This "Recent Repair" was kindly contributed by Tim Pullin.
A three waveband (MW, LW and VHF) AC superhet from 1957. Valve line-up: 6L12 (ECC85), 6C12, 6F18, EABC80, EM85, 6P15 (EL84) and EZ80. Features magic eye tuning indicator, rotatable ferrite rod aerial for MW and LW, and internal frame aerial for VHF.
One of a pair of sets found via an ad in our local free ads paper, (see Unknown Stella/Cossor/Philips). A nice looking classic late fifties set, in pretty good nick. This particular set was, apparently in use for many years behind the bar of "The Clifton" pub in Bristol. The vendor's mum was the cleaner and acquired it when the set was superseded sometime in the late sixties or early seventies. It was then run in a shed for a while until it was again superseded, forgotten, found and flogged to me!
The picture is from Dave Grant's excellent Murphy site, why not take a look?
I knew this set worked (at least on MW) because the vendor was very keen to demonstrate, and managed to plug it in and switch on before I could sprint across the room and stop him! Like a lot of Murphy sets of this period, that I have found the HT capacitors seemed to form up ok and the set worked quite happily on MW.
Back in the workshop the set also performed well on LW, but not alas on VHF (also par for the course for sets of this period). Far from the VHF band being silent, the set squealed loudly (you couldn't really call it "motor-boating") the pitch being unaffected by the tuning controls. Classic symptoms of decoupling capacitors. But which one?
The chassis was removed with no great difficulty (the concentric knobs just pull off) and lifted clear of the cabinet. There was quite a thick layer of dust, which was brushed and vacuumed away. I noticed the set was fitted with a two-core mains flex which I made a mental note to change to 3-core when the set was fixed.
Voltage measurements around the 6L12 (ECC85) valve showed the HT was a little high, but normal elsewhere in the set. The grid feed capacitor to the 6P15 (EL84) was actually ok, with around 2 to 3v negative indicated on the grid. It looked a little modern for the set, so had probably been changed.
I suspected the ECC85 may be low emission and changed it for a known good spare. D.C voltages were all normal now, giving no clue as to the location of the fault.
I decided to use the oscilloscope (with a x10 probe, almost essential when dealing with valve circuits - not only does it expand the voltage range of the scope, it means that circuits such as oscillators can be monitored without affecting its operation) to try and locate the area where the oscillations were starting.
However the moment I touched the earth lead of the probe onto the chassis of the set, it drew a trail of sparks! There was about 25v A.C between mains earth and the set's chassis, so it appeared there was a fault was at the neutral end (luckily for me) of the mains transformer primary winding.
It just goes to show you can never be too careful! The importance of the pre-plugin checks cannot be over emphasised as this example shows. I will now carry them out on EVERY SET I WORK ON even if it seems to be working perfectly. I have lived to tell the tale, all because the short to chassis was at the neutral end, and I don't want to push my luck. The set was probably operating for years like this, and may explain why it was fitted with a two-core cable! The mains transformer on this set doesn't look quite right. It has quite a modern appearance, with a clear plastic wrapping around the coils. (If any one has one of these sets I would be interested to know what yours looks like.)
I wasn't able to use the scope so decided upon an alternative method. A very handy gadget in my workshop is a 0.1uF 1000v capacitor with about 4" of wire soldered onto each end, which terminate in crocodile clips. This was going to be used to provide an AC short (to any signals) on the grids of each stage.
The capacitor was connected across the grid of V1a (the RF amp half of the ECC85) but had no effect on the howling. Shorting the grid of the oscillator half of the ECC85 had no effect either, suggesting the oscillations were being generated further "forward" in the set.
Upon shorting g1 of V2(the hexode section of the 6C12) with the capacitor, the oscillations disappeared. This was the faulty stage so time to narrow the search. This time the capacitor was connected between the anode of V2 and chassis, and the set sprang into life (well Radio 3!) proving the de-coupler C23 open circuit. It's a brown plastic type with its value of 0.003uF marked in paint spots. Replacement was a little awkward, as the capacitor only had short leads and was mounted under a fair bit of wiring. One lead of the faulty component was cut as close to the chassis as possible, the new capacitors lead soldered to the stub of the old, and used to pull the replacement into position, which saved a lot of fiddling around!
The set now works perfectly, and it's a real shame about the mains transformer. I was hoping to sell this set when It was finished, but as the transformer is faulty I will have to wait until a new one comes along. It's uncertain if the condition of the transformer will deteriorate, so it's best to err on the side of caution.
Text Copyright © 2001 Tim Pullin