valve, resulting in a measured switch-on DC voltage of 450V DC, re- Australia’s electronics magazine ducing to 375V when other valves begin conducting. The 2kW field coil replacement resistor drops the main HT line to 265V. The last operation before switchon was to add an Earthed three-core power line. I do this last because the cord gets in the way needlessly if done earlier. Initial switch-on was a singular disappointment – nothing happened! The faults included an open-circuit R8 feeding HT to the 6B6, and a dead 6B6 valve due to an open filament. Leakage through C18 reduced the 6F6 bias to 0V, a situation where the valve is forced into potentially destructive high conduction and is ineffective as an amplifier. C18 was the “new” capacitor that I had not bothered replacing. There was also a fault in the volume control resistance track; it had lost contact with the lug connecting it to the audio feed from the second IF transformer (L10). Once I had fixed all those problems, it came to life. I then aligned the IF transformers, resulting in significantly better performance. Finally, my tribulations were repaid by having a grand icon of its era working superbly well. SC October 2020  89