GB2037534A - Loudspeakers - Google Patents

Abstract

The diaphragm of a loudspeaker (Fig. D) feeds ports X, X', port X' being open and port X being closed by an air volume V. If the speaker is in phase with the resonance port X and volume V, it acts as an open port and port X' is ineffective. However, if the resonant amplitude builds up and increases beyond that of the speaker the air motion oscillates along port X', thereby lowering the resonant frequency and producing a phase lag. Other arrangements described require some acoustic energy from a speaker to reach an acoustic mass X and other acoustic energy from the speaker to reach an acoustic mass X'. Masses X, X' may be cylindrical or tapered ports, horns, auxiliary bass radiators, slots or further powered units. An L.E.D. may indicate overload. Infra sound below 20 Hz may be reproduced. <IMAGE>

Description

SPECIFICATION The 'Graham Holliman Velocity-coupled Infrabass Speaker' General Background Information For almost as long as it has been possible to produce sound by electrical means people have been trying to find ways to produce high quality, low distortion, high sound level bass frequencies in this way. The search has been intensifying over the last decade, with the advent of electronic musical instruments, with various research projects on the effects of infrasound, and in particular, the growing popularity of High Fidelity sound reproduction.

Two things have made the goal impossible to achieve. Firstly, all known systems produce considerable amounts of harmonic distortion. This is usually due to the fact that every factor that has any influence (listening room l.f. cutoff, drive unit resonant frequency, acoustic impedance mismatch, driver cone mass, enclosure air pressure and its effect on resonant frequency of the driver, increasing wavelength with decreasing frequency and therefore mismatch to the size of the driver, changing acoustic impedance with frequency due to listening room standing waves, changing polar radiation pattern pattern with frequency due to changing wavelength, low conversion efficiency etc.) is working together with every other to reduce the output amplitude for a given input progressively more as the frequency is reduced.Hence, one either accepts the lack in low bass volume, or attempts some form of low bass compensation by increasing the amplitude of these frequencies electronically. This latter course is doomed to failure because, due to the massive increase in signal level required, the driver would be required to execute such a large excursion that its motion would immediately start to produce gross distortion due to non-linearities in the suspension and/or magnet system. Even if this were not the case, interaction between the unit and its environment would produce similar effects. Increasing the size of the unit will worsen the effects that its enclosure will produce on its resonant frequency.

The second thing which is particularly important in the reproduction of music, (or its initial performance), which is not a steady monotone, but contains many transient changes in amplitude at any given frequency, is the correct rendition of frequencies below the lowest frequency which is considered to be 'audible', down to a very low frequency indeed (probably down to about 3 Hz., although 8-10 Hz. lower limit would seem to offer a worthwhile improvement. This is obvious in theory, as a perfect transient has an infinite bandwidth, but has never been able to be checked, as all previous speaker systems available have been inadequate.

In fact, most amplifier manufacturers have fitted rumble filters in order to avoid the excessive low bass overload distortion which otherwise results.

The use of these rumble filters does of course achieve the desired reduction in overload distortion but at the same time guarantees that bass transients can never be correctly reproduced.

The following is a description of a bass loudspeaker which I have invented, and would like to call the 'Graham Holliman Velocity-Coupled Infra-Bass Speaker', because it is capable of reproducing, from an electrical signal, frequencies lower than any other speaker of comparable size, with greater purity and higher efficiency, and has its l.f. cutoff (3 dB. point) well below the lowest audible frequency, (commonly accepted as being around 1 8-20 Hz.). This speaker also possesses the considerable advantage that it is far less affected by the listening room acoustics than conventional speakers, both in regard to its ability to 'not drive' room standing waves, and also to drive well into the room conditions prevailing at and below the room's lowest resonant frequency (the If. cutoff frequency).

I will limit the description to the reproduction of bass frequencies, although obviously further conventional drive units and crossover networks could be added to make up a speaker covering the whole range, if required. the unit under discussion in its simplest form, would be ideal as an add-on speaker to an existing system, either one or two being added to extend the low frequency response of the system by 3 octaves approximately.

As an introduction, consider diagram A. This represents a Helmholtz resonator, in which the resonant frequency is governed by the enclosed volume of air, V, and the length and crosssectional area of the port, X. In particular, if the length of the port, X, is increased, the resonant frequency of the system decreases, all else remaining equal. The Helmholtz resonator has been used for many years in base loudspeaker systems, and is known as the 'Bass-Reflex' or 'Phase-lnverter' system, see diagram B.This system has numerous disadvantages, some of the more serious being that the natural resonant frequency of the drive unit used is raised considerably by the springiness of the air, V, in the box, and also that the sound output from the port below its own resonant frequency is out of phase with that from the drive unit, with the consequence that, below resonance the total sound output falls extremely rapidly. Due to the requirement of a smooth frequency response at upper bass frequencies, the port has to be made so as to resonate at a frequency very close to that of the drive unit when mounted in the box, and this therefore means that bass-reflex systems cannot produce genuine deep bass, unless they are made to unmanageably large proportions.

