Fullscreen HTML5Med Res (??MB)HTML4

Vintage Radio HMV 42-71 dual-wave superhet receiver By Marcus Chick This radio by His Master’s Voice was made in Australia from 1954 to 1959, using HMV’s type 42 chassis. It’s a mostly standard mains-powered set with MW and SW reception, but a few surprises are hiding within. T his set came to me due to an acquaintance downsizing and moving into a retirement village. It was described to me as a 6V radio. However, on collecting it, it was apparent that it was mains-powered. Editor’s note: this set was previously reviewed by Rodney Champness in the August 2003 issue (siliconchip.au/ Article/5648). You can read that article for a more detailed breakdown of the model 42-71’s circuit. The model 42-71 is a typical Bakelite table radio of the day, featuring the then-recently-introduced miniature valves. For a multi-band set, it siliconchip.com.au is unusual as this set is in the higher price range. While it is designed as what the Americans called “farm radios”, it does not have the usual RF amplification valve preceding the frequency-­changer valve. Nonetheless, it is an attractive piece. Not uncommon for the era, the set shared its chassis with other cabinet shapes to become ‘different models’. After all, why reinvent the wheel? In this case, the chassis is type number 42. You can get the service manual for that chassis from Kevin Chant’s website at www.kevinchant.com/ hmv3.html Australia's electronics magazine Uniquely to HMV, it uses 457.5kHz for its intermediate frequency (IF), whereas most in that era used 455kHz. There was sound logic behind that. The objective was to ensure that none of the sub-multiple frequencies produced by the oscillator fell on a radio station, especially the one you were trying to listen to. There were plenty of national and international radio broadcasters back then. To accommodate that, there were three shortwave (SW) bands because changes in the time of day, sunspot activity and weather all conspired to render some bands inoperable. August 2024  87 SW1 covered 14.2 to 18.4MHz, SW2 covered 24.79 to 31.92MHz and SW3 covered 5.9 to 7.5MHz. SW3 is ‘band spread’, so fewer station frequencies are over a larger area of the dial. Then we have the broadcast band of 540–1600kHz, which had its station spacing reduced from 10kHz to 9kHz in 1978. That is why only 3SR and 2AY are anywhere near their original positions. Circuit details Fig.1 shows the set’s circuit, which follows the general plan for a superheterodyne radio of the day. V1 (6AN7), the ‘frequency changer’, is actually two valves in the same envelope. This was a 88 Silicon Chip step forward as the triode in the 6AN7 complements its hexode by becoming a separate exciter for the oscillator. Valves of this type were used in shortwave sets as they provided better modulation (and thus better performance) at higher frequencies. Restoration A cable with a mains plug is no guarantee of its actual operating voltage, so I needed to assess it first. I never plug a set in to see if it works; I cannot viably repair some sets when they self-destruct after being plugged in like this. Many of these sets were abandoned after they broke down, and as we who fix know all too well, Australia's electronics magazine certain bits deteriorate when the set is not used for a long time. I concluded that the incorrect description came from the fact that the visible valves have sixes written on them. In other words, the heaters ran from 6V, not the whole set. The set only had one knob attached, which was not entirely unusual, as they were the long-shanked plastic types with a clamp. Most of those were bad news. Rigid plastic and movement were never a successful combination, and that plastic does decompose and go brittle. Under the dirt, there appeared to be an almost-mint Bakelite cabinet. So, after separating it from the chassis siliconchip.com.au ◀ Fig.1: the circuit diagram for the HMV 42-71. Its most notable features are the multiple wafer switches for selecting between the MW, SW1, SW2 and SW3 bands and the somewhat unusual 457.5kHz intermediate frequency. Unlike some similar sets, this one lacks an RF amplification stage. The chassis diagram for the set is shown at lower left. and giving it a shower to get the dust off, I decided it was in better condition than I first thought. The chassis (which had not been showered) was also in reasonable condition. I took a good look at the chassis and noticed several waxed paper capacitors as well as obvious heat damage to the transformer wires. There were also three aged Ducon electrolytic caps, plus the mains cable didn’t seem to be in great condition. There was no way I was going to power it up in this state. General restoration advice I have no tolerance for wax paper and some oil-filled caps; they inevitably become electrically leaky. I do not siliconchip.com.au bother testing wax paper caps; finding one good one in probably five hundred is not efficient, so I replaced them all. My “Honor” (Lafayette) RC Tester manual quotes a non-polarised capacitor with a leakage resistance below 50MW (at valve working voltages) as unsuitable for screen decoupling, and less than 200MW unsuitable for coupling. Therefore, I will not tolerate a leaky non-polarised capacitor, regardless of whether it can be made to function. Consider that in a set like this, the grids will draw next to no current. So any positive voltage applied to the control grid will destroy its bias and can, or will, damage the valve. Even if Australia's electronics magazine it doesn’t, it will ruin its performance. I usually