Photographs courtesy of Derek Batten

You are now looking at a Series Two Melotone unit.

The curious mechanism in the top section of the left-hand cabinet is not a bench grinder! It is the generator assembly, which is seen to better advantage in the next photograph.

The wooden cabinet assembly beneath the generator shelf, houses the mechanism for time constants and harmonic integration. The time constant/voicing resistor trays (72 notes) sit on a shelf above the four keying relay trays (80 notes, including spares). Note the bottom tray has been drawn forward for the benefit of the photograph. Similarly, the vertical tag-board, on the right, has been swung forward to reveal a mere glimpse of the relay box controlling the capacitor network of the echo (sustain) circuitry.

The smaller cabinet, to the right of the generator cabinet, reveals one of several alternative sound systems used for Melotones – in this example, a pair of Ferranti amplifiers.

Series One Melotones appear to have started life generally housed in single cabinets. Then followed the Series Two, with a detached amplifier cabinet, while the Series Three Melotones were built on a single, tall frame. In the latter design, the amplifier was mounted at the top, with the generator assembly (this time, mounted vertically) and all else following the same format as a Series Two unit.

The mode of tone generation is electrostatic, as distinct from the Hammond electromagnetic principle. Each note is produced by an undulating electrical potential, generated by the relative rotation of two parts – one of which carries a wave or undulating form, and the other, a scanning array for scanning said waveform. A Melotone generator assembly consists of two stator discs – one either side of a double-sided rotating scanning disc. There are two octaves of undulating waveform rings on each of the stators – natural notes of the scale are on one, and sharps on the other. In order to cover the six-octave compass of the Melotone, two such generator assemblies are employed: a bass disc, generating the two bottom octaves from Tenor C (utilising two of the four available octaves), and a treble disc generating a further four octaves. They appear respectively left and right in the photograph.

The following selection of drawings are extracted from Leslie Bourn’s various patent applications, the earliest of which dates from 1932. This, and subsequent revisions, serve to address the general principle of electrostatic tone production in a musical instrument, and therefore do not specifically refer to the Melotone unit – or any particular pre-war organ. The italicised text references correspond with those used by Bourn in his drawings.

Bass and treble discs are identical, but speed of rotation - and consequently, pitch - is governed by the size of drive pulley (23). The treble disc revolves at four times the speed of the bass disc. The close-up photograph clearly shows the larger drive pulley for the bass disc, and a proportionally smaller diameter for the treble.

Each rotor disc surface overlies and rotates in close proximity to its allotted stator plate, much in the nature of an air-spaced capacitor. In practice, the undulating waveform rings (1) are formed by first providing the insulating disc (3) with an electrically-conductive coating.

In his patent applications, Leslie Bourn specifies a graphite coating, but in the actual production, a more satisfactory coating of nickel appears to have been used.

This coating is then divided into concentric rings by means of longitudinal sine wave cuts (a) of the required wavelength. The simplified diagram below represents a segment of a stator disc, but with the waveforms on a straight path. It will be noted that there are two engravings (a) per note - two sine waves in antiphase - the innermost (and consequently smaller-diameter) waveform being delayed by half its period. The engraving process effectively leaves the waveform rings (1) isolated from each other, and the active surfaces of the stators - including both the waveforms and the sections between - are electrically connected to wire pins, passed through the moulding. The waveforms (1) are then connected to the time-constant circuits and the sections (4) between linked together and rendered ineffective by being earthed. It will be noted that the highest note is to be found at the outer edge of the disc, and the lowest at the centre.

The concentric rings (2a) and radial scanning members (2b) of the grid electrode (2) are formed of graphite (or other conductive medium), located in the channels engraved in the insulating material of the rotor disc (5). After engraving, the disc is completely coated with a graphite or conductive layer, and the surface then rubbed down, to leave only the infill in the channels.

The common scanning electrode takes the form of a grid (2), consisting of a number of concentric ring supports (2a), which are superimposed over the respective spaces between the undulating rings of the stator disc. Short radial scanning members (2b) extend between each adjacent pair of these concentric rings, the radial members of each set lying transversely over a respective undulating ring (1). These scanning members are spaced by a distance equal to the wave length of the ring being scanned.

