The ubiquitous Leslie Speaker — or perhaps we should call it a Doppler Speaker — is so inextricably connected with various types of electronic music that we tend to take it for granted.
Listen to some really old recordings of electronic organs and you will hear a difference; the Leslie had not yet arrived. Electronic vibrato was then the order of the day and, with the possible exception of Hammond's delay time and scanner system, the effect was often rather ungainly.
In the early Hammond days (circa 1939), the loudspeaker was sometimes placed at the bottom end of a tall, square enclosure resembling a large scale church organ Bourdon pipe. Mounted just above the speaker was a butterfly valve, similar to that in a carburettor, which was rotated to provide a tremulant effect.
I wonder whether this principle set Donald Leslie thinking? The butterfly valve simply chopped the emerging sound waves into slices but he was looking for a method of imposing frequency modulation on the speaker's output so that vibrato (rather than tremulant) could be obtained. In my view he did better than that as the route he chose combines both FM (vibrato) and AM (tremulant) and it is the mixture of both types of modulation that give the Leslie Speaker its characteristic sound.
The Doppler Effect is widely used in scientific circles; astronomers, for example, use it to measure the speed of a star. If it is moving away from the earth, we see the star as slightly redder than its true colour. Conversely, when moving towards us the star will appear bluer. At the other end of the frequency scale, sound waves are equally affected.
A moving sound source, such as a train's whistle, will drop to a slightly lower frequency as it passes. The pitch of the whistle may be constant as heard by the driver in his cab but as the train approaches us the pitch we hear is slightly higher because sound waves are reaching us at the frequency of the whistle plus the speed of the train. After passing, the apparent pitch will fall because the train's speed is now subtracted from the true frequency. See Figure 1.
Another thing we notice is the increase and decrease in volume as the train approaches and then recedes. This phenomena has been translated musically in various ways. Mounting the speakers on a rotating baffle has been employed, or a speaker spun on the end of a counterbalanced rod. The problems here are that connections have to be made by slip rings, which can wear and become noisy; circular mercury troughs have been used to circumvent that problem but if the baffle is large there still remains the question of getting it spinning fast enough.
Donald Leslie's solution was to mount a downward-facing speaker in an (almost) infinite baffle, below which was a rotor. When this is turned, the listener hears sound approach and recede from him as the mouth of the rotor passes — synonymous to the passing train.
The shape of the rotor is shown in Figure 2 and the sound it emits is fairly directional. The rotor is circular for aerodynamic reasons but in effect it embodies a 45 degree 'sound mirror' under the speaker and the sound is thrown out through a wide port, almost in a centrifugal fashion (if such a thing applies to sound!). The width of this port and those in the tone cabinet or console are carefully designed for the smoothest effect. As the rotor's port passes, not only the Doppler Effect comes into play but also a momentary amplitude rise. Many electronic vibrato systems are excellent at imposing FM but the Leslie system's combination of FM and AM is quite unmistakable to hear.
Leslie's original idea has been copied by other manufacturers, perhaps with modifications of some sort, and this is the reason for the earlier suggestion that 'Doppler Speaker' should perhaps be a better generic term. If it is not obvious already, let me add that I wouldn't part with my Leslie 145 for anything!
The rotor should be made of dense material (heavy plywood is common) to ensure that all frequencies emanating from the speaker are projected through the ports. This requires a fairly powerful synchronous motor to belt-drive it. Motors with brushes are quite unsuitable as they will interfere with amplification.
When the system is used in an organ console, it is often turned on its side and the sound emerges from the side of the instrument. In this case the rotor is often made of polystyrene foam. Being light in weight, this can be driven fast by a modest-sized synchronous motor but will tend to absorb part of the speaker's output. This arrangement is reasonably adequate but better results are obtained from a separate Doppler cabinet with substantial speaker and rotor.
Although the tonal spectrum of keyboard instruments is not as wide as an orchestral recording, one speaker will be hard pressed to deal with 16 foot pedal and 1 foot manual pitches simultaneously. To emphasise modulation of the upper frequencies, Doppler cabinets are sometimes fitted with a horn pressure unit fed from a crossover network. Mounted above this are a pair of exponential horns (one of which is a dummy for dynamic balance), rotated by a second motor. (Figure 3).
The bell of the operative horn is some 4" across and the modulation is rather too transient. This being so, small dish-shaped diffusers are mounted a short distance from the mouths of the horns to give a more bland effect.
Chorale is also obtainable by running both rotor and horns at a slow speed — approximately half a revolution per second. The effect is improved if the same signal is applied to a conventional speaker in parallel with the Doppler. This results in a slow heterodyning of frequencies, the meandering of sound being reminiscent of a large cathedral organ whose pipes are always minutely out of tune with each other. Chorale is equally useful for light and secular music, providing a round and spacious effect.
Of course, it is not possible to get a heavy rotor up to full speed immediately: the average vibrato speed is 7.5Hz so the rotor must turn at 450 r.p.m. Hence the typical run-up to correct vibrato frequency takes 2-3 seconds. Some organists appear to dislike this aspect but I find it useful to sustain a note or chord while the rotations build up!
Maintenance of the mechanical parts of Doppler systems tends to get overlooked.
Motor bearings should be lightly oiled from time to time but not too liberally. The speed change from fast to Chorale is often by means of rubber-tyred pulleys, which may misbehave if they become oily. Methylated spirit applied lightly to the rubber surfaces will correct their grip in these circumstances.
Rotor and horn bearings also require occasional oiling and it should be noted that these are rubber mounted to eliminate noise. Belt drives should be inspected for wear and their tension adjusted if so required. Leslie belts are often made of cotton material but in an emergency could be replaced with light plastic belting; certain types of plastic belt can be cut and joined with a soldering iron.
E&MM readers will in most cases be highly practical, but think twice before considering construction of a Doppler system. The relatively high price of a commercial Doppler cabinet reflects the considerable know-how and skill that goes into manufacture. Be assured that a silent bearing is not that simple to make for this application. There is also the problem of finding a really powerful synchronous motor to drive that ideal but heavy rotor. Wind noise from the moving parts also has to be considered.
Weaknesses in this respect will come to the fore when you set up your stereo mics and Dolby tape. No matter how sweet the music, bearings that rattle and groan will be recorded faithfully. Finally, if you own such a speaker system remember the maintenance and if you don't own one forget about construction!