A Generation Onwards
The Electronic Organ Constructors' Society celebrated its 21st year of existence a few weeks ago. Those who decided to form this society were already actively engaged in building organs, but in those days it was necessary to be a diehard enthusiast and very little practical information was then available. Little did they dream that the 'organ-on-a-chip' would become a practical reality.
It is fascinating to consider the speed with which electronics has advanced over the past three decades. The transistor was born just over thirty years ago, but some years were to elapse before that first, rather flimsy point contact device was perfected and became a commercial proposition — the pnp junction transistor. Silicon replaced Germanium and new devices appeared in rapid succession, including encapsulated blocks of transistors and, of course, the IC itself. In a life-time, most of us have seen the transition from the era of the valve to micro-miniaturisation of complete circuit blocks consuming only small quantities of power. The LCD watch, digital multimeters, mini-computers and indeed space research are facts of life we accept, but none of these would have been possible without solid-state circuitry.
How has all this affected the organ constructor? E.O.C.S. has, of course, moved with the times. The club was formed to pool information originally and current editions of its magazine contain novel and interesting circuitry to aid its members. In today's realm of logic families, special purpose ICs and digital techniques, it may be interesting to look at the hurdles that faced the previous generation. My research may encourage someone to lay down the keel of an organ when he sees how much easier the process is today.
Publications dealing with organ construction were rather uncommon and, without knowledgeable guidance, it seems that it was necessary either to be an expert on electronics or simply operate on the 'wet finger' technique. Occasional magazine articles appeared which stimulated interest although they often dealt with a very basic single manual affair — and were based on valve circuitry. A number of the society's members were drawn in by a small ad. of the time which invited readers to send for a two manual and pedal circuit at a fairly inflated price: the end result was a very unstable instrument. Indeed, this particular organ circuit contained so many problems that it was the reason for starting up a constructors' society.
Books by the American author, Richard Dorf, and by Alan Douglas, doyen of the organ enthusiasts in the U.K., were followed eagerly — when they could be obtained. Both authors have continued to write on the subject and their books are now far less difficult to find. (Alan Douglas was an active member of E.O.C.S. from its inception and only recently retired as the society's President.) Having managed to glean a few ideas, the constructor had to decide on a method of tone generation. Those with engineering facilities cut toothed wheels to rotate in front of magnet/coil assemblies on the Hammond principle. The valve was the state-of-the-art device at that time so the majority of builders used this — or possibly neon lamp relaxation oscillators.
One of the advantages of a valve organ was that the player kept warm on cool evenings! A typical valve instrument generating six octaves would have required three double triodes for each of the 12 chromatic notes: one triode section would have been used as a master oscillator and the remaining five sections as frequency dividers. Throw in a few more valve stages for vibrato, mixing and preamplification and up to 25 amps of filament current was required, excluding the main amplifier.
The Eccles-Jordan (bistable) divider was invented in 1919, I believe, and of course the transistorised version of this circuit forms the basis of most of today's generators — probably encapsulated and combined with keying circuitry. In valve terms, two triodes were required and, although certain manufacturers used valve bistables, the constructor had to find a cheaper method. Consequently he would have chosen an R-C divider using a single triode stage, possibly a blocking oscillator. The latter method and neon relaxation dividers produced a sawtooth waveform containing the full series of harmonics, ideally suited to subtractive tone-forming.
The HT requirement for a valve organ was in the region of 150mA at 300V, perhaps dropped in stages. This was sufficient to sharply remind the constructor if he was careless! Testing a partly-completed circuit meant muting the power supply with hefty resistors to dissipate unwanted power — a very time consuming and tedious process. The eventual change to solid-state methods was joyous — no real problems with a breadboard and battery. The roles have been reversed since now the constructor has to avoid giving his CMOS a shock inadvertently!
Twelve master oscillators — one for each note of the chromatic scale — were required and each had to be tuned accurately. Based on a triode section, the oscillators were usually of the L-C type (typically Hartley) and occasionally of phase-shift configuration. There was no difficulty in making these stable but, before the Leslie speaker became popular, good vibrato was a prerequisite. Unfortunately, stable oscillators are difficult to modulate with vibrato so the 'Q' of these valve oscillators had to be diluted until a compromise between stability and vibrato could be found.
