Fullscreen HTML5Med Res (??MB)HTML4

Vintage Radio By Maurie Findlay, MIE Aust, VK2PW The Hotpoint Bandmaster J35DE console radio Over the next few months, veteran radio designer Maurie Findlay will go through the process of restoring a good “1940s wireless” to its original performance – and for those who are interested, he describes how to make it perform even better than new. The radio to be restored is a Hotpoint Bandmaster which was made in console (J35DE) and table (T55DE) models. Maurie takes up the story . . . The Hotpoint Bandmaster J35DE was a 1940s console radio that offered quite good performance in its day. This example is still in good condition, although the grille cloth needs replacing and the cabinet requires work. 92  Silicon Chip W HILE THERE were many run-ofthe-mill radios produced during the valve era, those with better performance were considerably more expensive and are now hard to come by. And while the sets made by AWA were highly regarded, those branded Hotpoint would these days hardly rate a second glance by vintage radio collectors. However, they would be missing out. Hotpoint-branded radios were made by AWA Pty Ltd (Amalgamated Wireless Australasia), Australia’s biggest electronics company in the 1940s. Which just goes to show that “badge engineering” was not confined to the automotive industry. The Hotpoint Bandmaster T55DE/ J35DE is a 5-valve radio offering AM broadcast band and shortwave reception, with provision for a pick-up to play records. The chassis may also have been the basis for radiograms made by AWA at the time. The Hotpoint J35DE/T55DE chassis was virtually identical to that in the AWA 721-C console radio and the 618-T mantle (or table) radio. A set of this general type, in good order, will have a reserve of performance for local broadcast stations and will receive the stronger shortwave stations. With care and patience, the valves and other components can be tested, replaced if necessary and the set realigned for best performance using no more than a multimeter. That said, the meter needs to be a modern digital multimeter. Multi-range meters available at the time the Hotpoint was designed mostly used a moving coil meter which required a current of 1mA for full-scale siliconchip.com.au deflection (FSD). Such a meter would give readings very much in error in many radio circuits because of the high resistances involved. For example, take a look at the circuit diagram of the radio featured in this article. At valve V3’s plate, it would read about one third of the actual voltage on the 100V scale. That’s because the relatively low impedance of a moving coil multimeter loads down the voltage when attempting to measure such a circuit. By contrast, a modern digital multimeter has an input resistance of 10MΩ (100 times greater) and would have very little effect on the voltage. Apart from a good digital multimeter (DMM), you will need spare parts, small hand tools and most important of all, some skill with a soldering iron. Still, if you have assembled a typical PCB, you should have no trouble soldering parts in an old radio chassis. However, you will need a bigger soldering iron to do some of the work. The Hotpoint T55DE is typical of 5-valve sets made in the valve era. It used good quality components which were operated conservatively and offered what most owners wanted: reliable reception of the local broadcast stations. More elaborate receivers, for use in remote areas, would have had an extra stage of amplification between the aerial and the mixer stage. For those needing high volume, a more elaborate audio system, perhaps using push-pull valves, would be prescribed. In addition, shortwave reception could be improved by incorporating a bandspread system so that particular frequencies can be tuned more easily, while an extra RF amplifier stage is also a big advantage at the higher frequencies. And so it goes on. The aim of this article and the one that follows is to give enthusiasts, with only a basic knowledge of radio, a systematic means of restoring vintage receivers to full performance. A particular set has been chosen in order to avoid a string of generalities which could easily have been confusing. I have redrawn the manufacturer’s circuit diagram, with component values marked, to avoid the need to refer to the parts list when studying the diagram. Circuit details Let’s start by going through the varisiliconchip.com.au This view shows the neat arrangement of the major components on the top of the chassis. A label on the dial backing plate shows the drive cord arrangement. The old Hotpoint featured a rather elaborate glass dial which carried markings for the Australian states, New Zealand and the international shortwave band. ous stages of the Hotpoint T55DE. Fig.1 shows the circuit details. Valve V1 is the mixer, sometimes called the 1st detector, and is a 6J8G. It takes the signal from the aerial (antenna) and converts it to an intermediate frequency which makes it easier to obtain the amplification and selectivity required. The 6J8-G has a special “heptode” construction which consists of a fine helix grid close to the cathode, a screen grid surrounding that and yet another screen grid followed by the suppressor grid and then the plate. In between the two screen grids is another grid which is connected to the grid of a separate triode element. This sounds complicated but this construction allows the local oscilJuly 2011  93 Fig.1: the redrawn circuit for the Hotpoint Bandmaster J35DE radio. It’s a fairly conventional 5-valve superhet configuration with AGC and a 455kHz IF. The set can tune both broadcast and shortwave bands. lator, using the triode section, to function with full efficiency, while mixing of the two signals takes place in the electron stream from cathode to plate. Several frequencies appear at the plate but the one we want, the difference between the signal and the higher oscillator frequency, is selected by the 455kHz tuned circuit. 