Fullscreen HTML5Med Res (??MB)HTML4

Vintage Radio Velco Velco 1937 1937 radio radio chassis chassis restoration restoration By Ken Kranz Back in the 1960s, I rescued this 1937 Velco radio chassis from the tip. I’m not sure what radio it came out of; it may have been a kit radio built into a custom cabinet. The cabinet was long gone, but I reckoned that one day, I could get the radio working again. Fast forward to 2019, and I finally had the chance to do just that. This chassis clearly was for a battery-powered radio, as it lacked a mains transformer and rectifier. I wasn’t sure which exact set it was from. Velco made several ‘kit’ radios so it could have been one of those. I searched the internet to get some information about the manufacturer, Velco. Arthur J. Veall Pty Ltd (247249 Swanston Street Melbourne, 302 Chapel Street Prahran) manufactured Velco-branded products from 1931 to 1955. From 1950, Velco Sound Systems Pty Ltd was at 65 Latrobe Street, Melbourne. siliconchip.com.au Velco-branded products included radios, signal generators, “set analysers”, volt-ohm meters, valve testers, multimeters and tape recorders. Velco manufactured a model 365B receiver in 1937. Its specifications were: Valves: 1C4, 1A6, 1C4, 1K6, 1D4 Intermediate Frequency: 175kHz Wave bands: broadcast band only Batteries: 2V (A) and 135V (B) Speaker: permanent magnet Case: timber Valve markings on my set indicate the valve lineup to be 1C4 (RF preamp), 1C6 (converter), 1C4 (first IF amplifier), 1B5 (demodulator & audio preamp) and 1D4 (audio output stage). A 1K6 dual-diode pentode had been fitted in the place of the 1B5, with the pentode triode-connected. This is very similar to, but not exactly the same as what’s specified for the 365B. The 1A6 pentagrid converter has the same pin connections as the 1C6 and very similar specifications. So I think that this radio is a 365B derivative. Circuit details Unfortunately, I could not locate a 83 Fig.1: the Velco 1937 radio circuit was traced from the original set and then modified. One modification was adding a set of diodes to clamp the filament supplies and protect it from any high voltages during repair. circuit diagram, so I had to create one by tracing out the circuit. It is shown in Fig.1. I drew the original version of this diagram using LTspice, so I am also able to simulate the operation of the radio. The component designators are unlikely to match the original schematics, as I had to number them myself. Excellent SPICE models are available for all the common audio valves and many RF valves, although it’s difficult to find models for pentagrid converters. You can download the files for my circuit diagram/model from siliconchip.com.au/Shop/6/5573 The download package also includes many handy valve models that could be used to simulate other sets. Looking at the circuit, there’s nothing really remarkable about this set. It does have an RF gain stage built around valve V1. The following mixer/oscillator is a conventional configuration, as are the IF transformers and sole IF gain stage. The two diodes in V4 are used to demodulate the audio signal and to derive an AGC control voltage, which is used to vary the bias conditions of the first three stages (V1-V3; RF amp, converter and IF amp). The audio output stage (V5) is a very basic Class-A configuration. Fixing it up The first thing I did after I got it on my workbench was to take a good look at it. I found that the tuning gang had a bent shaft, presumably due to the cramped nature of its storage location for the last 50 plus years. I removed the tuning capacitor and placed the shaft in a vice. Quite some force was required to make it almost straight. I feared one extra adjustment would break the shaft, so I stopped there and refitted to the chassis with new rubber mounting grommets, to replace the disintegrated originals. A number of the paper caps tested leaky, so I replaced all of them with Shown here is the underside of the Velco radio chassis before it was repaired, with the finished set shown adjacent for comparison. One of the biggest changes was the replacement of all the paper capacitors with newer film capacitors. 84  Silicon Chip Australia’s electronics magazine siliconchip.com.au modern types. The back-bias resistor was open-circuit; I was able to repair it by removing one turn of its wire and soldering that to its terminal. I also found some badly damaged wires and replaced them with 600Vrated black wires with silicone insulation. I decided to add eight silicon diodes, two sets of four in series in an inverse parallel arrangement across the 2V filament supply. This was to protect the radio against me accidentally connecting too high a voltage to this supply. The radio was initially designed to drive a loudspeaker fitted with an im- pedance matching transformer. So that I could drive a modern 8W speaker instead, I decided to fit a 100V line transformer, connecting the 0.5W tap (20kW impedance) between the anode of V5 (the 1D4 output pentode) and the B+ rail. I then connected my test speaker across the 8W winding. Testing the radio I applied 2V to the filament supply and measured the current drawn. It settled at around 700mA, which I thought was a reasonable figure to power the five valve heaters. I then slowly ramped up the B+ supply to 135V DC and measured a flow of about 10mA. Australia’s electronics magazine Some satisfactory noises were coming from the loudspeaker, so I fitted a short aerial and found that all local radio stations could be received with good fidelity, in spite of Pimpala (ABC 891kHz) transmitting 50kW only 3km from my location. I aligned the dial pointer with the actual transmitted frequencies and left it tuned to 1323kHz for some background music. A few days later, the radio was playing away in the background when the sound of silence hit me. The primary winding on one of the IF transformers had gone open circuit. My friend Andy (VK5AAQ) advised me that this was common on these 85 The coils for the replacement IF transformer was wound using a sewing machine with 0.1mm copper wire (left). The coils were placed on a wooden dowel, which is attached to the original mounting bracket as shown above. sets, as they only switched off the filament supply; the constant B+ voltage combined with moisture caused electrolytic corrosion, with this being a typical result. Rather than look for a replacement 175kHz IF transformer, I recalled that back in the early days of radio, many items were self-made. Inspired by B. B. Babani’s Coil Design and Construction Manual (1960), I decided to repair it with a home-made replacement coil. A replacement IF transformer The outer diameter of the coil former was about 10mm. A sewing machine bobbin is about 9mm; close enough for me. I measured the wire diameter at 0.1mm. So I ordered a reel of 0.1mm diameter enamelled copper wire and some clear sewing machine bobbins. I drilled a 1mm hole in the sidewall of the bobbin for the start of the winding, then placed it on the semi-automatic coil winder of a sewing machine. My daughter held a nylon rod with the spool of 0.1mm wire so it could unspool freely, and the machine ¾ filled the bobbin in no time. The spooling speed is fully adjustable from a crawl to frightening. In spite of the very low strength of the 0.1mm wire, we didn’t experience any breakages. I measured the inductance of the good coil on the old IF transformer at 7.4mH using a TH2821B LCR meter, and used a Fluke multimeter to determine that the DC resistance was 76W. As both trimmer capacitors had similar ranges (19-110pF), I decided to build the IF transformer using two identical 7.4mH inductors. I tested the freshly wound coil and found it to be over four times the required value, so we transferred about 86  Silicon Chip half of the wire onto another bobbin. I adjusted both coils by removing turns until the required 7.4mH was achieved. I found that the DC resistance and Q were very similar to those of the original coils. I cut the wires about 6mm from the coil former and soldered flexible ribbon cable leads onto the coil ends. I then covered the wire terminations in some ten-second ultraviolet cure resin. I then slid them onto a 6mm outer diameter wooden dowel, applied a generous amount of shellac to each end and fitted the assembly to the original mounting bracket. The wires were terminated as shown in the photos. The moment of truth: I powered the set on and turned the volume full up, but it was very quiet. A quick adjustment of the IF trimmer capacitors gave me lots of beautiful noise. I moved to a blank spot on the dial and used the trimmers to peak the noise at about ½ to ¾ compression. All the local stations came in loud and clear, including 5MU. An excellent result indeed. Although the slideable coils would allow adjustment of the coupling coefficient, the result was so pleasing that I immediately coated the former with shellac, including a small amount on the coil. As one would expect, the set stopped working. I did not re-tune the IFs and simply let the set dry out for a few days. At switch-on a few days later, the radio was again playing 1323kHz; most satisfactory. I refitted the IF cover and that hardly affecting the tuning. A quick re-tweak and the set ran for about a week until bench real estate required its movement back into storage. Conclusion All that’s left is to put the original can back in place. The repair of this set may offend some restoration purists. The radio was saved from landfill in the 1960s; it now works and might be enjoyed by somebody in the future. I saved all the components I removed. It could be reinstalled in a period cabinet by someone with the skills and inclination to do so. It still needs a replacement 2V dial lamp (not shown on the circuit diagram). I might make a screw-in replacement using a white LED and resistor. Some time in the future, I am hoping to find a diecast box that will locate over the large square hole, paint it a similar colour, and build a power supply into it so the set will run from a 12V plug pack. The radio consumes less than 3W, so I might even be able to power it from a 1A USB port. Velco references: siliconchip.com.au/link/ab31 siliconchip.com.au/link/ab32 www.kevinchant.com/velco.html SPICE references: siliconchip.com.au/link/ab33 siliconchip.com.au/link/ab34 siliconchip.com.au/link/ab35 Plotting valve curves using LTspice: https://youtu.be/VV3e_mNQ-dQ SC Australia’s electronics magazine siliconchip.com.au