The Electronic Keyboard (Part 2)
An historical and technical survey of the development of the electronic keyboard from its origins to the present day. This informative series will be of interest to anyone who has played an electronic keyboard instrument, whether an organ, electric piano or synthesiser, and its practical examples give the electro-music engineer a valuable insight into early music technology.
We have seen that by 1934, all the valid methods of organ tone production had been exploited, although not all were equally successful. One of the first photoelectric systems, using pre-recorded tracks on photographic emulsions, showed great promise; the German Welte company developed an excellent organ using large coated glass discs, about 3ft in diameter. This allowed them to rotate slowly. Alas, deterioration of the emulsion due to age, dirt and the light from the exciter lamps resulted in a short life, and the advent of the Second World War stopped development. But interest was revived, and Kimball tried the same idea but with modern technology, using tracks etched out on metallic-sputtered glass discs. However, mechanical problems dogged the design and it was abandoned. The Baldwin company also had a photoelectric system out on field trials, but this too came to nothing. So since the early experiments of Eremeef in 1932 and Toulouse & Midgeley in the mid-thirties, no one has succeeded in bringing the system onto the market. We might note that Mercadier tried a crude form of photoelectric generator in 1890.
The Hammond, Compton and Wurlitzer electro-mechanical generators had so far held sway because they were so much more reliable than other methods. But research under the pressure of war had enormously improved components like resistors and capacitors, and metallurgical advances improved laminations and iron cores in general. The C. G. Conn company introduced a full-scale 2 manual organ in 1947, with 164 independent oscillators each giving two independent waveforms; the company never departed from this very successful system, as not since the organs of Coupleux & Givelet had such full and rich choruses been produced. However, there is no doubt that the tone forming circuits could have been improved, and it is rather surprising that the Americans did not use any of the tone research pioneered by Frederich Trautwein or Meyer-Epler, or even Oskar Vierling.
Shortly after this the Baldwin company introduced an ingenious blocking oscillator divider system. I was invited by the company to examine this instrument in their Cincinnati factory and found they had developed tone shaping circuits almost to perfection. To this day, I believe these circuits, if used with the same waveforms, are unsurpassed. A larger organ soon appeared and the company moved orr to other models. Baldwin were the first to devise modulation of selected organ frequencies, 1, 2 or 3 Hz, to produce a random chorus effect. They also pioneered some very effective reverberation systems.
Wurlitzer were refining their earlier electrostatic reed organs and indeed as pipe organ builders their workmanship was unsurpassed. We should note here that all the instruments so far mentioned were intended for classical or church use. Wurlitzer, however, as pioneers of the cinema organ, realised there was a market for small entertainment organs for the home. They took a step towards this by making all the reeds vibrate continuously, and keying the polarising voltage. This provided a rapid attack, more suitable for popular music. They also fitted a pickup at the tip of the reeds thus obtaining twice the frequency compared with the pickup above the tongues. All the same, tonal variations were limited to those which vibrating reeds could supply.
But with saturation in church sales approaching, the home market became very tempting. Several smaller companies like Lowrey, Estey, Thomas, Minshall etc. entered the field, having found that a cheaper organ could be made by using frequency dividers supplied from just twelve oscillators. With plenty of vibrato and care in playing, quite acceptable results were obtained. The Wurlitzer company were the first to supply a complete rhythm generator unit, and it is interesting to compare the size of the "side man" with that of present day units.
All the foregoing used valves. The first organ to be transistorised was the Gulbransen. The first models used independent oscillators, but soon turned to blocking oscillators; which in turn gave way to 2:1 frequency dividers. It was inevitable that all other makers should turn to transistors but since they were only suitable for small currents, valves continued to be used in power amplifiers; it is only a short time since Philips abandoned these. However, many people in other walks of life still prefer valve where rough treatment can be expected!
A feature which is almost universal in entertainment organs is the rotating Leslie loudspeaker. Donald Leslie originally patented his many variants on this method in 1945, but for some reason it did not become popular until 1950, when it was re-issued. We might note that the John Compton organ company had obtained a British patent for the same device in 1934 and 1936. Pity it was not taken up.
