Sickness caused by an old PyeAuthor: Steve Harris
When I was about 10 years old, a friend at school had an old radio given to him by his grandmother, rather than throwing it away after the radio shop man had pronounced it not worth repairing. It was a big heavy thing, with a door on the back that opened to reveal the mysterious innards, and a cut out fret of a stylised sunrise on the front.
Over a period of time we systematically pulled it to pieces in the shed, finding out how far the fine wire from the speaker coil would stretch round the garden, whether or not the valves would withstand being dropped out of the bedroom window, and other such educational activities. The case was used as a handy storage box until bonfire night.
Little did I know that the relatives of this radio were going to come back in later life to haunt me for this desecration of their brother, but like the Ancient Mariner, I am now condemned to do penance by repairing them. The man in the local radio shop obviously knew a thing or two when he condemned that one to death by proxy at the hands of small boys.
The Pye MM, for that is the Albatross of which I speak, is one of the Rising Sun series of sets made by Pye of Cambridge in the early 30s. It is probably the more common of the mains sets in this group, either because it sold well, or more likely because they carried on working longer than most of the other models. I suspect that this is the case, because of various factors. Most of the capacitors are big wax paper things in metal cans, which survive with quite bad leakage for a long time without going dead short. The metal rectifier ages, going high resistance but not keeling over entirely like a valve, reducing the HT, but keeping the set working. There is no AF coupling condenser to go leaky and overdrive the output valve, the biggest killer of all, so it is often found that these sets are still working, after a fashion. If you are offered one, and it is described as working, but just gets the local station at low volume ("It's an old radio, what do you expect?"), then it needs a full overhaul. If the wavechange switch "needs a drop of oil", then it needs major surgery. Another thing is that the mains connector, a strange flat pin device on most models, is always missing, so if this was lost in 1957, then it has probably not been used since then. The wavechange switch brings us to the big problem with the MM and the similar battery Q.
Monkey BusinessIf you are familiar with old cars, either the Dinky type or the full-size variety, or gramophone soundboxes, you will have encountered Mazak. Not the music they play in supermarkets, but Magnesium Zinc Aluminium alloy, also called diecast zinc, 'monkey metal' and a lot of worse things. It has a tendency to self destruct, almost as spectacularly as the tape recorders in Mission Impossible, supposedly due to impurities in the original castings. Somebody at Pye must have thought that it was a jolly good idea to make all sorts of bits for their latest model out of this easy to cast, light, easy to machine and probably cheap metal. Well, I suppose they didn't think anybody would be repairing them in 65 years time, even if anybody knew about the inherent problems with the metal.
Some castings remain perfectly OK, (if you can call a metal that is brittle, strips it's threads and has air bubbles in it OK), while others buckle, expand, disintegrate and eventually turn into grey powder. Hopefully the pieces that are still good will remain so, but it is still dreadful stuff. The tuning assembly of the MM, consisting of the variable condenser, of the worst design ever made, the wavechange switch, which is simple, consisting of a mazak shaft which turns in "bearings" in the mazak plates that make up a sort of box, the ends of which hold the tuning condenser shaft in place, with one adjustable pointed centre. The outer plate is another casting, in a bronze finish, as are the mazak knobs for the volume control and wavechange. The potential for big trouble in this area is obvious. The knobs tend to crumble away, the switch shaft siezes in the end plates, which warp, making the condenser shaft loose, when the vanes, which are probably all buckled anyway, short together(one stator has HT on!), or the switch shaft breaks.
If you are lucky, only one of these happens, but any combination is possible. You cannot really find out until you have removed the top chassis, which is another story. The set has two chassis, the lower one has the mains transformer and rectifier, plus a big block of probably dodgy condensers, and the upper chassis has the rest, except for the speaker, and the frame aerial. All these assemblies are connected to each other with various looms and bits of wire, which are probably crumbling rubber insulated cable, all slightly too short to keep the sections apart and still connected. This is more important thasn it seems, because the two chassis are not even at the same earth potential! Added to this, the chassis fixings are all over the place, and before even starting to undo them you have to take off one of the back screw register plates, otherwise the chassis will only come half way out.
Fools rush in...
...Where angels fear to treadWhy didn't I just tackle this chassis first and leave well alone? Why did I buy them at all? The ghostly presence of the long dismembered MM of my youth reaches out to tap on my shoulder. Oh all right. File down broken bit and drill out center of knob. Dismantle both assemblies until bench covered in bits of aluminium, as though an aircraft has crashed. Re assemble with better front plate. Condenser vanes now touch. Bend, lever, prod and mutter until whole band tunable without shorting. Swop volume control (track O/C), reassemble. Works. Replace cover with dozens of screws. Doesn't work. Remove screws. Works. Tightening the screws is distorting the frame, causing vanes to touch. More fiddling. Each time I turn the chassis over, another wire from the snake's wedding joining the units breaks off. Then the frame aerial falls over, and one end of the winding comes off. Eventually it all seems to work, but the volume control still has no effect, and the tuning control only turns the condenser for part of the way. Studying the circuit diagram, which alludes to circuit modifications in a way that suggests that no two sets are the same, I find a problem. The resistor bit of the volume control works by varying the bias on the AC/SG, from zero (w.r.t.cathode), i.e. the top chassis 'earth' to a negative voltage, i.e.the bottom chassis 'earth'. This is applied via the frame aerial to the grid. The wires on this one have been extended by an exasperated engineer in the past, but the earthy end is connected to the top chassis. That can't be right, I say, and try it to the bottom chassis. Now the control works. The tuning drive, well that's just a mechanical problem. Three hours later the whole assembly is in pieces again. The inner alloy plate has warped, making the spindle out of line with the drum. The only answer is to strip it all down and swop the whole plate over from the other chassis. This will mean realigning the tuning condenser all over again. I really cannot go on describing the entire horror, it is probably best to forget the whole thing. Suffice it to say that the next day the set is reassembled and working, and I never want to see inside another one. The other restorable set is for sale as it is, with the remains of the third. If you want something to do on a cold winter's night or two, (or three), then this is the project you are looking for. - My penance is complete.