This old KB's very compact considering the generous proportions of its bits. It's survived quite well over the years although I understand it was retrieved from the back room of a now defunct cycle shop where I'd swear it's gathered close to its own weight in dust and fluff and dead spiders.
It really is incredibly dirty. I set to removing the back and soon the air (and the workbench and the floor) is thick with dust. Thankfully only a couple of stray woodworm had a nibble at it (presumably there was so much dirt they got indigestion). Well, the first surprise is that the mains tap's set to 205. I wonder when Bournemouth changed over to AC? Did anyone ever legislate on those two-pin 5 Amp plugs, one of which still terminates the rather tatty mains lead. There must be some houses still wired for those and three pin 5 Amp sockets.... I remember designing this air defence system for South Africa and they wanted old fashioned 15 Amp sockets (do they have plastic bits in their plug pins now or are they still unenlightened?). Sorry- back to the job in hand- unfortunately I note that the valves have been got at. Top cap clips, bar one, are all off, caps and glass are wobbly and the valves are only just poked into their sockets. Out with them one by one for a wash and brush up and a spot of superglue to make the bakelite and top caps secure. The one with the clip still in place looks awfully wobbly so that gets a dab on the top with a hot iron as the glue's setting. Phew!, where does it warn you that superglue shouldn't be heated up with a hot soldering iron, I think to myself, as I nearly keel over with the effects of what smells like phosgene.
The resistor markings don't really equate too well with their measured values but maybe it won't matter that much, judging from the advice given in contemporary articles on set design. I wonder how much the audio coupling capacitor leaks? I unsolder it noting that it was only half-heartedly soldered, then noticing it was new (well 50's or early 60's- someone's been here before- maybe the same guy that poked a screwdriver through the speaker cone?). 50 Kohm- replacing that will certainly cool down the old CBL31. In goes a nice new polysomething or other. "Why do these coupling condensers always go leaky", I absent mindedly think? Perhaps its some sort of crystalline growth akin to the Mullard "AF" transistor problem -someone ought to do a thesis on it.
Why are the dial lamps 6.5 volt 0.3
Amp? That doesn't look right. Wait till later, but I'll put small
bulbs in because the large ones are rubbing on the back of the
dial. Sure enough when the set's switched on the light from the
bulbs wouldn't get through the celluloid.
The repaired 9D2 looks a bit dim. The filament voltage reads 2 which can't be right. That tell-tale white powder inside says its gone soft. It may have won its battle to keep its clip mated with its top cap but it died in the attempt. I retrieve a new 9D2 from the junk box. I recall that this one came either from a wartime aluminium creation, stuffed with them (of which only the vaguest memories remain) or perhaps it maybe from that little two valve receiver that used to be fitted in the Dakota? No-one will ever know.
The new 9D2 lights up and now a few stations appear. A twiddle on the TF2008 and a response from the speaker tells me the IF's somewhere between 150 and 200Kc/s or so. I see the Light Program had settled on 1500 metres- that must have happened just before the dial maker sorted out his design.
Strange to see the crude wavechange switch. I imagined yaxley switches had been around for ever.... I wonder why they went to the expense of fitting a three gang tuning condenser? I suppose bandpass sections dealt with the problems of the low IF and the difficulties of dx reception when regional transmitter sites were close to the main large population areas. Perhaps they had to use up a lot of condensers left over from the old TRF sets? Why did they use such a low IF? There's nothing wrong with the IF amplifier-isn't the 9D2 really a 6K7 in disguise? I suppose because they had lots of coils left over from really early superhets! I digress again. A 15K resistor, that really was bad, feeding the screen of the IF amplifier comes out. In the process one of its end caps came off clean as a whistle revealing a perfect carbon end. It now measures 15K but I can't solder to carbon so an old wartime 15K that really did measure 15K goes in. I think that one came from an "RDF-1", does anyone remember that one- it could get ITV!
This set's back proclaims its got "rejectostats". They must be the bits of twisted wire stuck to the valve holders. Must look into that sometime- for some reason, unlike transistors, the invention never caught on (see below). Talking about the back- there's a half-clever idea, the back has these two chunks of brass riveted to it which engage with beefy sockets in the rear of the chassis. The idea is that you can't operate the set with it's back off. The brass bits complete the mains circuit. Come to think of it, the control knobs have grub screws, and being an AC/DC set, wet fingers could easily earth the live chassis via one's vital bits and the nails in one's boots. Oh! I remember, wasn't there something about turning the mains plug round if that happened?