As the cost of components especially transistors continued to fall, a market opened up for the home constructor. Several kits came on the market, persisting to this day. If you know what you are doing, this is a very economical way to acquire an organ and leaves one free to design any kind of console to suit one's surroundings. It also encourages experiments with tone circuits for some people will want more, some less and some quite different kinds of sounds.
The market for classical or concert organs now begins to show an upturn. This is largely because the cost of new pipe organs is prohibitive. The principal examples now available are made by Allen, Ahlborn, Rogers and Copeman Hart. The Allen is digital, more of that later. The others either use independent oscillators for each note throughout, or occasionally introduce a frequency divider rank if it is appropriate. Because of the versatility and low cost of present-day integrated circuits or micro-chips, there is a tendency to over-elaborate some of the circuits, for what purpose is not clear, since the principal function of any organ is to make music and nothing else. Although the original company is now defunct, we should not overlook the Miller organs. These originated as English versions of the Constant-Martin, but Miller soon branched out on his own and there are many examples of these separate oscillator valve organs still in use; some very large ones, as at the Sprowston church in Norwich and Khartoum cathedral. Miller always used separate ranks for flute, diapason, reed and string, keying the HT line through relays. The system was kept at a constant temperature, hence the organs held their tune.
The last attempt of the Hammond company to re-enter the concert field was the grand 100, an enlarged version of the tone wheel organs; this was not successful and was succeeded by the Concorde, their first solid state generator instrument. By this time, the electro-mechanical organs were obsolete, although The Parie in France and the Harmoniphon in Spain continued to be made for several years more.
Bracketed with the above must be the original Allen organs. Alone amongst U.S. manufacturers, Allen built to any specification. All organs had separate oscillators for every note, interesting because the inductances were toroidal. Toneforming units on trolleys were used in the larger organs, easily interchanged. The writer visited their first large instrument at Catasaqua in 1949, and found it astonishingly lifelife. Smaller entertainment organs appeared later, notably the "Carousel", which in common with the Rogers, used mechanically-struck metal bars for the celesta or glockenspiel. However, about this time the advances in integrated circuits opened the way for mass production of divider generators, and the market was rapidly saturated with instruments from Japan and Italy; strange to say, the countries once in the forefront, France and Germany, made no real contributions to these developments.
When the ingenious patents of Ralph Deutsch appeared, Allen decided to go digital and use his methods. Although it was not now possible to have an independent tuned source for every note (since these are derived from dividers), more than one complete generator is used in the larger organs.
Philips were the originators of the digital system in 1969/70, their complete generator measuring only 4½" x 4" x ½" - an extraordinary achievement. Deutsch enlarged upon the method, bringing in tone forming, attack and decay circuits, and other features, all digital. The interesting features of the current organs are that analysed components of selected organ tones are stored in a diode memory circuit, which is repeatedly scanned by a read-out system when a key is depressed and the appropriate stop gate is opened. The prime movers are oscillators which feed very accurate multivibrators which in their turn feed 2:1 dividers. The multis are kept in accurate tune by frequency comparators which hold the division ratios at 185 and 196, because 196/185 = 2 1/12 which is the interval of a semitone on the equally tempered scale; the error is extremely small, approximately 4 x 106. The selected analysis points on any waveshape could be as many as 48, giving excellent fidelity; but of course this would apply to all notes on the keyboard; so that the differences in voicing which occur in a real rank of pipes, over its compass, will not be reproduced. As in any digital device, it must be exact and precise. Shift registers provide attack and decay and though these are step functions they occur so rapidly that the ear hears them as a continuity. The inevitable high frequencies produced as the steps switch are removed by filters. Digital to analogue converters feed the audio outlets and this instrument well illustrates the many virtues of integrated circuits.
Feature by Alan Douglas
Previous article in this issue:
Next article in this issue: