The following text is of a technical nature. This has been done in the hope that it will help any other radio enthusiast who is contemplating the refurbishment of the A11/M11/K5/N9

The A11 & M11 under test

  The museum acquired via a generous donation, twelve M11 amplifiers, three A11 tuners, and some parts of three K5 heterodynes from a house in Fetcham. They had been stored in damp conditions for decades and all had been partly dismantled. Some parts were missing from all of the sets. It was decided that the best approach was to make up a working A11 and M11 from the wreckage available. A power supply and audio amplifier (as a substitue for the N9) would be made from modern parts to get things working. The K5 would be replaced by a signal generator until something better could be obtained.

To gain an idea of what we started with have a look at these images Donated Equipment

  The units are built on front panels made from ½ inch thick ebonite sheets. This material breaks if it is hit hard enough. Several examples had shattered front panels from rough handling. Ebonite is based on natural rubber that has had sufficient sulphur added (vulcanisation) to make it hard. It is a glossy black colour when new. Upon long exposure to ultraviolet light, it turns brown, and the sulphur leaches out. In damp conditions, this forms sulphuric acid, which firmly corrodes fixings in place. In many places. thick crusts of corrosion products had formed around fixings (and even tally plates!) The crust had to be chipped off with a screwdriver. Application of a soldering iron to the screw heads expands the metal and breaks the bond and allowing the screw to be undone. The panels are still black in areas that have not been exposed to light, such as behind control knobs. It was found that they could be polished with car body polish to give a reasonable appearance.

Nomenclature

  I have referred to components by the item numbers used in the diagrams in BR222, with an L, R or C added. For instance item 29 on the A11 circuit diagram is a fixed 3.5 jar capacitor. I refer to it as C29. The capacitances in naval documents of the period were given in “jars”, where 1 jar is 1111pF. The inductances are given in “mics” where 1 mic is 1 micro henry. Control knobs were referred to as “handles” in the old documents, presumably because the word knob was considered vulgar!

Sources of Information

  No handbooks could be found for the outfit CJ. Luckily the RN Communications Branch Museum/Library situated in HMS Collingwood, hold a copy of BR222 dating from 1930. supplemented by the Candadian Edition circa 1946. This was a training manual that was updated and revised over the years and gave much useful information. The famous “Admiralty Handbook or Wireless Telegraphy” was also revised every few years. It shows the circuit diagrams of typical receivers in service at the time, although it never names the set being referred to. Careful study shows diagrams for the outfit CJ in both the 1931 and 1938 editions of the handbook.

A11 Tuner

A11 Tuner

  This consists of a pair of coupled tuned circuits, and provides the main selectivity for the receiver. The first tuned circuit can be series or parallel tuned to suit the aerial impedance. The A11 tuner can be used in “Standby” as a single tuned circuit for searching the bands for signals. It can be switched to “Tune” where it functions as a coupled double tuned circuit. Tune gives much better sensitivity and selectivity, but requires two capacitors to be adjusted along with two inductors and a coupling control.

  Three A11 tuners were obtained. Sadly all had their brass tally plates missing. The newest had a circular ebonite plate on the top left of the panel which contained the “Valve Equivalent Capacitor “ (VEC) . This compensates for the input capacitance of the M11 amplifier when switching from standby to tune. It also had two spark gaps and the holders for two gas tubes. The corners of the front panel were broken off, so sadly this tuner was declared unrestorable.

  A second example had been largely dismantled. The coils and most of the wiring were missing. The tuning capacitors were mounted on thin aluminium formed sections, as opposed to the massive castings on the other two tuners.

  The third unit was almost complete. The R/C combination joining the FILs connection to EARTH (items 21, 22) were not fitted. There were holes in the back of the panel where the board holding these parts would have been fitted, so presumably this is a modification. The FIL terminal was connected to earth. This arrangement was left in place as it seems sensible with the modified power arrangements of the M11 as discussed later. This unit was the one chosen to be restored.

  The VEC consists of a pair of plates mounted on the rear of the front panel near the aerial terminal. It can be adjusted by slackening 6 nuts and bolts and sliding the labelled plate until the aerial tuning capacitor resonates at the same setting in “Tune” and “Standby”. This tuner had a single spark gap between aerial and earth.

