AU707281B2 - Loudspeaker system incorporating acoustic waveguide filters and method of construction - Google Patents

Description

-1-

AUSTRALIA

Patents Act COMPLETE SPECIFICATION

(ORIGINAL)

Application Number: Class Int. Class Lodged: Complete Specification Lodged: Accepted: Published: Priority Related Art: APPLICANT'S REF: COMPLETE-ASSOCIATED- WITH-PROVISIONALS PN 0290 and PN 1457 Applicant(s): GRAEME JOHN HUON, GREGORY KEITH CAMBRELL and WALTER MELVILLE DOWER 39 Monomeith Crescent, Mt. Waverley, Victoria 3149, Australia; 62 Lohr Avenue, Inverloch, Victoria 3996, Australia; and 19 Summerlea Road, Mt. Dandenong, Victoria 3767, Australia, respectivley S Actual Inventor(s): GRAEME JOHN HUON, GREGORY KEITH CAMBRELL and WALTER MELVILLE DOWER Address for Service is: PHILLIPS ORMONDE FITZPATRICK Patent and Trade Mark Attorneys 367 Collins Street Melbourne 3000 AUSTRALIA Complete Specification for the invention entitled: "LOUDSPEAKER SYSTEM INCORPORATING ACOUSTIC WAVEGUIDE FILTERS AND METHOD OF CONSTRUCTION" Our Ref: IRN 393704 The following statement is a full description of this invention, including the best method of performing it known to applicant(s): INN C:\WINWORD\JENNY\WORK\CAPFORMS.DOC 16 LOUDSPEAKER SYSTEM INCORPORATING ACOUSTIC WAVEGUIDE FILTERS AND METHOD OF CONSTRUCTION The present invention relates to a loudspeaker system incorporating one or more acoustic filters, in particular waveguide filters, and to a method of construction thereof.

The term waveguide originates from general Electromagnetic wave theory wherein it describes the general form of a bounded region for guided propagation of waves. This general electromagnetic case and the acoustic case considered herein are analogous.

The present invention is restricted to the dominant mode of propagation of sound waves'and therefore the filter sections could alternately be referred to as transmission line sections, as the mode of propagation obeys frequency independent phase velocity (analogous to Transverse Electro Magnetic (TEM) modes in waveguides).

However, the term transmission line in audio parlance has become associated with particular enclosure designs wherein heavy loss is introduced for one side of the transducer. The common use of the term transmission line is thus no longer understood to relate to supporting a mode of propagation over the 20 frequency range as is the case with the electromagnetic domain. Hence the term acoustic waveguide is used herein in preference.

There 'are many applications for acoustic filters and the techniques described herein are relevant to many of them. Accordingly, although the present S:invention is described herein in relation to and in the context of loudspeaker S 25 systems, it is to be appreciated that it is not thereby limited to such applications.

Loudspeaker systems include a combination of one or more electroacoustic transducers or drivers together with a partial or complete enclosure (referred to herein as an enclosure) and may be categorized by the nature of their frequency response when driven from a constant amplitude sinusoidal voltage source. Closed or sealed enclosures, also known as infinite baffle or acoustic suspension enclosures, provide high pass systems with second order roll off at relatively low frequencies. Vented enclosures, also known as bass reflex or MJP C:\WINWORD\MARFE\GABNODEL\ACOUSTI3.DOC -2- Helmholtz resonator enclosures, provide high pass systems with fourth order roll off. Conventional closed/vented or double vented enclosures provide bandpass systems having an additional second order roll off at relatively high frequencies.

Variants of these basic enclosures have been proposed which use additional chambers and/or additional vents between chambers and/or folded ducts exiting internally or externally.

Many attempts have been made over the last seventy years to construct low frequency loudspeaker systems with optimum performance subject to various physical constraints. The most sophisticated prior art loudspeaker designs are based on techniques which utilize lumped component equivalent circuits. All of the categories of enclosures described above can reasonably be modelled with lumped equivalent circuits. However, these models are restricted in usefulness to special cases. One known design disclosed by Schreiber in US Patent 5092424 incorporates a multi-chamber enclosure. However, the latter system still only 15 employs one resonance per chamber and has attendant space utilization problems.

A disadvantage of prior art lumped equivalent circuit modelling techniques is that they fail to assist designers to recognize or exploit the benefits of undamped waveguides or waveguide sections in the design of acoustic filters.

20 An object of the present invention is to provide an improved method of constructing a loudspeaker system. The method of the present invention includes a relatively sophisticated modelling technique which may assist designers to recognize and exploit a more diverse range of acoustic elements including elements having predominantly distributed natures.

25 A further object of the present invention is to provide an improved loudspeaker system and in one embodiment a loudspeaker system having improved acoustical response for reproducing frequencies at the low end of the audible/infrasonic spectrum.