touch a battery across the output transformer’s primary to ensure it is working and then perform resistance checks on primaries and secondaries. That eliminates a lot of time-wasting and potentially damaging rework later. Most Australian mica caps from the late thirties are reliable, and they should not touched, unless one end is out of the circuit. If one of its wires is out of circuit, I perform an insulation test to ensure there is no leakage. I absolutely do not perform this test on any capacitors that are across tuning coils etc (the coils and gang are a matched set). Since the mica capacitors are installed during manufacture, tampering with them is liable to cause the set to not work properly. The coils and gang are a matched set. They have mica caps installed during manufacturing to meet their specifications and tampering with them can have catastrophic results. Amazingly, there was one early modern type capacitor in this set. It stayed while the rest went. I was surprised to find no resistors worth changing as I went through the caps. As the transformer wires had succumbed to the heat of the rectifier and output valves, I cut the wires short and spliced in new lengths. Adding some heat shielding I cut out a piece of spare sheet metal from some shed doors to make a heat shield to fit between the 6N8, 6M5 & 6V4 valves and the transformer, protecting the new transformer wires. You can see it in place in the photo of the rear of the chassis. The mains cable had also been affected by the heat from the rectifier, so I took the opportunity to cut off the supply wires to the gramophone socket and, as there was room, reroute the new Earthed mains cable along the side of the chassis with clamps (you can also see this new arrangement in that chassis photo). August 2024  89 90 Silicon Chip Australia's electronics magazine siliconchip.com.au These three photos show the HMV 42-71 set after cleaning it, replacing all the old paper capacitors and some of the resistors. The mains cable was also replaced, as the previous one was too degraded, and lastly you can see the heat shield behind the 6N8, 6M5 & 6V4 valves in the photo at lower left. The P-clamp that holds the mains cable is too large, and uses a ‘not-to-standard’ cable tie to prevent it from pulling out. Best practice would be to use an appropriately-sized P-clamp to hold the cable securely. This was also the point that I decided that I needed to find out what model the set actually was, to ascertain the IF. As mentioned earlier, it is actually 457.5kHz, rather than the common 455kHz of most sets of that era. I noted that the circuit required R11, R12 & R13 to be three 10kW resistors in parallel. Interestingly, this set only had two. I left it as was. After completing a refit, I attached an analog meter across the B supply to monitor it and powered up the set via an isolation transformer with a kill switch. I also have neon lamps in bezels to the primary and secondary of the isolation transformer so I can quickly see if voltage is present, along with fuses on both primary and secondary to protect the transformer itself. Powering it up The start-up was perplexing. There were no dial lights; the heaters were glowing, but there was no HT. Blown dial lights are typical, so I fitted new globes. There was power on the rectifier, but nothing coming out of it. I hunted down another 6V4, as my “Knight” tester will not check a 6V4. With the new valve, I got a reading siliconchip.com.au on the B voltage, although it was low. That was obviously due to the missing 10kW resistor, so I replaced the two in the set with three new ones in parallel. Calibration It quickly became clear that someone had been playing with all the adjusting screws. Initially, the calibration did not go well; a symphony orchestra of various noises was getting in via the 36m antenna. A new G10 LED light was sending out a mass of RFI. Like the other noisy LED lights I’ve purchased, I returned it for a refund. That is why I use a halogen desk light. The computer’s UPS was also chipping into the EMI cacophony. Clearly, regulations around RFI no longer exist or are being ignored. AM radio is being pushed out to hide the fact that RFI is out of control. After killing power to all the noisy lights, plugpacks and such, I managed to get the calibration done. That improved things immensely. I did it with an entry-level signal generator, calibrated with a Fluke frequency counter and an oscilloscope as the meter. The ‘scope is also helpful for Australia's electronics magazine checking for gross distortion and, if present, helping to find its source. The signal generator must not be calibrated with the modulation tone on. My oscilloscope, counter and signal generator are coupled via a dedicated attenuator box in a shielded metal housing. Naturally, there was that annoying hissy crackle of a bad connection audible on start-up, which I quickly traced to pin 3 of the 6M5 socket. After sorting that and burn-in testing the chassis, I treated the cabinet with a beeswax furniture polish and put the set back together. It does pay to remove and replace valves and to scrape any oxidation from the pins. They should also be tested before putting them back in. Note that the knobs shown in the photos are from HMV but are not correct for the model. Conclusion This is an attractive radio that cleaned up well. You have to be wary of people having messed with the set previously and introduced faults. Most importantly, if the condition of the set is unknown to you, make sure it’s safe to power up before doing so! SC August 2024  91