Consider diagram C. The resonant frequency is determined by X and V as before, but it can be instantaneously changed to a lower frequency by the closure of the flap F, which introduces an extra length of port, X'.

Now consider diagram D., in which the flap F has been replaced by a speaker drive unit. If the speaker is in phase with the resonance (X:V) it acts as an open flap, in which case the rest of the port, X', takes no part in the resonance, and the unit tries to resonate at the frequency determined by X and V. As resonance starts to build up, the resonant Amplitude increases beyond that of the unit driving it. The excess air motion can not pass through the cone and therefore it osciilates along port X'. This automatically Lowers the Resonant Frequency until there is sufficient Phase Lag for the speaker driving frequency to be less efficiently coupled to the resonator. The amplitude then falls off, and the phase lag reduces.This phase lag/transient amplitude relationship is very important, as it both helps to maintain an efficient power transfer from the amplifier to the room, and also produces a musical sounding bass register, as it functions in a manner similar to the acoustics of a large auditorium rather than like the resonance of a small box, when fed with transient information.

If the driving frequency lies anywhere between the resonance (V:X) and (V:X+X'), the box will automatically 'Lock on' to the driving frequency in the above manner, similar to the 'Phase Lock Loop' principle now widely used in F.M. radio reception, producing a slight phase lag as a consequence of its mode of operation.

The system can now be seen to act in a similar way to a tuned resonator, but with the very important difference that the resonant Frequency is Signal Dependent, in such a way as to produce a slight Phase Lag while the Resonant frequency Tracks the Driving frequency within the low frequency pass band, thus maintaining a high electrical input to sound output conversion efficiency. Furthermore, because of its very high efficiency and its substantial independence of cone mass, it will deliver very well into very low acoustic impedances, without giving a peak in output at high acoustic impedances, thus making its output much less affected by the listening room resonances and If. cutoff than any conventional system.Its transient performance, too, is a useful feature, as it to a large extent tends to compensate for the transient effects introduced by the listening room, see diagram E.

In order to make the system even more efficient, and also to gain a further advantage in the directional radiation pattern (Polar diagram) at upper bass frequencies, I have carried out a number of experiments aimed at improving the 'Velocity Coupling' between the drive unit and the air moving in the port X--X'. The system does not in fact have to be entirely 'Port-loaded', either port X or port X', or even both X and X', can be replaced by other types of mass loading ... tapered port, (or horn in its extreme form). Auxiliary Bass Radiator, slot radiator, or even (a) further powered units could be used. The only criterion is that the resonance of the system should 'track' the drive signal frequency, with a slight phase lag which increases with resonance amplitude.

When other types of mass loading are used, the drive unit can be conveniently mounted in a very small enclosure between the two masses (equivalent to X and X') see diagram F for various alternative arrangements.

The technique which has so far proved to be most spectacular uses tapered ports, combined with a means of increasing the air velocity emanating from the drive unit. These two factors combined produce very strong Velocity Coupling between the drive unit and the ports, which in turn produce an extremely efficient device with an extra-ordinarily low bass cutoff frequency, and extremely low harmonic distortion. However, the units using slot radiators and/or aux. bass radiators instead of ports are also extremely effective, and have the added advantage of being more compact, and also of allowing less unwanted mid-frequency noises to escape through the A.B.R's or slots than would be the case with ports alone. (Note: The term Aux. Bass Radiators is the name usually given to a solid mass, such as a speaker cone, which is compliantly suspended but not driven by any electrical signal.This is allowed to flap about, hence its other common name of Flapping Baffle or Drone Cone. It acts in the same way as the mass of air in a port, resonating at a particular frequency in conjunction with an enclosed air volume which acts as a spring).

As the theory of operation is the same for all these systems, and has already been described above, I will go on to describe actual, practical cabinet designs, starting with the one using tapered ports and air velocity enhancement to achieve extra strong Velocity Coupling, mentioned previously. Consider diagram G.