It has long been understood that selected multiple resonances are a means of improving performance in loudspeakers. The problem has been that the obvious means of providing multiple resonances namely by adding resonators has consequent cost, space and complexity problems and in any case provides MJP C:\WINWORD\AARIE\GABNODELACOUSTI3.DOC -3only one beneficial resonance per resonator in the pass band, this resonance being the fundamental resonance for that resonator.

The present invention addresses these limitations by utilizing as a filter at least one acoustic waveguide or waveguide section. A waveguide or waveguide section may be arranged to provide a plurality of resonances. Significantly the resonances may be non-fundamental resonances. The resonances also may be substantially undamped, ie. they may not require any damping. The present invention also provides a technique for beneficially selecting and methodically placing such resonances to shape the response of the filter. The filter may be adapted to improve flexibility in alignment of response characteristics of a loudspeaker system and/or to enhance performance of the system. The filter may be integrally build into the loudspeaker system or otherwise attached thereto.

Whilst it is not possible to say with absolute certainty that prior art loudspeaker enclosure designs (especially those which were experimentally derived) have not accidentally incorporated a non-fundamental resonance in one of its air paths for a beneficial result, applicant is not aware of any prior art enclosure designs that have consistently incorporated substantially undamped non-fundamental resonances in a methodic way or that have optimally placed a single non-fundamental resonance to maximum benefit. To applicants best 20 knowledge, in no prior art enclosure designs has placement of non-fundamental l resonances been a cause of improved performance over conventional methods and no designer has to date deliberately designed non-fundamental resonances into enclosures.

As indicated above a variant of prior art designs has been the addition of 25 folded ducts, sometimes called "transmission lines" or "quarter wave labyrinths".

Although the latter terms are borrowed from transmission line theory they are not the result of any generalised modelling technique as is the case in the present invention and do not incorporate the advantages offered by the present invention.

On the contrary prior art transmission lines are designed to be deliberately highly absorptive, lossy and of low efficiency. Any non-fundamental resonances which may exist are so heavily damped that labyrinthine ducts are regarded as one of the least resonant enclosures known. In contrast an important feature of the MJP C:\WINWORD\MARIE\GABNODELACOUSTI3.DOC -4present invention in one embodiment is the use of waveguide filter sections which are substantially reactive, undamped, low loss and therefore high efficiency.

Conventional acoustic horns have also been used as filters as they provide certain advantages over other designs in controlling the behaviour of drivers.

Finite acoustic horns represent a monotonically increasing cross section filter, which limits their usefulness for low frequency applications where impractical dimensions are required for effective horn loaded operation. Some instances have achieved limited compaction by the use of folded horn configurations but size is still a limitation. The present invention provides an alternative to conventional horn design with potentially improved and more compact design characteristics by the use of waveguide filter sections.

The present invention provides a new and fundamentally more accurate method for modelling and constructing acoustic filters. The method of the present invention differs from previous modelling methods in that it allows recognition of filter elements that are predominantly distributed in nature and their beneficial incorporation into the design. This may be achieved by treating appropriate components as waveguide filter sections defined by characteristic impedance and length and analysing the components accordingly.

The above approach may provide a suffficiently accurate analysis to allow controlled utilization of non-fundamental resonances in the waveguide filter sections. The approach may also allow partial or complete elimination of damping material in the waveguide filter sections. Partial or complete elimination of damping material is desirable because it facilitates creation of relatively efficient filter designs. The above approach also inherently creates new degrees of freedom in the modelling/design process.

Utilisation of substantially undamped multi-resonant waveguide sections has advantages in facilitating adjustment of response characteristics and minimising driver cone excursion. The latter has the effect of reducing distortion.

The former has the effect of extending the frequency range of the filter.

A loudspeaker system is essentially a combination of one or more acoustic filters and one or more electro-acoustic drivers. The filters and drivers may be Q MJP C:\WINWORD\MARIE\GABNODEL\ACOUST13.DOC iy designed around an enclosure to produce a specified acoustic output over a defined frequency range when operating into a specified acoustic environment.

The or each acoustic filter according to the present invention may include one or more waveguide filter sections. Where a plurality of filter sections are used, these may be arranged in cascade or in parallel or in combinations of cascade and parallel or more complex configurations.

The initial step in the construction of a loudspeaker system according to the present invention is the specification of performance criteria and design objectives. The initial specification for a system may include: a. frequency range of the pass band with desired tolerance on amplitude and phase response; b. asymptotic amplitude roll-off slopes above and below the pass band (thereby also specifying asymptotic phase response); c. maximum design figure for rated acoustic power output into a specified acoustic environment (for example 27 steradians); d. overall desired maximum physical size for the enclosure; e. desired distortion performance; and f. desired efficiency.

This information may indicate a preferred choice for an enclosure 20 configuration including the combination of an open or closed chamber (acoustic S• waveguide) on one or both sides of one or more electro-acoustic drivers. It may *also indicate approximate dimensions.

Like mechanical systems, acoustical systems may be more easily modelled as equivalent electrical analogues to facilitate an optimization process.