Any commercial circuitry that could be obtained was scrutinised for ideas. Both Allen and Conn used free phase methods, where each frequency was separately produced and tuned (analogous to the pipe organ, in fact). The chorus effect obtainable from this system was an enviable feature and many home-brewed instruments employed this principle. However, they were hardly portable due to the weight of the inductors used at the lower end of the frequency scale. The free phase organ is still popular today, the transistor version with pot core inductors being far less unwieldy.
Keying the generated frequency was problematical before the advent of transistors. Both the power requirement and cost of valve operated 'gates' were such that the principle was feasible only in expensive commercial instruments. The constructor avoided metal to metal key contacts because, due to the fairly high impedances involved, clicks and thumps could be clearly audible. Resistive switching was often used to hide the transients: the key contact could be made to traverse a piece of pot. card or dip into a conductive liquid, such as commercial anti-freeze! Another method used at the time was to place a chain of high value resistors between the generators and preamplifier, the playing key (at rest) earthing out a centre tap: this method was unfortunately prone to causing embarrassing ciphers. Diode keying began to take over, allowing several pitches to be keyed from one pair of contacts carrying a keying voltage: some control over sustain was another advantage of diode keying. Free phase organs, as is still the case today, used DC switches under the playing keys to supply the various individual oscillators. An R-C network after the keyswitch allowed the generated tone to rise and fall away in volume, giving an effect similar to the pipe organ.
Professionally made keyboards were less easy to obtain than today and often had to be sprung and contacts of some sort fitted. The keys were often heavy and suffered from considerable inertia, unlike the modern plastics key with metal extension. Those tempted to cannibalise old pianos found themselves with unmatched keyboards and pivot points that were in the wrong place.
Tone forming methods do not seem to have changed radically. The generated waveform has, however, tended to move from sawtooth to square wave for reasons of economy and the ease with which bistable dividers can be synchronised. Keyed signals from the generators were applied, by means of the stop switches, to various filters — much as they are today. Low-pass, high-pass and tuned filters subtracted unwanted overtones from the rich sawtooth waveform to give flute, string and reed tones.
A sawtooth wave is almost ideal, except when trying to imitate a clarinet or similar stop. To surmount this difficulty, earlier instruments incorporated an outphaser. This circuit mixed 8' and 4' (sawtooth) pitches in phase opposition and produced an 8' square wave which was ideal for those woody, hollow voices.
Outputs from the various departments had to be applied to further valve stages for mixing, preamplification and possibly reverberation drive. It was not unusual to use a ni-chrome electric fire element as the reverb spring, with an ordinary crystal cartridge at each end acting as transducers. Rather Heath Robinson perhaps, but the idea often worked adequately!
Before the LDR arrived, the Swell Pedal had to consist of a Meccano-geared pot. This often became noisy after a short period of use, especially if the instrument was played by someone with a rhythmic right foot! The main amplifier itself had to be treated with respect, with possibly 450V on the anodes of the output valves. Fortunately, Messrs Tobey and Dinsdale soon ousted that bulky and heavy chassis from the organ and were the forerunners of current small but powerful power amplifiers.
Yes — organ construction was a massive commitment a generation ago. Also, there was nowhere you could actually go to hear your organ before you built it: once started, the process could take a couple of years because of the time involved in winding hundreds of coils and working out a suitable method of keying.
If fathers were game for the task, it's that much easier for their sons. There is no lack of published detail, solid state circuitry has reduced the risk of a good 'belt' to those unused to the art of electronics and purpose-designed devices have telescoped both the time and space involved. Specialist suppliers such as Wersi, offering a vast range of superb instruments, and Maplin are extremely helpful to those venturing into construction.
If building from a kit, it is possible to listen to the final version before deciding if it is to your musical taste. Finally, it is probably true to say that, thanks to the mass production of solid-state devices, the cost of building an organ has risen slower than inflation over the past thirty years.
Perhaps the only aspect to cause the constructor to hesitate today is the choice of possible avenues — but that must be his decision!
Feature by Ken Lenton-Smith
mu:zines is the result of thousands of hours of effort, and will require many thousands more going forward to reach our goals of getting all this content online.
If you value this resource, you can support this project - it really helps!