Air from the front of the speaker cone is driven at high velocity through hole Y. This hole Y is about 1/4 of the dia. of the drive unit cone, and is in a sheet of very thin, rigid material such as Formica or duralumin. This sheet is bonded to the panel on which the speaker is mounted, which is itself a 1" sheet of marine plywood, with a hole in it large enough to give the speaker cone completely free travel in the forward direction.

The thin, rigid panel increases efficiency by controlling airflow and reducing the air friction.

The high velocity airstream from hole Y induces a similar movement in the air in port X, which communicates the movement, as described previously, to the air in ports X' in the proportions required to maintain the slight phase lag desired.

The taper in ports X' helps avoid upper bass standing waves in the ports; it also allows an increase in airstream velocity in the vicinity of the speaker hole Y and the port X, this providing closer coupling between drive unit and box. Port X is also tapered to allow greater freedom of movement of the airstreams, and closer coupling between driving and driven streams of air.

Port X is commonly about 1/3 to 1/2 of the driver cone diameter, and chamfered at about 45 degrees, as shown.

For listening room acoustic compensation, the interior volume of the main enclosure is chosen (from readily available data on the resonant frequency of Helmholtz resonators) to resonate with a port area X at a frequency of about 1 zF to 2 times the lowest frequency resonance of the listening room (or a frequency whose wavelength is about twice the length of the room's longest diagonal). Ports X' are then chosen to be as long as possible consistent with having as few bends as practicable, only a minimum of taper, and to produce a pleasant shape for the finished product.

If ports X' are opened to the air at some distance either side of the cone, it will be found that the upper bass frequencies are made directional, and in particular do not radiate well sideways, this is a very great advantage, as it means that far less upper bass is reflected from the listening room walls to colour the sound. It also has the advantage that the speaker can be positioned to avoid driving some of the room resonances/standing waves at all (practicaily speaking). All interior surfaces of the ports must be very smooth to prevent drag on the airflow, the box itself should be made Pressure Tight and the walls should be incapable of flexure . . . 1 " marine plywood is recommended. Any bends in the ports should be filleted to make a smooth transition.

Total cross-sectional area of ports X' where they join port X and hole Y, should be equal, or only just larger than, the area of port X.

A prototype of the above cabinet has been built with a 15" driver, and reproduces pure, distortion free bass down to below seven cycles per second with minimal loss of amplitude or efficiency, and in fact it is as efficient at these frequencies as a normal high efficiency full range speaker is in the upper registers. The lower limit is not limited by theory. If it were found that a lower cutoff frequency were desirable, this could easily be achieved with the same techniques.

One additional advantage of the small hole Y is that it helps lower the fundamental resonant frequency of the drive unit when mounted in position. It must however be pointed out that this design is only intended to work properly below about 120 Hz. or so in the form described, and needs a special crossover network to prevent frequencies above this figure from reaching the drive unit, otherwise considerable unevenness would result.

As these extremely low frequencies can be dangerous if maintained for long periods at excessive volume levels, an indicator (Visual) is desirable to show the operator when this is taking place. (The output from the speaker being too pure to be detected by hearing alone if the frequency is below about 18 Hz.). Such an indicator has been deveioped, derives its operating voltage from the signal, is of relatively high impedance so as not to degrade the signal, and in fact operates an LED (Light Emitting Diode) via a low pass filter, a rectifying circuit, and an 'integrating' R/C network.The final supply to the LED is current limited by a series resistance, the LED is reverse voltage protected by another diode (which could even be a zener diode for added forward protection and more constant light output) and the integrating network has a further diode in its supply to improve the visual effect by delaying the turn off time when once the LED has been lit. A variable potentiometer is also useful to control the initial 'turn on' sensitivity of the LED.

The circuit described is quite sophisticated, and all that is readily necessary is a current limited, reverse voltage protected LED with its operating voltage suitably adjusted, running from a high impedance circuit connected to the Infra Bass drive unit terminals. However, the added complexity is considered to be worth the trouble, as it gives a much more acceptable visual display, and under 'normal' circumstances remains out (whereas by comparison, the simplified circuit keeps flashing on and off with the signal).

A practical design based on Velocity coupling using Auxilliary Bass Radiators is shown in diagram H. Exactly the same reasoning applies. A design of more conventional appearance using less efficient Velocity coupling and Slot Radiators is given in diagram J. Again the same description of the mode of functioning applies, although in this case no means of increasing the velocities of the various airstreams is used.

Due to its highly efficient, signal dependent, 'tracking' tuned resonance, the 'VelocityCoupled Infra-Bass' speaker doesn't need high acoustic impedances to operate efficiently. In fact its efficiency is affected much less, by the changes in acoustic impedance which accompany listening room resonances, than practically all conventional types of speaker. Hence, the bass register sounds much more smooth and natural, uncoloured by the listening room acoustics.