25 The equivalent electrical analogues may be obtained by comparing differential equations of motion for both systems. The acoustical and electrical systems are o. considered analogous if their differential equations of motion are mathematically the same. Corresponding terms in the differential equations of motion are analogous to each other. Equivalent electrical circuits may then be created and analysed using standard circuit analysis techniques.

The next step in the construction of a loudspeaker system according to the present invention is the design of an initial acoustic waveguide with resonance at MJPC:\W'NWORD\ARIE\GABNODEL\ACOUST3.DOC T -6the lowest desired frequency to be reproduced. This design may create an acoustic filter utilising multiple resonances obtained from each acoustic waveguide section. These resonances may provide a series of minima in the driver diaphragm displacement(xD) and in the Enclosure Characteristic Curve of Driver Displacement (ECCDD). For an enclosure having a lumped vented chamber on one side of a driver and a chamber and single acoustic waveguide of length d and characteristic impedance Zo on the other side of the driver, it may be shown that a series of minima in the driver diaphragm displacement will be produced in accordance with the formula: xD BI 1 V 2- XC VG S Z E v C/F ZAD ZAF where

C

1 sin *kd 1+ O i o- 20 [l+C cos-M+ ca CB r eNote: The various terms used in the formulas herein are set out in the Appendix.

The resultant driver diaphragm displacement will contain a defined extended series of maxima and minima. Those resultant peaks and dips in the 25 displacement (and ECCDD) may not necessarily correspond to peaks or dips in the acoustic output from the enclosure, but may assist in controlling the response.

Each identifiable acoustic waveguide will result in similar behaviour. When acoustic waveguides are combined in cascade or in parallel, each section may produce such behaviour, and the overall combinations of sections may exhibit a multiplicity of resonances.

If, for example, a single acoustic waveguide is subdivided into two waveguides with equal overall length or delay but two differing characteristic MJP C.WVNWORD\ARIE\GABNOEL\ACOUST13.DOC -7impedances, the series of resonances of the original overall section will be modified into two sets of resonances corresponding to the two new sections. The lowest frequency resonances will still be related to the overall length of the two combined waveguides. The resonances so formed will behave in accordance with the values of delay and characteristic impedance attributable to each filter section and combinations thereof in accordance with the above equation. Adding acoustic waveguides by division of the original acoustic waveguide will add sets of multiple resonances.

Since the behaviour of the enclosure at the lowest frequency of the pass band is predominantly governed by the overall dimensions of the acoustic waveguide filters specified in the initial specification, the shape of each chamber may be chosen within the constraint of the overall maximum physical size of the enclosure, to provide the first ECCDD minimum near the lowest frequency of the pass band.

15 The acoustic waveguide section may be modelled as having the overall length d, which provides a one way propagation delay kd 1 d, c and a nominal characteristic impedance:

S

PoC 2 1 25 St where St can be regarded as the ratio of the volume V of the acoustic waveguide to its overall length, d. The filter may be implemented as two or more acoustic waveguide sections in parallel or as a single set of cascaded acoustic waveguide sections.

The design approach may enable synthesis of the required response by utilising multiple acoustic waveguide sections, by choosing where their MJP C:\WINWORD\MARIE\GABNODEL\ACOUSTI3.DOC -8resonances are placed and by manipulating their values of delay and characteristic impedance.

Single uniform sections of acoustic waveguide may provide multiple higher order minima of ECCDD. The next step in the construction of a loudspeaker system according to the present invention is to partition the initial acoustic waveguide into waveguide sections, either as cascades or branches or combinations of both, with the intent of: 1. providing maximum benefit from multiple resonances and associated ECCDD minima whilst; 2. enabling tailoring of the resultant response to meet the specifications for amplitude and phase; and 3. preserving the ECCDD minima derived from the initial single or parallel acoustic waveguide section design.

The overall modelling approach may be iterative and based on analytical or 15 numerically simulated solution of an equivalent electrical circuit of the o driver/enclosure/external load combination. The topology of the enclosure may be modelled as one or more inter-connected acoustic waveguide sections in conjunction with the electro-acoustic driver(s) and the appropriate resulting air load for the enclosure.

i' 20 The interconnection of acoustic waveguides of differing characteristic impedances creates a discontinuity at the interface and may require inclusion of circuit simulation elements to represent the discontinuity. Theoretical and experimental studies have shown that for two acoustic waveguides of differing cross-sectional areas ST1 and ST2 at the discontinuity, the lumped acoustic S discontinuity masses M determined empirically from: MC =0.26po- 1-1.4 +0.4 for S 2 The resulting network can then be analysed by using standard circuit analysis techniques such as node (Kirchhoff Current Law) or loop (Kirchhoff Voltage Law) analysis. However, the circuit analysis technique chosen should be MJP C:\WINWORD\MARIE\GA8NODEL\COUSTI3.DCC -9capable of incorporating distributed structures. A computer-based circuit simulation package such as SPICE may be used for this purpose. Each acoustic waveguide section may be simulated as a lossless transmission line section characterised by Zo and The parameters of the acoustic waveguide filter sections may be varied in an iterative manner from coarsely chosen initial to target values and the resulting effect on the acoustic output and ECCDD across the pass band calculated, until a specified response is obtained.

A closed-form optimisation of the response can also be undertaken but will not be described here.

Once the desired characteristic impedance and one-way propagation delay for each acoustic waveguide section are defined, the acoustic waveguide sections can be given physical dimensions. The area (St) of the acoustic waveguide section is given by: St (p 0 c)/Zo for each section Each acoustic waveguide section can then be specified in terms of length (corresponding to time delay) and area.

The construction process may also consider other factors such as practical limitations of electro-acoustic driver construction and peak air particle velocity in 20 each acoustic waveguide section to minimise distortion and noise resulting from air turbulence.

The method of construction according to the present invention addresses the limitations of lumped element models and allows a designer to take advantage of the distributed nature of chambers, ducts, vents and other 25 structures found in acoustic filters and loudspeaker enclosures. In summary the construction method may include the steps of: 1. selection of an enclosure and acoustic filter configuration to suit the desired design criteria; 2. creation of an equivalent circuit model incorporating the equivalent circuit of the driver and all filter elements, including discontinuities; MJP C:\WINWORD\MARIE\GABNODEL\ACOUSTI3.DOC (4 ~ii ~C 3. representing the filter elements as waveguide or transmission line sections of desired characteristic impedance and length or lumped components according to design requirements; 4. analysing the model using any suitable analysis technique or computer simulation package for electrical circuits (the analysis will show the frequencies where resonances are located); fine tuning the model to provide the desired results, paying particular attention to beneficial placement of resonances including non-fundamental resonances; and 6. converting the electrical analogy to acoustic components within the size and shape constraints of the design criteria.

Loudspeaker systems covering a wide variety of embodiments may be constructed according to present invention, including embodiments having: any enclosure design, whether conventional or not, having one or more waveguide acoustic filters built into the system or attached thereto by any suitable means; any enclosure having waveguide acoustic filters coupled to one or both sides of the speaker driver, or to one or more chambers, vents, ducts, cavities, passive radiators or other sound generating or transmitting 20 structures forming part of the enclosure, or any combinations thereof; and any enclosure having any number of waveguide acoustic filters arranged in any branching combination including cascade and/or parallel configurations.

t z #i MJP C:,WINWORD\MARIE\GABNODELACOUST13.DOC T 10a According to one aspect of the present invention there is provided an improved acoustic filter for a loudspeaker system, said system having an enclosure in combination with at least one electroacoustic driver and said system having a first response, said acoustic filter being incorporated with said enclosure and having a second response so as to produce said first response from said loudspeaker system, said acoustic filter including: a first acoustic waveguide section having a first characteristic impedance and at least a second acoustic waveguide section having a second characteristic impedance, each acoustic waveguide section having an impedance discontinuity at each end thereof and being connected in a configuration in which said acoustic waveguide sections interact with one another; said configuration giving rise to resonances based on resonant paths between pairs of impedance discontinuities in said configuration, said resonant paths being tuned and said resonances being excited to produce said second response; wherein said resonant paths are tuned and quantities, frequencies and excitations of said resonances are adjusted by modelling the first and the at least second acoustic waveguide sections so that the sections are represented in the model as distributed elements, and by optimizing the model to produce an 20 alignment of said resonances including non-fundamental and out of band resonances which gives rise to said second response.

According to a further aspect of the present invention there is provided a method of constructing an improved acoustic filter for a loudspeaker system, said acoustic filter including a first acoustic waveguide section having a first 25 characteristic impedance and at least a second acoustic waveguide section having a second characteristic impedance, connected to and arranged to interact •*:with said first acoustic waveguide section, said system having an enclosure in combination with at least one electroacoustic driver and said system having a first response, said acoustic filter being incorporated with said enclosure and having a second response so as to produce said first response from said loudspeaker system, said method including: MJP C:\WINWORD\MARIE\GABNODELAACOUSTI3.DOC modelling the first and the at least second acoustic waveguide sections such that the sections are represented in the model as distributed elements wherein said modelling includes simulating acoustical elements of said system including each waveguide section by means of an equivalent electrical circuit wherein the acoustical elements including said distributed elements are represented in the circuit via electrical components; and optimizing the model to produce an alignment of resonances including nonfundamental and out of band resonances based on resonant path between pairs of impedance discontinuities in said filter and, wherein said optimizing includes analysing said circuit and changing values of said components from roughly chosen initial values until said resonant paths are tuned and quantities, frequencies and excitations of said resonances are adjusted to simulate said second response by target values of said components.