  All three A11 tuners had the variable coupling coil missing. The magnetic coupling from this coil adds to, or subtracts from, the inductive coupling between the primary coil and the secondary coil. Correct coupling gives a classic “flat top” bandpass response. If the coupling is insufficient, the receiver is insensitive. if the coupling is excessive, the frequency response has a double humped shape and the receiver is also insensitive and prone to interference on other frequencies as the bandwidth is larger than it needs to be. The “Coupling” handle moves the coil up and down and rotates it. There are two positions of the coil where it is close to the secondary coils, in one position the coil is rotated 180 degrees with respect to the other so that the coupling can add or subtract to/from the coupling between the fixed coils.


  A new coil was made using a piece of grey plastic soil pipe which was to hand. It looks OK on the existing mechanism. Luckily one tuner had the broken base of a coil that contained the slip rings. The coupling handle on the front panel operates one arm to raise and lower the coil. A second arm driven by the same handle, rotates the coil so that coupling can be minimised or reversed. The coil has 38 turns of 0.2mm ECW, giving the correct inductance of 200 mics.



  The A11 tuning capacitors have a capacitance of 2 jars (2222 pF). The minimum capacitance was measured as 50pF. They actually consist of two 1 jar capacitors in parallel. They are of very unusual design. There are two sets of stator plates, and the rotor is made in two semi-circular halves insulated from each other. There are two slip rings, one at each end of the rotor and each connected to a rotor half. The slip ring brushes are each connected to one stator. There are two rotor positions at which the rotor and stator vanes are fully meshed. In one position, the rotor halves are connected to the opposite stator, giving maximum capacitance. In the other position the rotors mesh with the stators they are connected to, giving minimum capacitance. This cleverly gives a large capacitance in a small physical volume. The vanes are made from aluminium. The stator vanes have over mouldings in aluminium over their corner sections to electrically and mechanically join them together. The frames are a black insulating material I presume to be Bakelite as it does not turn brown when exposed to the light, in the way ebonite does, but I don’t really know what it is!

The capacitors can be dismantled by removing the 4 large bolts, but care needs to be taken to number the stator plates and Bakelite frames before disassembly using a scriber. Everything in these sets was hand made with limited use of jigs and tools. Note also that each stator is made in two parts. If the capacitor is re-assembled in a different order the spacings between parts will be different, and it is very hard to prevent the plates from shorting as the shaft is rotated. The parts were washed in domestic laundry detergent. The insulation around the slip rings was carefully cleaned as the dirt gave poor insulation and poor capacitor Q . The slip rings and their wire brushes need careful de-greasing to prevent intermittent contacts. Luckily they had been greased at some time which had prevented corrosion. On assembly the plain bearings were greased with castrol LM grease and the slip rings with Electrolube CG53A contact grease. The bearings need to be adjusted to centralise the capacitor plates to prevent short circuits as the shaft is turned. They were set to give a tiny amount of axial play.

  The slip ring brushes are fitted towards the front panel on the tops of the capacitors, and towards the back panel on the undersides. Spacing washers are fitted between the Bakelite frame and the non-brush side of the stator. The shafts at the front panel ends of the capacitors are cross drilled to take a taper pin when fitting the handle. The pin is vertical when the dial reads 0 degrees. The shafts at the back of the capacitors are not cross drilled which helps when re-assembling. The cross drilling was done without any attempt to make the drill perpendicular to the shaft, so that a handle cannot easily be swapped from one capacitor to another.

Care needs to be taken in re-assembly to ensure that both capacitors in one gang have maximum capacitance in the same shaft position! The inter-capacitor coupling on one capacitor may need to be turned 180 degrees to correct this.
  The 3.5 jar capacitor (item 29) measured 3.92nF with >20M ohms insulation resistance.

M11 Amplifier

M11 Amplifier

  This is a 4 valve unit. The first 3 valves are un-neutralised RF amplifiers, the fourth is a leaky grid detector. As all the valves are triodes, the RF amplifiers are very prone to oscillate. This is partially prevented by biasing the grids positively from a potentiometer connected across the LT supply. This only applies to the second and third stage. When the first stage oscillates, you are in trouble! When the grid is positively biased it draws grid current and damps the tuned circuits. I thought it unusual that there was no facility for reaction, as was typical in many old straight sets. The Admiralty handbooks explain that at LF, reaction reduces the tuned circuit bandwidth to the point that if tuned for a reasonable beat note, the signal would be outside of the receiver passband, losing sensitivity. To receive a CW station, the K5 heterodyne is used to provide a beat note. In modern terms the set then becomes a direct conversion receiver.