6J8-Gs cost more to manufacture than other valves designed to do the same job but this valve worked better than most, particularly on the shortwave bands. It was often used in quality receivers manufactured at the time. The next stage, V2, uses a 6SK7GT pentode. The internal shielding between the control grid and the plate is provided by the usual screen and suppressor and the valve is able to amplify in a stable fashion. Other valves, such as the 6U7-G, available at the time, could have done the job equally well. An important requirement for this 94  Silicon Chip stage is that the valve has a “variable mu” characteristic; the gain reduces as the negative bias on the grid increases, which allows for automatic gain control (AGC). An interesting point about the design of the Hotpoint circuit is that AGC control is applied to both the 6J8-G and the 6SK7-GT on the broadcast band but only to the 6SK7-GT on shortwave. This allows greater amplification for the weaker shortwave signals. Another special design point is the filter in the broadcast band aerial circuit. A 50pF capacitor in series with a high-Q inductor forms a series tuned circuit at 455kHz, effectively shorting out the receiver input at that frequency. Not many designers would have considered this necessary, because 455kHz is kept clear of highpower transmitters. As was conventional at the time, this set has four circuits tuned to the intermediate frequency of 455kHz, two before and two after the amplifier. Coupling between the circuits was loose enough for the circuits to be tuned individually without affecting each other. The resultant selectivity caused attenuation of the higher sidebands and hence a reduction in the higher frequency audio. At the time, few designers would have incorporated a bandpass arrangement. People seemed to think that radios should have a “mellow” tone. Detection & AGC The next valve, V3, a 6SQ7-GT, incorporates two diodes and a triode. It recovers the audio from the intermediate (IF) signal, provides the automatic gain control and amplifies the recovered audio signal. Other valve types capable of doing the same job were available at the time. For example, a pentode double-diode could have been chosen for higher gain. But the triode provides a reserve siliconchip.com.au of gain anyway, with a very simple circuit. The small amount of negative bias required is obtained from a high-value resistor in the grid circuit (10MΩ). The diode connected to pin 5 of V3 rectifies the 455kHz signal from the IF transformer and the recovered audio signal appears at the lower end of that transformer. At this point, the audio is mixed with the 455kHz IF signal and a filter, consisting of a 47kΩ resistor and two 100pF capacitors, removes this 455kHz component. With the function switch in the “radio” position, the recovered audio appears across the 500kΩ (0.5MΩ) volume control potentiometer. AGC is developed by the diode connected to pin 4 of V3. However, pin 4 is returned to a voltage that’s negative with respect to the cathode via a 2.2MΩ resistor and therefore does not start developing an AGC voltage until a certain signal level is reached. This is “delayed AGC” and ensures that maximum gain is available for very weak signals. Output stage V4, the audio output valve, amplifies the signal further and provides power to drive the loudspeaker. It is a 6V6-GT and was the best choice for the job at the time this set was designed. In this set, it is operated with a cathode bias resistor that’s slightly larger in value than usual. This reduces the power dissipation and audio output of the valve but would make for longer life. The optimum load resistance with the higher bias resistor would be higher than the usual 5kΩ and is probably somewhere around 7kΩ. Design fault This circuit has a serious design fault concerning the arrangement for connecting the speaker. The output transformer is mounted on the back of the speaker and is connected to the output valve via a plug and socket arrangement on the chassis. As a result, if the set were to be accidentally switched on without speaker connected, the 6V6-GT screen current would be very high and this would probably ruin the valve. A better arrangement would be to have the speaker transformer permanently mounted on the chassis and the voice-coil leads extended. Alternatively, a solution such as that described on page 91 of the August 2010 issue siliconchip.com.au The chassis mounts vertically inside the cabinet, so that the glass dial and control shafts face upwards. Note that the output transformer is mounted on the speaker frame. This means that the 6V6-GT output valve could be destroyed if the