Preferred embodiments of the present invention will now be described with reference to the accompanying drawings wherein: Figure 1 shows a cross-sectional view of a prior art sealed enclosure loudspeaker system; Figure 2 shows a cross-sectional view of a prior art vented enclosure loudspeaker system; 20 Figure 3 shows a cross-sectional view of a prior art double-vented enclosure loudspeaker system; 4" 9*

I.

<NTO*

-11 Figure 4 shows a cross-sectional view of a loudspeaker system according to one embodiment of the present invention; Figure 5 shows a cross-sectional view of a loudspeaker system according to another embodiment of the present invention; Figure 6 shows a cross-sectional view of a loudspeaker system according to a further embodiment of the present invention; Figure 7 shows in schematic form, an equivalent acoustical circuit for the loudspeaker system of Fig. 4; Figure 8 shows a graphical representation of typical ECCDD values (normalised driver displacement) for the loudspeaker system of Fig. 4; Figure 9 shows a perspective view of a six section acoustic waveguide filter according to a preferred embodiment of the present invention; Figure 10 shows a perspective view of a loudspeaker system including a single vented enclosure and incorporating an acoustic filter as shown in Fig. 9; 15 Figure 11 shows in schematic form, an equivalent electrical circuit for the ooo i loudspeaker system shown in Fig.10; and Figure 12 shows a graphical representation of comparative frequency response for the loudspeaker systems shown in Figures 1 to 4.

Referring to the drawings, Figures 1, 2 and 3 show prior art sealed, vented 20 and double vented enclosures, respectively described at pages 1 and 2.

Figure 4 shows in cross section a loudspeaker system according to one embodiment of the present invention. The system of Figure 4 includes a single waveguide filter section attached to a vented enclosure 40. Enclosure 40 includes a baffle 41 dividing enclosure 40 into an interior chamber 42 and at least one duct S 25 43 which provides an acoustic waveguide filter section. Interior chamber 42 is vented to the exterior of enclosure 40 by at least vent 44 which terminates into exterior space 45 by way of duct 43. At least one electro-acoustic driver 46 is mounted on baffle 41. Chamber 42, duct 43 and vent 44 may be realized with various aspect ratios and by use of cleating, folding and/or bafffling techniques.

One wall of enclosure 40 such as base wall 47 of duct 43 may be formed by placing the enclosure on a flat surface such as a floor.

MJP C:\WINWORD\MARIE\GABNODEL\ACOUSTI3.DOC -12- Figure 5 shows a loudspeaker system according to another embodiment of the present invention being a pair of systems as in Figure 4 facing each other.

The design and operation of the latter embodiment is substantially identical to that described with reference to the embodiment of Figure 4, except that the air load is shared between the two systems resulting in greater output.

Figure 6 shows a loudspeaker system according to a further embodiment of the present invention. The Figure 6 embodiment is a modification of the embodiment shown in Figure 4 in which vent 60 terminates into the exterior space 61 directly. Since the position of open end 62 of vent 60 relative to mouth 63 of duct 64 is different from the embodiment of Figure 4, the radiation pattern of Figure 6 embodiment differs from that of Figure 4. In other respects, the design and operation of the Figure 6 embodiment is similar to that described with reference to the embodiment of Figure 4.

The loudspeaker system of Figure 4 can be represented by the equivalent acoustical circuit model of Figure 7 with the acoustic waveguide filter having a characteristic impedance and length shown as a two-port network The equivalent circuit of Fig. 7 can be analysed to show amplitude response, phase response, delay response and other performance criteria. Values of impedance and length can be adjusted from coarsely chosen initial to target values to fine 20 tune the model for optimum performance according to design criteria.

When the desired result is achieved, the model may be realized in an acoustic domain by giving the target values physical filter dimensions. Other factors such as peak volume velocity may be taken into account at this stage. If design criteria dictate or if further improvements to performance are required, a 25 more complex waveguide acoustic filter can be used.

When the circuit of Figure 7 is analysed for driver cone displacement with a particular set of values, the result is a curve of cone excursion minima as shown in Figure 8. In Figure 8 cone excursion minima 1 and 2 correspond to fundamental resonances of the vented chamber and waveguide filter section, whilst minima 3 and 4 correspond to non-fundamental resonances of the waveguide acoustical filter. The non-fundamental resonances are a consequence of the characteristic impedance and length of the waveguide section.

MJP C:\WINWORDMARIE\GABNODEL\ACOUST13.DOC -13- In Figure 8 curves A, B, and C represent the Enclosure Characteristic Curves of Driver Displacement (ECCDD) for prior art sealed, vented and doublevented enclosures respectively, for a given enclosure size. Curve D represents the ECCDD for an enclosure according to the present invention. It is to be appreciated that a lower displacement value indicates less driver displacement for a given level of acoustic output It may be seen that driver displacement for a given SPL for an enclosure according to the present invention (curve D) is considerably improved over the frequencies of interest when compared to the prior art sealed and vented enclosures (curves The present invention exceeds the SPL for a given driver displacement for the double-vented (curve C) prior art enclosure over much of the range of the frequencies of interest.

