Ekco UAW78 LW, MW & SW radio 1937

Ekco (Southend-on-Sea) UAW78 (Sir Misha Black design) radio 1937. My rarest find yet. Spent the last 2 weeks restoring this Bakelite beauty, cabinet was mint but the inside was a horror movie. All that's behind me now, I just need to find a space for it.
It had been neglected for many decades, stored in a cardboard box at the home of a garage owner in Corley Ash, near Coventry. It had belonged to his late Father and I said I would update him when it was repaired. The level of rust inside was exceeding anything I've attempted to repair before. But not deterred I divided the restoration in chapters like a book and this helps in understanding how it was eventually brought back to life again.
1937 colour shop flyer for the range of receivers on offer.
The  radio's designer Misha Black here in 1935 showing off his battery powered bicycle, he also was responsible for UK roads signs and British Rail diesel trains throughout the 1970's. 

The repair

Chapter 1: getting the rusty chassis out of the cabinet
It's rare to find a 1930s radio with it's back still in one piece. From the 30's radios and televisions had cardboard backs (not very fireproof) but saved on the weight of the set. I must say Ekco put quit a bit of effort into weight reduction on this model. Heavy mains transformer was replaced with a newly introduced series of valves designed to have their heaters wired in series and this required a standard 200 mA current draw from the heaters, most valves here have different heater voltage requirement, but when all in series with each other their different resistance can be achieved if they have the same current requirements, a large mains dropper with voltage taps for 210, 230, & 250 volt mains was also added to the chain. They also made the chassis with three sided thin steel instead of thick aluminium 5 sided norm until now. Steel though will rust quickly. The battery version of this set has a completely different circuit and valve selection for lower current valves.
All my cabinet screws were rusted that bad I had to hacksaw them off, re-drill holes and use new bolts. I used a slitly smaller drill bit and forced a new thread for the bolt in the chassis frame.The control knobs were the same, I tried soaking threads in WD40 overnight, but no good. To remove the knobs I used brute force with two screw drivers as levers, taking care not the crack the front or the knobs.
Forced knob removed -  due to rusty grub screws
Once the knobs were removed I heated the grub screws with soldering iron temperature of 350' and this was enough to release the totally rusted threads.
View looking underneath the chassis from the back, opposite is the view after removal, lol Soon cleaned up with a car vacuum and paint brush. Use a brush with a soft gentle action and take photos of the chassis after
Rust was a feature in this cob web


Chapter 2: Grunge removal
The next job was to remove all the grunge from within, trust me nobody will ever see this, but it's so rewarding. Any future repairs will require access to the inside and confidence to repair the set comes from the look within. Use a car vacuum and paint brush to get into all those hidden places. Next wire brush and wire wool the rust to a smooth finish, you'll need to remove all the valves, so take a photo of the layout. Then coat the rusty areas with rush inhibitor paint, this will turn the metal black in colour, but it will look more tidy and last a few more years. Sadly though the underneath can't be treated, take care not the paint the soldered areas. when you clean the valves don't rub the printed areas because that will rub off. While the valves are out use a fine modelling drill attachment to clean the inide of the valve pins, wire wool them on the valves if needed.
Restored inside

Chapter 3: Loudspeaker repair
The loudspeaker was beyond repair and I replaced with this 8 inch oval one I had lying around. The original was a 6 inch circular type and was also a new lightweight design, with a small magnet instead of a powered magnet coil in previous models. Don't forget to clean the inside of the cabinet too, there was some woodworm hols in the speaker baffle, but I think they are long gone. The speaker bolts needed to be sawed off too. While removed I found the rear of the speaker cloth was mint, showing the gold braids in the cloth, so i carefully removed it and applied glue to the outer edges to stop it from fraying more and re-fitted. Cloth on line will cost around £20 a yard for the period stuff. The secondary winding on valve transformers is around 1/2 to 1 ohm resistance, so most impedance speaker will work, the lover the resistance the louder the volume, only by a small percentage though.
While the chassis was removed for repair, I fitted a small temporary speaker th the chassis with a bolt, this aloughs free movement while fixing the set.

Chapter 4: Mains dropper restoration
I't still a long, long way from powering up this old gentleman, next I tackle the mains dropper resistor. All the copper straps with the voltage tags had all broken, so I used 4 wraps of 18 swg copper wire, tightly wrapped around the porcelain tube, these also act a solder points for the voltage selector. Luckily all winding's had continuity so I carefully re-wound the lose sections after repairing the central support bolt that (you guessed it) had rusted solid. Where I broke the bottom clip away I soldered a nut to hold the treated bolt back in place. Take care to connect the wires right, red input was at the base and the output from the top feeds through to the mains fuse found on the smoother LT choke. Take care to scrape the insulation of the resistance wire before soldering them to the terminals I do it with a small modelling knife, scrape all around, until shiny. The resistor values are 532 ohm (large winding) 104 each for 2 small winding's giving 740 ohms total for 240 v AC mains. (valves resistances add up to 107 ohms.) Non of this info is in the manual.
Chapter 5: the power supply.
It's a good idea to start repairing a radio from the right hand side of the circuit, one valve at a time. For a job a bad as this you'll need a signal generator and an oscilloscope.