There are three important innovations which the Velocity Coupled Infra-Bass speakers described make use of.

1. Already mentioned, the ability to produce pure undistorted infra bass frequencies at high efficiency and high volume levels when required.

This has not previously been seen as an advantage below 20 Hz., and no other known system can do this properly even at this frequency. A prototype Velocity-Coupled Infra Bass cabinet, fitted with a 15" driver has been built and tested to below 7 Hz., as already stated.

2. The second innovation is the avoidance of audible T.P.D. at the extreme bottom end of the audible spectrum. T.P.D. or transient phase distortion, is a new term which I have coined and refers to the shift in frequency of transient signals which occurs when the transients are passed through a circuit which introduces either a phase lead or a phase lag around that frequency. Almost universal is the phase lead produced by amplifiers and transducers at the lower end of the frequency range, caused by the limited bandwidth which all these devices possess. This will automatically raise (by up to 3 octaves or more in extreme cases) the frequencies of transients within this range. Some of these transients may be at frequencies below the audible limit at that volume level, and these become not only audible, but discordant. Others, already audible, would just become discordant and louder.With the low cutoff frequencies normally employed, however, the frequencies involved would be so low that only the former effect would be significant. This new discovery goes a long way towards explaining why all Hi-Fi systems up until now could not produce convincing spaciousness, and has not previously been understood.

The Velocity-Coupled Infra-Bass principle has been designed to produce a phase lag at the bottom end of the sound/vibration spectrum. This will not only help cancel the effects of L.F. phase lead elsewhere in the system (and incidentally in the listening room acoustics due to L.F. cutoff) but will also ensure that any remaining T.P.D. effects remain inaudible as their frequencies will be shifted downwards and therefore will remain outside the audible bandwidth. This deliberate introduction of phase lag at the low end of the bandwidth is therefore very important and I am looking for particular protection on this point. The 'signal-dependent control' of the cabinet's resonant frequency is one particulariy good means of achieving this end, but there are certain to be other electronic analogues which produce the same effect.

3. In all cases where a small enclosure is required to produce an alternating movement of air(/gas) and where the total displacement is a fairly large proportion of the total enclosure volume, the resultant output will be found to oscillate about a mean which is not the same for different output levels, and in particular is different from the 'rest' or no signal condition.

This is because the pressure inside the enclosure is raised by a greater percentage by a given movement inwards than it is lowered by an exact equivalent movement outwards. The mean about which oscillation takes place therefore moves outwards as a direct result, and to a greater extent the larger the driving signal. This phenomenon is only really significant at very low frequencies, as only here are sufficient quantities of air required to be moved. Speakers required to operate at high volume levels below 100 Hz., and to give very low distortion at these frequencies need some form of correction for the effect, and the Velocity-Coupled Infra-Bass cabinet does this by utilising the way in which the air from the driver is coupled to the air in the port X.Coupling is stronger in the 'inward' direction than in the 'outward' one, and this ensures that the mean position remains substantially constant. If any remaining movement of the mean position occurs it is better that this should be in the 'inward' direction, as this does not represent the sound of an enclosed space, to the ear, and is therefore psychoacoustically more acceptable.

Compensation for this phenomenon is also never considered by speaker manufacturers, probably because their products do not possess such an extended L.F. response, and consequently it has never occurred to them. As the Velocity-Coupled Infra-Bass speaker is the first device to take this effect into account, I am also looking for protection on this point. Again, although Velocity coupling is a good way of achieving the desired result, the same effect can be produced by a fairly simple electronic circuit involving an R.C. filter and a rectifying element, as I have proved recently. (Protection required).

I am, of course, also seeking protection for the Velocity-Coupling principle itself, and any speaker/cabinet employing this principle.

In addition I am seeking protection for the particular designs described, and the details therein (for example the use of twin tapered ports X', tapered port X., velocity enhancement device consisting of port Y. in thin, rigid sub-baffle fixed to main, thick drive unit mounting baffle. Hole in latter large enough to give full clearance to the cone movement. Likewise the use of auxiliary bass radiators, slots, etc. as described, and the L.E.D. indicator device in any form. Also the deliberate use of inductors capacitors rectifying devices etc. in an attempt to produce the 'Signal dependent' tuned resonance effect, the 'Phase lagging' effect or the 'Stabilising of the mean centre of oscillation' mentioned in 3 above).