  The set was originally made for NR13 bright emitter valves fitted with an inter-service base. The valves were fitted with the base above the glass envelope. The valve holders have 4 prongs at right angles to the panel and the valves slide in. At some time the valves were changed to NR17 dull emitters, with a conventional B4 base. The navy had adapters made that enabled the new valves to be fitted to these sets. Later M11 amplifiers had right angled B4 bases fitted through slots in the panel just above the coil boxes. The valves are mounted with the envelopes upwards. The M11 we restored is one of these later versions.

  The original power arrangements shown in BR222 for the M11 are unconventional, although similar circuits are shown in the Admiralty handbooks. The -LT rail is connected to the A11 (labelled “FIL”). The -HT connection goes to the +LT terminal. No power supply connection goes directly to earth. This gives some fault protection in the case that a terminal is accidentally earthed, but means that neither the -LT or -HT can be deliberately earthed. The A11 tuner originally had an isolating capacitor and resistor between the FIL and earth connections. If the -LT was earthed and a fault earthed the +HT, all valve filaments would be destroyed. This setup is also worrying if a mains PSU is used, as a primary to secondary short in a mains transformer would give mains voltage on the terminals.

  Some of the M11 amplifiers have been modified to a more conventional power connection arrangement with -LT and -HT being common and earthed. A paxolin sheet is fitted under the power terminals with the new labelling. One set with the later valve holders has the terminals in the new arrangement as standard. This modification must have happened after 1930 as it is not mentioned in BR222. The Admiralty wireless handbook of 1931 page 577 also shows the old arrangement. Perhaps the A11 modification to connect the FIL terminal to earth was to be used with the modified amplifier. I restored the M11 to this state. The admiralty handbook 1938 paragraph F25 shows a complete outfit CJ with the modified power arrangement.

  Page ND2 of BR222 is dated 31/1/35 (see References), and shows a power distribution scheme for a number of receiving bays. The -LT and -HT are earthed. The M11 was supplied with 4V and 48V HT, the then current arrangement with dull emitter valves. This arrangement is referred to as the “central battery” arrangement.

M11 Component Values

  BR222 does not give component values apart from the tuning capacitors and coils. The following are the values that were fitted:

  The detector grid leak (item 63) was missing from most M11 amplifiers, presumably because it was only held in with clips. One early set had a 2Meg resistor, but it was longer than the part missing from most sets. A modern 2M2 resistor was fitted into a 1920s Ediswan glass cased part. The brass end caps were drilled end caps to take the resistor wires. The end caps are held onto glass tube by solder, as it was originally built.
  Some examples of the grid leak bypass condenser (item 64) were still present on the M11s. They were paper dielectric condensers made by TCC in green painted tin plate boxes. They were marked 0.2 jar = 222pF. A modern 220pF silver mica capacitor was fitted into one of the old cases. A new lid was made from FR4 PCB material. The lid was soldered to the case.

  Decoupling capacitors (items 67 and 68) were missing in all examples of the amplifiers. WT handbook 1938 paragraph F25 states “C5 and C6 2uF”. The same values are given in the 1931 handbook. It is presumed that the original capacitors were also paper dielectric in tinplate boxes.

  The dimensions of the original capacitors could be determined from the clamps and the black areas left on the ebonite panel where the original capacitors had blocked the day light. New capacitors were made from blocks of wood painted green, with channels milled out to take modern 2.2uF 250V polyester capacitors.

  The Filament rheostat measured 15 ohms and needed dismantling and cleaning. The RF amp grid bias potentiometer measured 1000 ohms and needed similar treatment.

M11 Coil Boxes

Edge of Coilbox		Top of Coilbox

M11 Circuit and Wiring diagram and Coil Box Terminals

M11 Tuning Capacitors

  21 coil boxes were acquired with the M11s. There are two designs of coil box. The original design has fixed brass contact studs. The later boxes have sprung contacts. The M11 coil receptacle originally had sprung strip contacts. Later units had fixed studs. BR222 states that the sprung contact coil packs can be used with the sprung strip amplifiers, but the stud coil packs cannot be used with the later stud coil boxes.

