This "Recent Repair" was kindly contributed by Ranulph Poole.
He sent it in the form of his restoration notes plus an accompanying explanation. I have left it in the same form here.
To help with following the description,
the circuit diagram is provided here.
Circuit Diagram Part 1 Circuit Diagram Part 2
Some restorers may feel that the modifications carried out by Ranulph to improve the sound quality of the this set are a little excessive. However the modifications could be reversed at a later date if required, and Ranulph clearly wants a set that is pleasant to listen to with today's broadcasts. The subject of modifications is discussed on the Modifications page in the Repair and Restoration Information section of this website.
This set, released in August 1955, is a first generation AM/FM table receiver. It was bought for £5 at the local junk shop. Although dirty and slightly scratched, it seemed in basically sound condition; there was little evidence of previous ‘repair’ work, except that the double-triode of the FM front-end was missing. Fortunately, the wiring was all PVC insulated and would not need replacement.
The case is of simple construction: the sides, top and bottom are made of rectangular veneered 3/8" plywood; the front is a cloth-covered loudspeaker baffle with a full-width glass tuning scale beneath. Some beading covers the join between the sides and the baffle/tuning scale. The back and inspection cover underneath are of perforated cardboard. A 6 ½" loudspeaker is fitted to the baffle.
The electronic design is very standard for sets of this era; it includes a permeability tuned ECC85 FM front-end and magic eye tuning indicator. The audio amplifier has anode-to-anode feedback, designed to provide additional bass boost on FM. The wavechange switch and ECH81 circuitry are built on to a subassembly, alas making access to the associated components difficult. Generally, the standard of construction is cost-effective but of reasonable quality.
Repair Work Carried Out
Modification Work Carried Out
The AM and FM distortion was checked by applying a modulated signal to the aerial input and measuring the output on the loudspeaker terminals at a level of 1V RMS. The FM distortion was about 0.7% at 100 kHz peak deviation, and the AM distortion 1% at 95% modulation depth.
2W of audio power were available at the onset of clipping.
The AM bandwidth was about +/- 9 kHz at -3 dB. (This is rather wide for today’s circumstances: broadcasters restrict the modulation bandwidth to 5 kHz.)
The FM sensitivity was about 100 mV to achieve limiting. The audio output remained usable well below this level.
Unfortunately, the output transformer didn't survive initial testing. When I took it apart, there were about 3 breaks and some evidence of corrosion. I had a lot of fun winding 3,500 turns of 40 swg wire in neat layers! Fortunately, the insulation of modern wire is much better, and you don't need to worry too much if adjacent layers touch. I used PTFE plumber's tape between the layers - it seems to have ideal electrical and physical characteristics. In another guise, I repair antique clocks: a mainspring winder - really just a handle and spindle mounted in a frame - is useful for coil-winding. It would be easy for the amateur wireless restorer to make a similar arrangement.
I didn't measure the distortion introduced by the FM discriminator, but it must have been greater than 10% at full deviation. Increasing the coupling as mentioned in the summary reduced it to well below 1%, and had the added advantage of making the tuning of the set much less critical.
The manufacturers had forgotten to add de-emphasise, and the sound was bright, to say the least. In an attempt to compensate, they had added a rather nasty bass boost circuit, which started to work at about 1 kHz. With the correct de-emphasise, the bass boost can be removed, or arranged to work at a much lower frequency where its effect is more pleasant. The tone control was another horror. It only had an effect - an excessive one - during the first few degrees of rotation. With an external low-impedance input (through the Gram sockets) it had no effect at all. (A quick look at the circuit diagram shows why.) It can be made usable very simply as detailed in the repair summary.
When aligning the FM front-end, it is not a good idea to adjust the beehive capacitors with your fingers: they have the full HT on them! The first IF transformer (part of the front-end) is best aligned by loosely coupling the IF signal into the grid of the FM mixer. The alignment will need to be checked again if the neutralising capacitor C11 is adjusted.
The AM detector introduced severe distortion above 65% mod depth. This was the result of the additional AC loading imposed by the volume control and the AGC. (With AM-only sets, the AGC was usually derived from a second diode, and AC loading was less of a problem.) A small amount of forward bias (via a 10 M resistor from the HT supply) helps greatly. Ideally, the bias should be proportional to signal level, and it is possible to arrange this by returning the 10 M resistor to the screen grid of the IF stage. At low signal levels, the screen voltage is also low. As the AGC takes effect, and the current through the valve falls, the voltage rises towards the supply rail. The AM IF bandwidth of +/-9 kHz is rather too wide by today's standards, as the broadcasts are restricted to +/-5 kHz. However, I didn't attempt to alter this.
The service sheet recommends aligning the AM front-end by applying a signal to the aerial input through a 200 pF capacitor. This value is much too large, and heavily loads the RF tuned circuits. It is much better to plug in the aerial that the set would normally use, and loosely couple the output of the signal generator into this.
By far the most troublesome problem was leakage across the wavechange switch. Somehow it had become coated with a conductive film, which produced a variety of unpleasant and obscure symptoms. Cleaning the surfaces of the switch wafers adequately was difficult (to say the least) because of the limited accessibility.
Anyway, the set now works pretty well. If nothing else, the effort was educational!
I received the following information about the repair of another of these sets from Richard Newman
I read with great interest one of your 'recent repairs' regarding a Baird 301. I have had one of these for some time and have just got round to working on it.
Like the report, I too changed all the Hunts capacitors and fitted correct values where incorrect substitutes had been made. Before all this the set worked but sounded rather 'wooly'. I found that a lot of the Hunts caps had changed in value. One in particular in the de-emphasise had gone from .01uF up to around 0.4uF so that would account for the wooly sound on FM. Also for some strange reason, the tone correction cap had been changed from .002uF to 18000pF.
Once these had been changed, the FM sound was superb. I don't consider it to be too 'bright' as in the report. I haven't carried out any of the modifications suggested by your contributor as I don't consider them necessary.... plus it detracts from the original design.
The only thing I would agree with is the tone control which is just as described. However I have left it as is because it doesn't bother me. Perhaps the speaker I am using is less 'toppy' than the one used in the report. Whatever the reason, I am very happy with the sound of my own Baird 301 I just thought you might be interested in my own findings.
Text and Diagrams Copyright © 2001 Ranulph Poole