In a preferred embodiment, a six section waveguide acoustic filter as shown in Figure 9 is added to the model, the enclosure then having the configuration shown in Figure 10. Each section of the filter is labelled 1 to 6 15 respectively in Figure 9. To design with best accuracy, the entire enclosure is •treated as waveguide sections and an equivalent electrical circuit model is created.

The equivalent electrical circuit model is shown in Figure 11. In Figure 11 logo block 110 represents the driving source for the loudspeaker system such as an 20 amplifier and block 111 represents the electro-acoustic driver. Since the "diaphragm of the driver has front and rear surfaces, inductances LMAF, LMAR represent masses of air associated with the front and rear surfaces respectively of the driver diaphragm. Block 112 represents the six section acoustic waveguide filter of the loudspeaker enclosure. This is modelled with six two-port filter 25 sections TLF1 to TLF6 respectively. Each two- port filter section is defined by values of characteristic impedance and length. Block 113 represents the internal chamber and twin port tubes. The internal chamber is connected to the exterior of the enclosure by distributed masses of gas in the port tubes and is modelled with two two-port filter sections TLR 1 and TLR 2. The port tubes terminate into the exterior space by way of filter section TLF 6 in part, and this is modelled by connection of inductance LMAR 23 to output node 114 of filter section TLF 6.

Finally, block 115 represents the exterior air load. Initial values are selected in the MJP C:\WINWORD\MARIE\GABNODEL\ACOUSTI3.DOC

I

-14model and the response characteristics are determined by analysis followed by fine tuning to optimise performance. Analysis is preferably performed by means of a computer based circuit simulation package such as SPICE. When the desired results are achieved the target values defined in the model are converted to physical filter section dimensions.

Referring to Figure 12, curves A, B and C represent a specific comparative example of the frequency response of prior art sealed, vented and double vented enclosures respectively, with the same electro-acoustic driver being used in a common enclosure size in each case for comparative purposes. Curve D represents a specific example of the frequency response of a loudspeaker system according to the present invention, for the same enclosure size and electroacoustic driver.

Comparing the frequency responses, it can be seen that the present invention provides improved bass response over that of the sealed and vented 15 configurations (curves Comparing the frequency response of the present oolil invention with that of the double vented (curve C) prior art, it is apparent that overall bass response is extended at higher bass frequencies whilst the response is more linear with frequency.

Although for the specific comparative example shown in Figure 12 the 20 output from the prior art displays a peak output which exceeds that of the present S"invention, inspection of the corresponding ECCDD curve in Figure 8, reveals that driver displacement is significantly lower for the present invention over this range, facilitating higher acoustic output for a given level of driver displacement and/or induced distortion, than for the prior art.

25 The acoustic output of the loudspeaker system of the present invention can be adjusted quite flexibly by choice of parameters. For example, the waveguide filter can attenuate or increase output at the expense of linearity of response, or bandwidth can be adjusted for linearity of response at the expense of efficiency.

The present invention provides an improved method of construction of acoustic filters. The invention also provides a substantially reactive acoustic filter employing substantially undamped duct sections resulting from application of the MJP C:\WINWORD\MARIE\GABNODEL\ACOUST[3.DOC improved method of construction. The invention also provides an improved loudspeaker system resulting from incorporation of the improved filters. The improvements may include minimisation of cone excursion over a relatively wide bandwidth, consequential reduction in distortion, better filtering of harmonics, new degrees of freedom in the alignment of performance objectives, controlled delay response including the option of constant delay, the option of shorter delay than obtainable by prior art modelling, controlled amplitude and phase responses, and elimination of the need for lossy filtering. Consequentially a loudspeaker system according to the present invention may exhibit relatively high efficiency, good transient response, and a structure that is compact and cost effective in comparison to conventional designs.

Finally, it is to be understood that various alterations, modifications and/or additions may be introduced into the constructions and arrangements of parts previously described without departing from the spirit or ambit of the present 15 invention.

0 s *0 6 MJP C:\WINWORDXMARIE\GABNODELACCOUST3.DOC

Claims (44)

1. An improved acoustic filter for a loudspeaker system, said system having an enclosure in combination with at least one electroacoustic driver and said system having a first response, said acoustic filter being incorporated with said enclosure and having a second response so as to produce said first response from said loudspeaker system, said acoustic filter including: a first acoustic waveguide section having a first characteristic impedance and at least a second acoustic waveguide section having a second characteristic impedance, each acoustic waveguide section having an impedance discontinuity at each end thereof and being connected in a configuration in which said acoustic waveguide sections interact with one another; said configuration giving rise to resonances based on resonant paths between pairs of impedance discontinuities in said configuration, said resonant paths being tuned and said resonances being excited to produce said second response; wherein said resonant paths are tuned and quantities, frequencies and excitations of said resonances are adjusted by modelling the first and the at least second acoustic waveguide sections so that the sections are represented in the model as distributed elements, and by optimizing the model to produce an alignment of said resonances including non-fundamental and out of band resonances which gives rise to said second response. 4* 2

2. An improved acoustic filter according to Claim 1, wherein said modelling includes simulating acoustical elements of said system including each waveguide section by means of an equivalent electrical circuit, wherein the acoustical elements including said distributed elements are represented in the circuit via S":".electrical components.

3. An improved acoustic filter according to Claim 1, wherein said optimizing includes analysing said circuit and changing values of said components from MJP C:\WINWORD\MARIE\GABNODELXACOUST13.DOC -17- coarsely chosen initial values until said second response is simulated by target values of said components.

4. An improved acoustic filter according to Claim 3, wherein said circuit is analysed via a technique capable of incorporating a distributed structure.

An improved acoustic filter according to Claim 3, wherein said circuit is analysed via a technique based on Kirchhoff Law.

6. An improved acoustic filter according to Claim 3, wherein said circuit is analysed via numerical simulation means.

7. An improved acoustic filter according to Claim 6, wherein said simulation means includes a digital computer.

8. An improved acoustic filter according to claim 7, wherein said simulation means includes a commercially available computer based simulation package.

9. An improved acoustic filter according to Claim 3, wherein said circuit includes one or more interfaces representing discontinuities between waveguide sections having differing characteristic impedances.

An improved acoustic filter according to Claim 3, wherein said second Sresponse is realized in said system by attributing physical dimensions to said target values.

11. An improved acoustic filter according to Claim 1, wherein said quantities of said resonances are determined by the number of acoustic waveguide sections included in said acoustic filter. MJP C:\WINWORDMARIE\GABNODEL\ACOUSTI3.DOC -18-

12. An improved acoustic filter according to Claim 1, wherein said frequencies of said resonances are determined by distances between pairs of impedance discontinuities in said configuration.

13. An improved acoustic filter according to Claim 1, wherein said excitations of said resonances relative to a driving excitation applied to said acoustic filter are determined by magnitudes of impedance mismatches at the impedance discontinuities of said configuration.

14. An improved acoustic filter according to Claim 13, wherein the magnitude of the or a respective impedance discontinuity is determined by reference to the characteristic impedances of the acoustic waveguide sections which interface the impedance discontinuity.

15. An improved acoustic filter according to Claim 13, wherein the magnitude of at least one impedance discontinuity is greater than a discontinuity caused by a bend.

16. An improved acoustic filter according to Claim 1, wherein said configuration includes acoustic waveguide sections connected in cascade.

.17. An improved acoustic filter according to Claim 1, wherein said configuration includes acoustic waveguide sections connected to form one or more branches.

18. An improved acoustic filter according to Claim 1, wherein each acoustic waveguide section is formed by a duct or vent of constant cross sectional area and wherein the or each impedance discontinuity represents a point of abrupt °t :change in cross-sectional area of the associated duct or vent. SMJP C:\WINWORD\MARIE\GABNODECOUST13.DOC -19-

19. An improved acoustic filter according to Claim 1, wherein the non- fundamental resonances which contribute to said second response are preserved by an absence of significant damping in the loudspeaker system.

20. A loudspeaker system incorporating an improved acoustic filter according to any one of the preceding claims.

21. A method of constructing an improved acoustic filter for a loudspeaker system, said acoustic filter including a first acoustic waveguide section having a first characteristic impedance and at least a second acoustic waveguide section having a second characteristic impedance, connected to and arranged to interact with said first acoustic waveguide section, said system having an enclosure in combination with at least one electroacoustic driver and said system having a first response, said acoustic filter being incorporated with said enclosure and having a second response so as to produce said first response from said loudspeaker system, said method including: modelling the first and the at least second acoustic waveguide sections such that the sections are represented in the model as distributed elements wherein said modelling includes simulating acoustical elements of said system including each waveguide section by means of an equivalent electrical circuit wherein the acoustical elements including said distributed elements are 9~e represented in the circuit via electrical components; and optimizing the model to produce an alignment of resonances including non- fundamental and out of band resonances based on resonant path between pairs of impedance discontinuities in said filter and, wherein said optimizing includes analysing said circuit and changing values of said components from roughly chosen initial values until said resonant paths are tuned and quantities, frequencies and excitations of said resonances are adjusted to simulate said second response by target values of said components.

22. A method according to Claim 21, wherein said circuit is analysed via a technique capable of incorporating a distributed structure. I MJP C:\WINWORD\MARIEGABNODEL\ACOUSTI3.DOC 1~'

23. A method according to Claim 21, wherein said circuit is analysed via a technique based on Kirchhoff Law.

24. A method according to Claim 21, wherein said circuit is analysed via numerical simulation means.

A method according to Claim 21, wherein said numerical simulation means includes a digital computer.

26. A method according to Claim 25, wherein said numerical simulation means includes a commercially available computer based simulation package.

27. A method according to Claim 21, wherein said circuit includes one or more interfaces representing discontinuities between waveguide sections having differing characteristic impedances.