  The early coil boxes had their wavelength ranges engraved in the ebonite front panel. The serial number was engraved on the right hand side handle. It is all too easy to swap handles between coil boxes when working on them.
  In later years, as communication channel definitions, changed from wavelength to frequency, these boxes were modified by the addition of a nickel plated brass tally plate, giving both the frequency and wavelength range. Inconveniently the plates cover up some of the fixing screws, they were not always fitted in the same place on the panel! Later sprung contact coil boxes had frequency and wavelength engraved on the panel, along with the serial number.

  Coil box dimensions are not very consistent. Some boxes don’t fit some M11s. The boxes contain three RF transformers, identical for each range. The coils are wave wound on cardboard formers, and are insecurely fixed to the coil box lids using 4BA screws and cylinders of ebonite. When removing lids it is easy to rotate the cylinders and tear off the coil terminations. Several boxes had open circuit coils, due either to mechanical damage or corrosion.
  It was possible to swap coils between boxes. Mechanical parts such as lids have holes drilled in slightly different places, meaning that they need filing to make them fit. The sprung contact coils have slots milled in the front panel strip to take the wiring. The lowest frequency coil box was only issued in small numbers. Only one example existed amongst the 21 boxes. It had an open circuit winding, but this was repaired by removing a turn until the break was found.

K5 Heterodyne

  This was used to bring the receiver onto frequency and as a heterodyne to receive CW. The buzzer modulated note is essential to set up the seven adjustments on the A11 and three on the M11 to get the set on frequency. A collection of bits including three partially stripped front panels, three of the six plug in coils and some spare tuning capacitors were acquired. There were no buzzer or case parts. With no working original K5 I have used a Signal Generator. However, I originally used crude croc-clip leads to make a wafting coil to tune the Outfit CJ and although this did work to some degree some improvement was needed. To this end I have made a 'wafting' coil from various spare parts. The ebonite came from some pieces of an M11 amplifier front panel that appeared to have taken on the Scharnhorst and lost.... The winding is a temporary 20 turn coil using enamelled copper wire which once trialled will determine if we need more or less turns and the final coil will have turns made from cotton covered wire. The images below show the stages of manufactur - click on an image to enlarge it.

Update

  After a long hard search for a working K5 it was decided to try and build one using the collection of bits we had inherited. Although not strictly an exact copy of an original K5, one has to admire the skill and attention to detail of Andrew Sinclair in producing a near copy of an original K5. For the time being we use the frequency of 198kHz (Radio 4) to demonstrate the Outfit CJ but this also required the correct coil - once again Andrew came to the rescue by manufacturing a new coil

Original Coupling Coils	M11 Panel Section	Boring centre hole	Cutting slot	Finished parts	Finished coil

K5 Coils

    The following K5 coils are held:
      22000-4000m 14kHz-75kHz
      6000-2200m 50kHz-136kHz
      340-230m 882kHz-1304kHz
      One incomplete, tuning range unknown.

N9 Note Magnifier

  The M11 amplifier will drive high impedance (Browns type F) head hones directly, although it is not very loud. The Navy optionally supplied an “N9 Note Magnifier” with these outfits. This was an audio amplifier that drove low impedance telephones (head phones in our parlance) via a separate telephone transformer. No volume control was fitted, although the unit could be switched from no valves in circuit through to 3 valves to give some control of volume. Unfortunately we have so far failed to find an N9. We therefore built our own equivalent to the N9 and also fitted it with a power supply.

'Homebrew' replacement for the N9 and Power Supply Unit

Power Supply / Note Magnifier

  To demonstrate the equipment to the public, we needed a power supply to produce 50V HT and 2V LT to operate the valves in the M11 and (one day) the K5. We also need to amplify the M11 output to drive a loudspeaker. In with the M11 amplifiers was a wooden box that contained parts removed from the equipment. It was about the right size to contain a power supply and amplifier using modern technology.

  The HT supply is variable as a means of taming the M11 amplifier, It is wildly unstable above 35V and often needs less. A transformer produces 30V AC which is rectified to give around 42V DC. The HT regulator is probably more complicated than it needs to be, but it gives a very clean output and is current limited to 12mA for safety. 35V is generally considered safe to touch. The HT is carried on brass terminals which might come into contact with the visiting public.