I could not find a circuit diagram for free on-line, so paid £1.99 to a site called www.servise-data.com, well worth the money for a full high definition manual. V5 is the power supply rectifier valve UR1c 1/2 wave rectifier to be exact. It takes 70 seconds to warm-up giving the rest of the valves in the set time to heat-up before applying HT to them, thus avoiding all the distortions that you would haer as it warmed-up. It too a while to discover but when this valve warmed up it was arcing inside causing the breakdown of many 0.1uF 'c's inside the set, but more annoying it would blow the 6,2 volt dial lamp every time I turned it on. A 500 ohm 10 watt resistor it fitted across the bulb, but even this had gone open circuit, and when replaced still blew bulb when valve flashed inside.The answer was to fit a BY127 or similar silicon diode across the anode and cathode terminals (later I fitted a solder strip) The HT will be around 280v now instead of 185 volts, this is because the silicon diode has less forward resistance than a valve. I fitted a 500 ohm 10 watt in the HT rail and a 5k  1/2 watt between V1 & V2's HT feeds, this returns the voltages to circuit manual specifications. There is also a company fitted 100 ohm wire wound 10 watt resistor in series with HT (this is fitted on the HT choke panel.)
Before restore! showing original speaker
Paper smoother block
It's safe to say all smoother components will be suspect, so I replaced most with 44uF 400 volts electrolytic's found on line. The originals were paper types,  two found in a large white brick type bolted to the chassis, this I removed for extra space. If you keep the valve running use similar types (8, 24 uF) either side of the smoother transformer (Choke).
Volume knob before and after restoration, 400 degree heat to grub screw loosened it

Chapter 6: The sound output stages V4 (Pen 36c)
A few of the 0.1uf capacitors were dragging the HT rail down, and removing bias from the sound valve are located around this valve. After this I replaced ALL the 0.1 uF 'c's' in the set. four are located in 2 metal green boxes? this is for screening I found when swapping them (hums appeared). Bit tricky because they pot riveted to the chassis. R20 (250k) is the grid bias for the output valve I replaced that and now have sound  from that point, tested with a signal generator. No sound from the volume control though, so it looks like trouble with V3 stage, it's a double diode (used to rectify RF from last IF stage and or AGC) and triode pre-amp for the output valve. It took me quite a while to realise that somebody in the past had swapped this valve around with V1 (always a good idea to check all the valves are the right ones, lol) When the valves were wrongly fitted the screening can had crushed the feed wire to the top cap that contained a hidden 250k resistor, this was in two halves. With this done I now have signal from the volume control, yippee! 1/2 way there, but still a way to go yet.
I fitted some white paper inside the light box to increase brightness

The extent of the rust damage on the thread of the tuning knob!! - heat applied to bolt with ann iron helped freed the jammed screw

Chapter 7: RF stages V1 & V2 (TH21c & ARP3)
After injecting 460 Khz signal to the 2nd stage grid (top cap) I re-aligned IF coils for maximum signal, same with 1st IF stage, I could at last get Radio 4 on the long wave band  (cheers of joy to know that the IF transformers are working!! - rewinding them is not much fun)
The input switching and coil arrangements for this set are a complete nightmare, although the manual gives good layouts and some help like numbering the switch contacts and then not numbering it on the circuit layout??? Similar trouble finding where coils are located to check them, in the end I gave up and fed a signal from the aerial socket to the grid of the first stage and got medium wave to work too. I would have liked to improve sensitivity but I think the elements had attacked the input coils, so with a long aerial wire I have at least 2 of the 3 bands working.
Band switch (bit of a weakness)
A view here of the band switch, near is the oscillator switch, far is the aerial coils switch. Really this needed to be removed and the contacts re-tension-ed, but the amount of wiring held me back, perhaps another day I will attempt to do it.

A, shows temporary speaker attacked to side while repair taking place, B, shows HT smoother choke with 100 ohm and 500 mA fuse on top, C, shows dial light box, it has 3 tiny slits of various sizes to illuminate station names on the three ranges. Photo on right shows the cob web & rust free inside with the black rust coating. The gold RF valve (V1) was wrongly installed in the screened can at rear in the audio stage. I removed and rust treated the mains dropper screen can, held in place by 3 nuts underside
Power supply valve V5  base, valve was faulty so I re-directed the anode & cathode wires to a terminal block with a BY127 diode 47uF smoother and a 500 ohm 10 watt HT reducer. Right side coil is a Droitwich filter coil, stops over spill on MW bands.

No comments:

Post a Comment