Design of the original Series One/Two discs proved to be a disaster. An acetate-based form of plastic appears to have been used, which created the problem of it having to be kept at a relatively stable temperature. This was all the more critical as the acetate layer was riveted onto an aluminium backing (flywheel) disc. Ian Macnaught (Scottish Cinema Organ Trust) sums up the situation thus: “If the disc temperature was allowed to go up and down – like hot during the day, and cool at night – it kneaded the plastic, causing a catalytic action, which resulted in a vinegar-smelling gas to be given off. The loss of mass in the material then allowed decay of the acetate, into a ‘toffee’ consistency, which gradually poured itself out of the bottom of the disc assembly. The discs became warped, corrugated, and because of the tight limit [between rotor and stator], allowed them to scuff, eventually sticking firm. Normally, the Melotone cabinets were kept stable by the chamber heaters, or by a small bulb in the base of the cabinets”. The accompanying picture – also courtesy of Ian – dramatically illustrates this point, the casualty in this instance being a disc from the Mayfair, Aigburth Melotone.
 
With the Series Three Melotone, came a more satisfactory Bakelite disc. However, while not as sensitive to operating temperatures, the more brittle nature of the material has been known to cause cracking.

Despite these faults, Compton must be commended on achieving the extremely close tolerances to which these components were initially machined and assembled. In such a design, concentricity, flatness, and freedom from lateral wobble are crtical factors.

As notes are keyed, a DC voltage – around 400V at source – is passed to the appropriate rings on the stator plate. Each key contact is connected to its proper generator ring through an adjustable resistance network. The resistors, resembling cartridge fuses in construction, are contained in a series of tray assemblies (three resistors per note), enabling values to be substituted, in order to regulate each note. A capacitor (18) is also connected between each generator ring and earth. The purpose of this is to cause the potential of each ring to build up slowly, thereby avoiding a sharp attack.

From this source of potential, an undulating potential is induced on the scanning element (2), of a frequency depending on the speed of rotation and the distance apart of the undulations of the ring. This undulating potential is then collected from the drive shaft (22) via a carbon brush, and passed to a common amplifer and speaker.

A second capacitor network, affording a degree of sustain, may be brought into play by depressing the Echo Control stopkey at the console. This effectively inserts a further capacitor in parallel with each leak resistor, thereby causing the potential to diminish more slowly, on release of each note.

Vibrato is produced by an eccentric drive, which effectively rocks the two stator plates back and forth. This mechanism is clearly visible in the front centre of the generator assembly photograph, and is driven from the main motor shaft by a belt, connecting with the large split pulley on the vibrato shaft. One side of this pulley is coupled to the shaft, while the second side is an idler. When Vibrato is not selected at the console, the drive belt runs by default on the idler side of the pulley. Depressing the Vibrato tab, operates a relay arm, bearing a forked channel, through which the vibrato belt passes. The travel of this fork piece is sufficient to guide the belt onto the fixed side of the pulley, thus activating the vibrato drive.

In a typical theatre organ specification, the Melotone is provided in its basic (tibia-like) tone colour at 8, 4 2-2/3 and 2ft pitches. Two or three synthetic voices are usually included:

Krumhorn (8 + 2-2/3 + 1-3/5)
Cor Anglais (4 + 2-2/3 + 2 + 1-3/5)
Musette (4 + 2-2/3 + 1-3/5 + 1-1/7 + 8/9 + 9/11)

Though lacking in essential percussiveness, the following percussions are derived from the Melotone, economising on installation of the real instruments in a Melotone-equipped organ:

Marimba (Melotone 8, vibrato cancelled) Vibraphone (Melotone 4 + vibrato + echo control)
Carillon (8 + 6 + 4 + 2-2/3 + 2 + 1-1/3 + echo control + vibrato cancelled)
Chimes (6-2/5 + 4 + 2-2/3 + 2 + 1-2/5 + echo control + vibrato cancelled).

Comptons did experiment with different Chime harmonics, and there are known cases where the 1-2/5 pitch in the above example has been modified to 1-1/5.

Much has been spoken of the Melotone’s unique sound. This is largely due to there not being a separate disc for each of the 12 notes of the scale. On a single disc, it proved impossible to produce the absolute correct number of cycles for all notes. Consequently, some cycles had to be added or chopped, in order for them to meet around a circular path. As a result of a conversation which took place many years ago, between Ian Macnaught and Wally Fair (Leslie Bourn’s assistant), it is now known that the correct sine wave was produced on an oscilloscope. From this, bent wire was used to reproduce the waveform into a long length, and then formed into the required circular path. Where the peaks and troughs of certain notes did not always meet in the middle, some ‘editing’ was required, in order to produce a continuous path. The actual number of cycles used was the nearest possible. Equal temperament tuning is, therefore, not possible, and the Melotone is slightly out-of-tune within itself, apart from the C octaves. For this reason alone, organ restorers have failed to fully reproduce the sound of the Melotone unit by using other types of electronic organ generator – including Compton 12-disc generators.

Add to this the unique combination of 1930s Vitavox horn speakers, and the levels of distortion and colouration present in valve amplifiers of the period, and that is what makes a Melotone special!