  The LT supply is derived from a transformer with two 7.5V secondaries in series. The centre tap is connected to earth giving supplies of about +/-10V. The positive rail supplies an LM317 regulator fitted to a heat sink mounted outside the box. It is capable of supplying currents of up to 1A. This gives spare capacity to run an N9 and K5 in the future.

  The audio amplifier (note magnifier) consists of an op-amp pre amplifier, a 3kHz low pass filter and an LM386 power amplifier driving the speaker. The detector terminal on the M11 is intended to drive the primary winding of a transformer connected to the HT rail. R18 replaces the transformer primary. Additional decoupling is required to keep noise from the HT regulator from getting into the audio chain. Back to back diodes protect the op-amp input from spikes from the HT rail. The second op-amp is a low pass filter. It is needed when receiving on the lowest frequency band to prevent the RF carrier frequency being heard as a whistle. The amplifier had considerable hum problems. It was found that a star earthing arrangement almost eliminates this.

Outfit CI/CJ Production Volumes

  The three M11s that have tally plates are:

Serial Number	AP Number	Date	Valve Base	PSU
P50 (Bob Finch)	7484A	1930	Inter-services 2	-HT connected to +LT
G34	7484A	1929	B4	Built with -LT, -HT common
G36	7484A	1929	B4	Modified for-LT, -HT common
P81	7484A	1932	B4	-HT connected to +LT

  It is baffling that the later set had the older power supply arrangement. Why does a set with a 1930 date have the old valve bases? Perhaps they were built to order for specific ships?

  The book "Handbook of Signalling Equipment" dating from 1939 lists every ship in the fleet and it’s radio fit. The A11, M11, K5, N9 form outfit CJ. The A11, M11, K5, N9 with a short-wave tuner and amplifier form outfit CI. Some Outfit CNs also had A11 and M11 instead of the specified units for this outfit!

  The only capital ships with CJ or CJ were Nelson and Rodney. 91 were fitted in cruisers, 25 in aircraft carriers. 18 in flotilla leaders, 71 in smaller craft. This makes a total of 207 still in service at the start of the war! It seems surprising that they hadn't all been replaced by then.

  I imagine they would have made several hundred of these sets between 1926 and 1932. Presumably many would have been replaced by 1939. The IEE paper of 1930, listed below shows what is likely to be an outfit CI and states it was “widely fitted in the fleet”. Even more worryingly, the receiver outfit C dates from 1912 and makes the outfit CJ look modern. Amazingly they still had 57 in service, 9 of them in cruisers.

Royal Navy LF receiver timeline 1912 to 1946

Outfit C

  The receiver outfit C was introduced in 1912 and saw considerable modification and upgrades throughout it’s service life. In the version in BR222 it covered 55kHz to 1400kHz. It consisted of a tuner consisting of coils condensers and switches, all in individual boxes. There was am M3B amplifier and a N9 note magnifier. About 20 units to make a complete receiver.

Outfit CJ

  The outfit CJ was in service from roughly 1926 to 1940covering 12kHz to 545kHz. It consisted of an A11 tuner, M11 amplifier and K5 heterodyne and N9 note magnifier. Four units were needed to make a receiver. An instruction to dispose of the A11,M11,K5 and N9 was issued in late 1944.

Outfit CN

  The outfit CJ was replaced by the outfit CN (A19 tuner, M19 amplifier, N19 note magnifier). The M19 had screened grid valves (presumably solving the stability problem) and the heterodyne built in. Three units to make a receiver.

Outfit CAA

  The CN was replaced by the outfit CAA (A50 tuner and M50 amplifier) which in block diagram form doesn't look much different to the CN, apart from having a pluggable arrangement of up to 4 front end tuned circuits. The description states that all four were needed for receiving when the ship was transmitting. Two units to make a receiver.

B29

 &msp;The B29 is similar to the outfit CN, but has two front end tuned circuits. It can be switched from standby to tune as per the A11. The B29 is a complete receiver in one box! Presumably the outfit CAA was the only receiver that could receive when the ship was transmitting.

B41

  In 1946 the B41 receiver was introduced. It was a superhet covering 15kHz to 700kHz.

Outfit CJ Operating Instructions

  At the time of writing the museum does not have a K5 or N9. To operate the receiver, you will need:
    An AM signal generator and the PSU/Note Magnifier.

    An aerial cable (BNC socket to red and black leads with 2BA solder tags.),
    Two 18 inch brown leads to connect the A11 to the M11.