Well it makes for an unusual job, servicing the set with it's back on out of the cabinet! The lowest bit at the front is the bottom of the plastic dial (groaning under the weight of the chassis)-no thought for the serviceman. I'll stick a screwdriver handle under the front of the chassis. Can't really make sense of the IF frequency. It seems wrong somehow. Fire up the computer and load the patent IF mystery program. Scrutinise the innards of the RF stages and put details into program together with the salient dial markings and fiddle around with the keyboard to optimise tracking....and 125 Kc/s is the answer. The best I can do with most of the IF trimmers is about 160 Kc/s. Try as I might without stripping the threads in the king size mica compression trimmers they just won't come down to anything like 125 Kc/s. Try padding a bit? Well when I add 47pF across all four trimmers and set the TF2008 to 125 Kc/s the trimmers tune nicely to the calculated frequency. Up goes the gain and the dial reads something like the wavelength of the stations appearing.
Now to start on fine tuning the oscillator tracking. No-one's ever fiddled with these trimmers because the back's in the way. To get at these trimmers one has to remove the back and stick a couple of thick washers in the holes where the safety device goes. This done I find these trimmers have also gone low. Another 47 pF does the trick and after a twiddle, up goes the gain again and now stations are exactly where they are supposed to be. At this point there's a fizzling noise and a grubby wax condenser connecting the audio output valve anode to deck starts exuding its vital fluids. Nip it out and the fizzling stops. Now try some fresh condensers and some resistors nearer the original marked values in the AVC circuits. It's always the 47K and 470K that seem to go off most. The set's now getting quite lively. Change some of the odd decoupling condensers in the screens and cathodes and now the performance is getting really good. I can get Droitwich quite well without an aerial.
Time to put aside the chassis and take
a closer look at the cabinet. Someone has caught the veneer on
the front and pulled off three little pieces. It spoils the look.
For that matter so does the rather scruffy varnish. On balance
I guess it should be stripped off- at least then I can make a
decent job of the patching to the veneer. Nitromoors gets the
old finish off and underneath the front panel is revealed as
walnut and the top and sides, nicely figured oak. Why did they
slap that dark varnish all over the cabinet and hide the walnut?
- perhaps walnut used to be cheap and dark brown varnish the
in-thing? I carefully clean up the edges of the broken veneer
and glue some patches in place. Lots of applications of French
polish stain, linseed oil and rubbing with wet and dry brings
out the nice patterns in the wood and returns the cabinet to
basically the right colour. What can be done with the speaker
cloth? What finished as a faded brown started as a sort of turquoise.
Under the protecting woodwork away from the ravages of ultra
violet bleaching of daylight the cloth is pristine. The exposed
bit is a faded brown. Dig out a small canister of dye, soak the
cloth for a bit then mix in some fixer and hey presto the whole
piece is back to a glowing turquoise. A bit of sticky glue applied
to the inside of the cabinet and the cloth goes back just as
it was in 1934.
Screw in the speaker, slide in the chassis,
put the cabinet back on its small wooden plinths and fix some
new feet into place and its all as good as new.
Plug it in and switch on. A faint hum heralds the roar of the stations. Hang on why are all the stations roaring. Peer over towards the computer. No that's off. Turn down the volume and the roaring goes away. Turn it up and the turbocharger starts up again. Quick think. I remember snippoing off that .002uF from the anode of the CBL31 but I never did get round to replacing it. Hopefully the stray bit of RF in the output will go away when a new one's in place? Cross fingers.
Post script...I found out much later that the IF coils are stagger-tuned to broaden the bandwidth, improve the audio and prevent instability!
vertical whip antenna for chimney mounting, with a balun in the base and
a length of twin-screened feeder. Rejectostat receivers had a 3-socket
antenna/earth panel (A1, E, A2), with a balanced antenna coupling coil
on the first tuned inductor.
downlead of an external antenna, and using the screened balanced feeder