This invention makes use of conventional loudspeaker drive units, which can be regarded as part of the prior art. The improvement comprises the manner in which they are used, the type and design of associated cabinetry and equipment required to use them in the manner described, in the numbered claims following, and any modifications carried out on the drive units themselves in order to fulfill the claims as listed.

Claims (20)

Claims

1. The technique of matching a loudspeaker to its acoustic load, to give its most efficient ouput in the configuration used when the environmental acoustic impedance is extraordinarily low, at any frequency or within any frequency range.

2. As in claim 1, by causing the drive unit/s to be coupled to air movement in its/their local acoustic environment, such as a 'listening room', by means such as viscous drag, vortex action, turbulence.

3. As in claim 1, by causing the drive unit/s, at low frequencies, to be coupled to air movement into and out of a cabinet, of any design, construction or type, by means such as those in 2.

4. Claim 1, achieved by the use of electromechanical feedback on the drive unit itself, any auxilliary mass loading used, or both.

5. As in claim 3, using one or more 'resonant box and port/s' configurations.

6. As in claim 5, where the resonant frequency/ies claimed is/are made signaldependent by use of (a) suitable driver/port/s configuration/s.

7. Any speaker claimed in any claim where the effective velocity of the air moved directly by any drive unit is increased in any way, for example, by the use of a constricting aperture or nozzle.

8. Any speaker claimed, where the ss shape/s of port/s or aperture/s are especially contrived to increase the coupling between the air moved by a drive unit and the airstream producing (indirectly or directly) pressure changes in the environment.

9. Any speaker claimed, using any other forms of mass loading, such as 'flapping baffles', in place of any of the moving air masses or ports described or implied.

10. Any combinations of claims 1-9.

11. Any speaker using one or more bass/infrabass drive units in the manner described in any claim relevant to the production/reproduction of low frequencies, in such a way as to make the acoustic output phase locked to the input signal; for example, by making the output phase lag the signal by an amount which is dependent on the amplitude of the output airstream velocity, for any given strength of input signal.

1 2. Any speaker design or construction, substantially as claimed in any claim, but using more than one drive unit, up to the use of multiple units.

13. The use of a visual indicator, such as a 'signal-derived' L.E.D. display, to indicate the presence of frequencies at the very low frequency end of the acoustic spectrum, which these speakers can produce at very high, but inaudible levels, and which might be dangerous in excess.

14. Any speaker fitted with a display as claimed in 13.

1 5. Any speaker claimed, using in addition another drive unit or units, and associated circuitry to enable its usable frequency range to be extended.

1 6. The manufacture and/or use of any speaker claimed for compensation of the bass/infra-bass acoustics of a room, by itself, or in conjunction with any other speaker, speakers, or additional drive units.

17. The manufacture and/or use of any speaker claimed, by itself or in conjunction with any other speaker, speakers, or additional drive units, for:- A. Music reproduction, such as for example, in Hi-Fi systems.

B. 'Live' music or sound production, such as for example, with electric guitar, organ, synthesizer.

C. Any other purpose, such as in Research, Industry, or for Medical or Scientific purposes.

New claims or amendments to claims filed on 28th Jun 1979.

Superseded claims 1, 13, 14.

New or Amended Claims:-

1. The technique of coupling a loudspeaker to its acoustic load to produce a large air velocity component of any frequency or frequencies within its range, while at the same time producing the smallest possible related pressure component, in the air immediately adjacent to the device (corresponding to a hypothetical 'zero acoustic impedance' condition).

13. Any speaker claimed in any claim, which uses a visual indicator, such as a 'signal-derived' L.E.D. display to indicate the presence of frequencies at the very low frequency end of the acoustic spectrum (for example, say, below 25 Hz, although the actual figure chosen will depend on the context) which these speakers can produce at very high, but inaudible levels, and which might prove dangerous in excess.

14. Any speaker claimed in any claim, fitted with a display as claimed in claim 13.

18. The manufacture or use of any sound system incorporating one or more speakers substantially as claimed in any claim, by themselves, or in addition to other more conventional speakers or speaker units, this including systems of one channel or more than one channel, where the number of claimed speakers is not necessarily equal to the number of channels supplied or in use.

1 9. Any speaker claimed in any claim, fitted with any non-standard modification, such as more than one voice coil, to more easily facilitate the latter part of claim 18, for example, (or to assist in carrying out claim 4).

20. Any speaker, substantially as claimed in any claim, where the design is modified in some way, for example by introducing a bend or bends into the ports, introducting extra means of guiding the airflow smoothly, changing cross-sectional areas, shapes, flare or taper rates, changing cabinet interior or exterior sizes or proportions, using different materials for construction, or in any other manner, for whatever purpose.