    One 8 inch brown lead connecting the detector output of the M11 to the input of the PSU/Note magnifier.

    Three 6 inch brown leads connecting the M11 to the PSU/Note magnifier HT+, LT+, COM-.
    A cable from signal generator output terminated in a coupling coil, 2 turns of wire 2 inch diameter.

    The A11 and M11 need to be separated by 12 inches or more, to reduce instability.

To tune the M11 to a frequency.

  Tune the signal generator accurately to the wanted frequency. Set the signal generator to AM with a modulation depth of 50% if this is adjustable. Set the output to 0dBm or a hundred millivolts. Connect the generator to the coupling coil. Fit the appropriate coil box to the M11.

Warnings

  1. Be patient. This set is not easy to tune, and it needs lots of practice to hear anything.
  2. Note the valves are numbered 1,2,3,4 left to right.
  3. When tuning, take care not to overload the detector as this makes it hard to hear the tuning adjustments. If necessary
    reduce generator output or move the coupling coil away from the set.
  4. Changing any control will affect all the other controls due to feedback in the valves.
  5. Do not let the M11 oscillate. When it does, turn the bias clockwise. Next reduce the HT voltage.

  Switch on the power supply. Set the volume control to about 1/3 maximum, the HT voltage to about 20V. On the M11 set the filament control fully clockwise and the bias control fully anticlockwise.

    Waft the coupling coil below the detector valve V4 and tune the right hand condenser for maximum signal. (You are
    coupling the coil to the right hand pair of coils in the coil box.)

    Waft the coupling coil under V3 and tune the middle condenser.
    Waft the coupling coil under V2 and tune the left condenser.

To tune the A11 to a frequency

  This assumes the M11 and signal generator are already on frequency and the aerial is connected.
    Waft the coupling coil above the left hand side of the top of the A11.

    Set the A11 right hand stud switch to “standby”. (This disables the secondary tuned circuit.)

    Select an aerial inductance as a first guess, 1 for high frequencies and 8 for the lowest. Sweep the aerial tuning condenser
    for a signal peak. If none is found, try changing the series/parallel stud switch and the aerial inductance. The set may start

    to receive signals at this point.

  To improve reception the secondary tuned circuit can be switched in by turning the right hand stud switch to “tune”. As a starting point, set the secondary inductance to the same number position as the aerial inductance. Set the coupling a bit less than the maximum position. Sweep the secondary tuning condenser for a peak. If none is found try changing the secondary tuning inductance. Once both circuits are peaked the coupling can be adjusted for loudest signal / least interference.

To use the outfit CJ to search for signals

  Just don’t try.

Safety Note

  The M11 uses open brass terminals for all connections. There is nothing to stop somebody touching these terminals. The power supply produces a maximum voltage of 35V which is considered safe for human contact in all standards we can find. In addition the power supply is current limited to 12mA which is unlikely to cause injury to anybody. If the HT was inadvertently connected to the LT terminal on the M11, the valves should be unharmed, although this has not been tried.

Hum

  The setup is very prone to hum. This is mainly caused by the detector stage of the unscreened M11 picking up mains hum. It cannot be filtered out as the detector is inherently non-linear and produces a spiky hum waveform containing harmonics across the audio frequency band. It is important to route mains cables away from the M11. Reducing the note magnifier volume and getting a large signal from the M11 helps. Remember the ships used DC mains in the 1920s!

References

   The Electron and Sea Power Vice Admiral Sir Arthur Hezlet p183 Lists typical numbers of receivers carried in ships.

   BR222 1930 (referred to as BR222 in this document). When using the BR222 links below, scroll to the relevant page.
   Page BA2 General notes on Tuners A4 A5 A9 and A11
   Page BA10 Tuner A11
   Page H14 Amplifier M11
   Page GE2 Heterodyne Unit K5
   Page ND2 Central battery wiring.
  Admiralty handbook of Wireless Telegraphy 1931 Page 577
  Admiralty handbook of Wireless Telegraphy 1938 paragraph F25 diagram resembles outfit CJ
  Handbook of Signalling Equipment 1939
  Proceedings of the IEE Wireless Section Volume 5 1930 Page 9 “Naval Wireless Telegraph Communications” Shearing and Dorling.
  Admiralty Fleet Order AFO6895/44 Disposal of Obsolete Stores December 1944

Yours aye
20/2/24
Andrew Sinclair