Introduction | First Impressions | Work | Testing and Using | Circuit | RF Stage | Local Oscillator | IF Amplifiers and AGC | PSU | Conclusions

Introduction

This radio was one of several communications receivers that fortuitously arrived at the British Vintage Wireless & Television Museum just as the previous policy of the late founder not to have such things was reversed, due to my influence. There is now a Comms Corner as I named it, with a number of important radios of Military, Commercial or Amateur Radio type.

I, and probably you, have known of this set from youth. They were advertised in many magazines ex-navy in the sixties and seventies, suggesting they did not do long service with the navy, as it is a post-war set. They have a very distinctive look, Fig. 1, with the bow-fronted tuning scale. That is of an interesting design. The bands slope, which is done so that the scale can be longer than one full rotation of the drum, thus allowing more frequency increments to be marked, giving higher resolution. Each scale is illuminated separately, showing only the one selected at the time. The crystal was dated 1952, the original introduction date is not known. The manual has all pages dated August 1956, and clearly is a re-issue therefore.

Fig. 1: Front view of the set, showing the curious tuning scale.

First Impressions

Examining the set for the first time − I had not encountered one ‘in the metal’ before − it is rather confusing. The reason is that this (D-version) was designed mainly for RTTY use. There is CW, and even AM, but it is not that obvious how to get AM out of it. I was not familiar with RTTY, so the mode switch was not understood. A later more careful examination of the mode switch revealed all. Note that AM is called R/T, Radio Telephony (as opposed to W/T, Wireless Telegraphy, though on here the more modern term CW is used). RTTY, Radio Teletype, is sent by FSK, Frequency Shift Keying. To make it more confusing there are two options for the shift frequencies. For those unfamiliar with RTTY, it is sent by sending two tones, and to make things more complicated instead of a standard for the tones, there are at least two. Well, why make things simpler when with a little thought you can make them more complicated! Things became much clearer on obtaining the loan of the manual, Navy BR1617, titled for B40D, but actually covering all five radios, both operation and service, including even how to rewind the coils − an unusually comprehensive manual.

There are five versions: starting with plain B40, then A, B, C, D, and they look rather different. The number of controls increased with each version and, additionally, the B, C and D versions had a significant cosmetic change with the addition of the black background panels under the controls either side of the tuning scales. They look much more interesting like that.
The way this article is written has had to allow for the fact that the circuit diagram would have taken up four sides of A3, so I have referred to it but it has to be found on the web on the excellent website of The Vintage Military & Amateur Radio Society.

This site has a very large number of manuals of Services equipment, including test gear and also some radios not used by the Services. It is a free service. There is no need to join. For those who do not want to look it up, I hope you will be able to get the feel of the set, anyway. The full manual there is 489 pages! To put that in perspective, the manual for a certain American power amplifier I repaired recently is also 489 pages. I found the circuits at last on page 522. No, I have not got that wrong!

Although not on the cover, there is brief mention of Receiver 62B. This set covers Long and Medium waves, and is designed for use with ‘Sound Reproduction Equipment’, in other words as an entertainment set when not needed for military duties.

Before I go further, a note on valve types. All those used are specified using their CV numbers. CV numbers are the Armed Forces’ Common Valve List types. For those not in the know, the three arms of the Services decided they would renumber commercial valve designations with their own individual numbers, and they tried to separate receiving and transmitting valves. As if the situation wasn’t confused enough by the makers’ having their own systems. Then they decided that they would introduce a joint-services listing system. Note that a valve may be listed under more than one CV number, but when cross-referenced, A may be equivalent to B, but don’t assume B is listed as equivalent to A. Some ‘valves’ are solid-state devices. Next came the NATO numbers………….. I have used mainly the book CV Register of Electronic Valves (including semiconductor devices) 1963, a book marked ‘Restricted’ but which appears at sales from time to time. There are also commercial equivalents books such as the Pinnacle valve book. In this article CV numbers are translated to their common ones.

Work

My initial test found the set dead. This was due to the mains switch sticking on one pole. A drop of oil cured that. Connection was a problem as it did not come with a mains-lead, and we didn’t have one so in the expectation of changing the connector, I soldered the mains wire on. Yes, not very nice, but it got it going. The set kind of worked. Lots of noises and all the signs of not being well stored as is so often the case, with intermittent and noisy controls, but there was definitely life.

As said above, the first thing was to get the mains switch to connect, though strictly, getting the cover off came first. A nasty flimsy piece of metal, which fought back, and replacing was worse! You then see there is a hell of a lot in here − three chassis, with the PSU and audio on the lower, plug-in diecast one. The later models use a very nice potted mains transformer while the earlier ones had an open-frame type, which failed. Also, there are two miniature full-wave rectifiers on here. They seemed determined not to use octal valves, and had found the original rectifier was unreliable, so they used two, each on one phase only. I assume the set pre-dated the GZ34, which would seem the obvious choice for long-term reliability. The dislike of electrolytics, which became fashionable, is shown here with only paper types for HT smoothing.

So, with a drop of oil on the open-frame mains switch we had life. To remove the PSU assembly, you need to unplug some Jones plugs to the main chassis, which were not located for the convenience of the serviceman. That said, the original sets were worse. The manual refers to them having been moved to make servicing more convenient. The mains connector was replaced with an IEC. This was time-consuming because dismantling was needed to get access. The mains filter caps were disconnected as they were old. They may have been perfectly fine, but I didn’t want to risk it. I tested the set to 1800V DC with my home-made PAT-tester; there was no leakage. I didn’t risk going all the way to 3kV, though it would probably have been fine. I replaced the 0.001µF coupling to the output valve grid – it could be relied on  to be at least a bit leaky by now. The value was very small. This set was not intended to have a bass response, so I fitted a 0.022µF, and also R306 (470kΩ) as it had drifted a long way up. It is essential to protect both valve and PSU so in old radios the output valve grid cap is at least checked, and also the grid leak is not left if it has gone high, and this applies to military and domestic sets alike. For the AF voltage amplifier pentode V301, I replaced the screen capacitor and resistors for the reasons given above. When I tested the power/AF section on its own, I found all volts were 336. Referring to the circuit, I realised that the earth connections go through the multiway connectors, and therefore it was ‘floating’. Once reinstalled, it was fine. I did some random tests of resistors on the main chassis: and as a result I changed a couple more capacitors and resistors. Very good for the age.

Testing and Using

To use this set you need to work out the connections, not as obvious as you may think. The connectors are labelled with numbers, not function. The aerial is expected to be balanced, so I modified it to be unbalanced (by the simple expedient of earthing one side) and changed the connector used, to one that wouldn’t be needed, but which had a plug in it. These are a screw-on design of coaxial connector. Next the speaker output was not convenient, so I used a jack socket mounted on an adaptor plate. To actually drive the external speaker, it must be turned on, by the switch above the mains switch. 

I tested some resistors. Apart from those mentioned and changed, they were found to be generally not far off value, and did not need changing. Checking leakage of some capacitors other than those previously mentioned, again I found them to be good. The same alas was not true of the internal speaker for while it was working for the initial tests, it went open circuit. A combination of damp storage and provocation by some nasty transient noises when the set was first tested. Luckily, I had one that fitted.

The AF gain was very scratchy, and sealed, so I drilled a hole in order to use Deoxit on it, which worked well. The bandwidth switch was very intermittent, again cured with Deoxit. I also ‘washed its face’, which made the set look rather better, and replaced two dial lamps one which had failed, and one was dull. It was a 12V bulb. Note there is a brightness control for these on the set’s rear. The crystal didn’t work. On removal, which was not easy, it rattled, but on opening (this is the old-type, which was screwed together not sealed), it was found that just a bit of SRBP (synthetic resin bonded paper) was loose. The pins were cleaned, as was the socket, which was very loose.

As is my habit, I measured signal-to-noise ratios (SNR) with the set switched to AM. Measured using a Marconi TF2019 phase-locked generator, 30% modulated, 1µV PD 50Ω output, with the set adjusted for 50mW into a Marconi TF340 power meter. The results are shown in Table 1.

Table 1: SNR Measurements.

Frequency	SNR		Frequency	SNR
650 Kc/s	>17 dB		1.6 Mc/s	16 dB*
1.6 Mc/s	17 dB*		4 Mc/s	>17 dB
4 Mc/s	17 dB		9.5 Mc/s	>17 db
9.6 Mc/s	17.5 dB		18 Mc/s	17.5 dB
18 Mc/s	16.5 dB		30 Mc/s	>17 dB

 

The frequencies are near the top and bottom of the range being tested. The figures all look satisfactory, and the work required to get this set going was not great, especially considering it was badly stored. It appears people think because a set is military, it is also water and damp proof and can be kept in an out-house. Not so unless it’s a field set.

Circuit

A very big feature of this set, and part of the reason it is so big, is the use of a turret-tuner, Fig. 2. You may be familiar with these, in a very small version, used in the old band 1 and 3 TV tuners, where ‘biscuits’ were mounted on a drum that rotated when you selected the channel. The ‘biscuit’ contained the pre-tuned RF and oscillator circuits. They had studs to connect it, which ran against spring contacts that were fixed. The colour of the plastic used was why they were called biscuits, I assume.

Fig. 2: The turret tuner.

This construction has several advantages. The connections are very short, the separate bands can be in cans to keep them screened from one another, preventing interaction, and the assembly can be made very rigid. It is however more expensive and larger. At VHF and above, you may not have a choice. Another set to use this is the Eddystone 770, where they even integrate the valve bases into the tuning capacitor to get shorter connections, and silver-plate the whole assembly.

This set is not of predictable architecture. While it has single conversion, a typical IF frequency of 500kc/s and two RF amplifiers, a triode-heptode mixer with the triode unused, and three IF stages, there are some surprises. For example, there are two volume controls, one marked gain, which is mainly the RF/IF control, but has an AF potentiometer on the back (see RV 305 & 309), the other being RV224, the conventional AF gain feeding the grid of V301, CV303 (6BA6/EF93). Note that this is a variable-mu pentode, normally associated with RF/IF amplifiers under AGC control, and is here under AGC control. A strange setup. The pure AF gain is the small knob on the left side. Further, RV309 controls attenuation after the output stage, on the secondary of the output transformer, and there are two of those, the ‘usual’ one being in the anode of the output valve, converting to line impedance, feeding TR303, which steps down to low impedance for the monitor speaker. 

RF Stage

The RF stage is also odd. While the second RF amplifier is AGC controlled, the first is a ‘straight’ pentode, CV4014 (CV6064, itself equal to our familiar EF91, 6AM6, 8D3, and Z77). It, however, was subject to overloading so an unusual control was added, labelled anti-cross modulation, RV101. This is basically a manual bias control, which allows the high cathode volts developed across R104 (8.2kΩ) to be opposed by positive bias. Yes, a ‘straight’ valve would have more linear characteristics, higher gain and lower noise than a variable-mu type, but only, you may think, marginally, and adding a control easily misadjusted, leading to an apparently ‘deaf’ set. The RF tuned circuit is configured for common-aerial operation, so needs to be high-impedance at frequencies other than that the set is tuned to, to allow other sets to work on other frequencies. 

Local Oscillator

The local oscillator, V104, CV4014 (EF91), is a pentode. Often these sets use a triode, or triode-connected pentode, but in this case it is pentode connected, with the coil having four connections, with both the anode and grid connections on taps, leaving the ends, which are where the tuning occurs, free from DC across them, and more importantly, less loaded by the valve impedances, thus better for getting high-Q. Output is taken from the anode, so the mixer, V103, CV2128 (better known as our old friend ECH81, or 6AJ8), does not load the oscillator’s grid circuit. All for best stability as is the stabilised HT from V305, CV1832 (0A2 or 150C2). The oscillator configuration is Hartley, as is the BFO. The triode in the ECH81 isn’t used. With RTTY using tones, it was important the set drifted little, or the decoding wouldn’t work.

IF Amplifiers and AGC

As is usually the case, the IF amplifiers have variable degrees of coupling between primary and secondary, to derive the different bandwidths. The narrowest position uses a pair of crystals, and unusually they are between the first and second IF amplifiers. Usually, the maximum selectivity is in the anode coupling to the first IF amplifier, to attenuate unwanted signals as soon as possible, to prevent overloading and spurious response.

AGC is not applied to the third IF amplifier. This is quite normal. While applying it would sharpen the AGC response, there is a problem in that it would also at this high-level signal stage cause the effective level of modulation to be increased, and that would cause distortion of the signal. The final IFT (intermediate frequency transformer) is not quite conventional. The BFO signal is normally coupled into the detector via a very small capacitance, sometimes even relying on strays, but here there is a separate winding on the transformer, and it’s in the anode of the BFO valve. The detectors are conventional, a capacitor from the valve’s anode feeds the AGC diode, and the signal detector is fed from the transformer’s secondary. This feeds an impulse noise limiter with automatic adjustment for the normal IF level, and manual adjustment of the threshold. All four diodes are in two CV140s (EB91 or 6AL5). From then as stated the signal goes to an AGC-controlled audio pre-amp pentode, CV454 (EF93/6BA6), then the output valve CV2136 (6BW6), to be followed by the unusual transformer arrangement already described.

PSU

This brings us back to the PSU, on which I have commented. A feature again unusual is the use of two smoothing chokes, the conventional one in the HT positive, though it is mysteriously specified as 18-25 Henrys, suggesting a ‘swinging’ choke, but they are used with choke-input filters in class AB or B amplifiers, and would be expected to have a much wider range of inductance between the minimum and maximum currents, and a second choke in the centre-tap of the HT winding, to earth, it being a straightforward 20 Henrys, with two capacitors across it. Possibly a tuned arrangement. The manual does not expand on why things are done this way, I won’t either as I don’t know!

The BFO deserves some further consideration. These are normally straightforward things. Not so here. Because tones need to be generated for FSK, and there are four of them, there are four trimmers to tune the inductor to the correct pitch. The narrow position on the system switch gives a 1kHz tone and is also used for CW reception. Finally, there is a crystal-controlled position, CAl (ibrate), for zero-beating the tuning to allow the cursor to be set exactly correctly. 

Sadly, I failed to take a picture of the PSU chassis, but Figs. 3 and 4 show the set from the opposite side to the turret assembly, and an overall picture from the top. The two side chassis are separate assemblies.

Fig. 3: The set as seen from the opposite side to the turret assembly.

Fig. 4: Top view.

Conclusions

A very solid set, not difficult to service if you have the circuit, which needed not much work considering its poor storage. It is not ideal for many amateur purposes. The particular requirement for RTTY by FSK probably does not suit modern computer-generated RTTY, though I may be wrong. For both AM and CW the filters offered are not ideal, unlike so many sets of ex-military origin, where you get the option of several bandwidths.

This all said, though, you get a hell of a lot of radio for probably not much money, and as part of our heritage it is interesting. With a nice clear tuning scale and intrinsic stability, it is well suited to the shortwave listener. For anyone who collects vintage services radios, it is a well worthwhile addition.