28. A method according to Claim 21, wherein said second response is realized in said system by attributing physical dimensions to said target values.

29. A method according to Claim 21, wherein said quantities of said resonances are determined by the number of acoustic waveguide sections included in said acoustic filter.

30. A method according to Claim 21, wherein said frequencies of said resonances are determined by distances between pairs of impedance discontinuities in said configuration.

:31. A method according to Claim 31, wherein said excitations of said resonances relative to a driving excitation applied to said acoustic filter are determined by magnitudes of impedance mismatches at the impedance discontinuities of said filter. 3 MJP C:\WINWORD\MARIE\GABNODEL\ACOUSTI3.DOC kx: a/I~ -21

32. A method according to Claim 31, wherein the magnitude of a respective impedance discontinuity is determined by reference to the characteristic impedances of the acoustic waveguide sections which interface the impedance discontinuity.

33. A method according to Claim 31, wherein the magnitude of at least one impedance discontinuity is greater than a discontinuity caused by a bend.

34. A method according to Claim 21, wherein said filter includes acoustic waveguide sections connected in cascade.

A method according to Claim 21, wherein said filter includes acoustic waveguide sections connected to form one or more branches.

36. A method according to Claim 21, wherein each acoustic waveguide section is formed by a duct or vent of constant cross sectional area and wherein each impedance discontinuity represents a point of abrupt change in cross-sectional area of the associated duct or vent.

37. A method according to Claim 21, wherein the non-fundamental resonances which contribute to said second response are preserved by an absence of significant damping in the loudspeaker system.

38. A method of constructing a loudspeaker system incorporating a method according to any one of claims 21 to 37.

39. An improved acoustic filter constructed by a method according to any one of claims 21 to 37. S'

40. A loudspeaker system constructed by a method according to claim 38. R g j MJP C:\WINWORD\MARIE\GABNODEL\ACOUSTI3.DOC AIT I -22-

41. An improved acoustic filter substantially as herein described with reference to the accompanying drawings.

42. A method of constructing an improved acoustic filter substantially as herein described with reference to the accompanying drawings.

43. A loudspeaker system substantially as herein described with reference to the accompanying drawings.

44. A method of constructing a loudspeaker system substantially as herein described with reference to the accompanying drawings. DATED: 5 May, 1999 PHILLIPS ORMONDE FITZPATRICK Attorneys for: GRAEME JOHN HUON, GREGORY KEITH CAMBRELL and WALTER MELVILLE DOWER a* "a A.' MJP C:WNODMREG*OE~OSI DC 'v O' Appendix GLOSSARY OF TERMS S* S BI c CA d k MA P r SD XD UD Zo product of magnetic flux density and voice-coil wire length [Tm] s peed of sound in air (343.38 m/s at 200C) acoustical compliance [m 5 IN] length of waveguide section [m] imaginary unit wave-number co/c) [m 1 acoustical mass [kg/m 4 sound pressure [Pa N/m 2 mean distance from radiating apertures (diaphragms and ports) to observation point [m] S piston area of driver diaphragm [m 2 S piston displacement of driver diaphragm (rms) UD/(jo)) [m] piston velocity of driver diaphragm (rms) (m/s) characteristic acoustical impedance [Ns/m 5 S ambient air density (1.2048 kg/m 3 at characteristic impedance of air (413.70 Ns/m 3 at 200C) ratio of volume to acoustic compliance of a container of air yPo) (142,058 Pa at 200 C) ratio of specific heats of air at constant pressure and volume ambient air pressure (101,325 Pa) time delay of propagation [s] normalised driver displacement for a given radiated SPL (enclosure characteristic) [S 2 angular frequency [rad/s] solid angle of radiation (assumed to be 271 steradians) generator voltage from amplifier blocked voice coil impedance p 0 Po poc poC 2 Y Po T XD o) Q VG ZEV MJP C:\WINWORD\MARIE\GABNODEL ACOUST13.DOC ZAD l/jc CAS RAS jc MAD 1/SD 2 ZMD (of driver) ZAF (BI/SD) 2 ZEV (assuming ZEG 0 of amplifier) MAPI acoustical mass of rear vent CAB1 acoustical compliance of rear chamber CAB3 front chamber ahead of acoustic waveguide (B1 1/CABI MAP 1 CAS acoustical compliance of driver RAS acoustical resistance of driver MAD acoustical mass of diaphragm ZMD mechanical mobility of diaphragm UG acoustic referred source input velocity ZAG acoustic referred equivalent source impedance ZAV acoustic referred voice coil impedance (BI) 2 /S D 2 ZEV Z 1 Z 3 rear chamber lumped impedance 15 Z 2 diaphragm to waveguide lumped acoustic impedance Z5 external terminating airload impedance Z6 driver diaphragm airload impedance Ic (I In I II'. *1 I. I C MJP C:\WINWORD\MARIE\GABNODEL\ACOUSTI3.DOC