speaker cable is disconnected from the chassis while power is applied. of SILICON CHIP, for an Airzone 612 console radio, could be adopted. This is what we eventually did with this Hotpoint set. Power supply A power transformer and rectifier valve, V5, are used to derive the 240V DC high-tension supply for the amplifying valves. However, the usual approach has not been taken. V5 is a 6X5-GT and this valve has special insulation, designed to withstand the high voltage between the cathode and the 6.3V filament. The 6.3V heater winding on the transformer also supplies the other valve filaments and is effectively at chassis potential. The alternative approach, and the one mostly used in sets at this time, was to use a 5Y3-GT rectifier which has directly-heated cathode supplied from a separate 5V heater winding. This could then “float” at the HT voltage which could be anywhere from 100-300V or more, depending on the circuit requirements. The 100Ω resistors in each plate circuit of V5 are provided to limit the peak current. The 8µF capacitor connected to V5’s cathode also affects the peak current and hence the life of the valve. It should not be replaced with a higher value. To complete the circuit description, note the function switch which allows the set to be switched for radio or record pick-up operation. There are July 2011  95 By contrast with the top, the underside of the chassis is quite crowded due to the bulky old-style components used. Note the primitive technique used to anchor the power and speaker cables, ie, by tying knots in them. The underside of the chassis is protected by a perforated steel cover, a rather unusual feature for radios of that time. tone control positions for both radio and pick-up. In the pick-up position, the screen supply to V1 and V2 is disconnected so that “play-through” from the radio stage, due to stray coupling, is eliminated. The pick-up input was designed to accept the high-output signals from the crystal (piezoelectric) pick-up cartridges used in the 1940s with 78 RPM records. Not every restorer will want to bring this back to life! In addition, the treble cut applied for radio listening is probably too severe for modern ears and could be reduced by choosing a smaller value for the associated .01µF capacitor. A power socket for a turntable motor is mounted on the chassis and is alive even with the radio switched off. When replacing the power cord, we 96  Silicon Chip used the socket as a convenient termination. However, there is a safety issue here in that the metal terminations in the socket are close to the metal surface on which the socket is mounted. If the bare wires are not pushed right into terminations, there is the possibility of them touching the metal chassis with disastrous results. Such a socket would definitely not meet approval today. Preferred value components In the original service manual for the Hotpoint Bandmaster J35DE, all the passive components, ie, resistors and capacitors, are in “non-preferred” values. For example, one resistor is specified as 2.5MΩ while others are marked 1.6MΩ, 50kΩ, 32kΩ, 25kΩ and 20kΩ. In the capacitor list, there is a 50pF unit, a 70pF (actually µµF) unit, some .05µF units and so on. This is because this set was made before the introduction of the “preferred value” system, which is now universally used for small components. With preferred value numbering, a designer can adjust a circuit value to a desired order of accuracy while stocking the minimum number of components. The numbers in the ratios 10, 15, 22, 33, 47, 68 and 100 would be stocked by a design laboratory over most of the range, except for very small and very large values. On the other hand, for very critical circuitry, a designer may need to stock values in finer increments such as 10, 11, 12, 13, 15, 16, 18, 20, 22, 24, 27, 30, 33, 36, 39, 43, 47, 51, 56, 62, 68, 75, 82, 91 and 100. However, extended over the decades, this could involve a huge number of components. So what do we do about, say, replacing the 50kΩ resistor in the grid circuit of the 6V6-GT output valve? You cannot buy a 50kΩ resistor at your usual supply store. The answer is that the exact value is not critical and a 47kΩ resistor will do the job perfectly well. This also applies to most of the other components in the radio. The 2.5MΩ resistor could be replaced with 2.2MΩ, the 1.6MΩ with 1.5MΩ and so on. In addition, the .0025µF capacitor from the plate of V4 to ground can be replaced with a .0022µF capacitor with negligible effect on the way the radio works. With this in mind, the circuit presented in this article has been redrawn with “preferred values” for most of the passive components. There are, however, some components where accuracy must be maintained. The 4000pF (4nF) capacitor in the shortwave oscillator circuit is an example. It modifies the tracking of the oscillator frequency to give the desired tuning range. The same applies to the capacitors in the 455kHz IF coils (unmarked). Next month Next month’s article will describe the practical side of getting the Hotpoint Bandmaster into operation. It is now 60 or more years since the set was manufactured and that meant that a great deal more than defective valves had to be considered. Many capacitors, resistors and even the wiring had SC deteriorated badly. siliconchip.com.au