IMPROVED MULTI-BAND LOUDSPEAKER SYSTEM

FIELD OF THE INVENTION

The present invention relates to a multi-band or multi-way loudspeaker system for reproducing sound having spectral content. In particular the invention relates to an arrangement of the system such that a transition to each frequency band correlates with demand for power based on the spectral content of the sound.

BACKGROUND OF THE INVENTION

It is traditional in the design of a loudspeaker system to split the system into several frequency bands. The bands are integrated by a crossover network to provide a full range loudspeaker system. This is done because it is not practical to build a loudspeaker system than can accommodate the full range of frequencies in the audible spectrum (20 Hz to 20 kHz) in a single band. Three band systems comprising a subwoofer, mid-band driver and tweeter are common in the prior art. Four or five bands systems are sometimes used in more expensive loudspeakers to facilitate operation of electro-acoustic transducers in their optimum ranges.

In a stereo installation it is common for a loudspeaker system to cover a full range of sound frequencies wherein all bands that are used to make up the full audible range are built into one enclosure and driven with a single amplifier for each stereo channel. In a surround sound or home theatre system it is common to separate frequencies below a subwoofer cut-in frequency (typically 80, 100 or 120 Hz) to a monophonic subwoofer channel and to retain separate channels for frequencies above the subwoofer cut-in frequency. Typically the monophonic subwoofer cut-in frequency has been chosen around 100 Hz because it was generally considered that sound frequencies above 100 Hz are too directional to be accommodated by a monophonic channel. Directional in this instance denotes that their location may be easily discernable. The effect of causing location to be discernable is known as localisation.

Although the audible frequency range may be split in the prior art into 2, 3, 4 or more frequency bands, it is nevertheless still common practice to use a single amplifier for each channel, with a possible exception of the subwoofer channel which may include multiple active subwoofers.

Thus in stereo or surround channels a single amplifier is commonly used for each channel and passive crossovers are employed to integrate multiple electro-acoustic transducers covering different frequency bands. It has been established that the demand for power associated with typical program (spectral) content varies according to the frequency band. However, no attention has been given in prior art loudspeaker systems to address efficient use of power based on frequency bands. Also no attention has been given to matching the cross over point in the audible spectrum to a monophonic or common channel with the demands of program content.

Consequently the present invention may address arrangement of the frequency bands to promote a more efficient use of available power.

SUMMARY OF THE INVENTION Fig. 1 shows a graphical representation of distribution of power relative to frequency for typical spectral content. The distribution is shown on a logarithmic scale and reflects a relatively rapid and more or less steady decline in demand for power from the lower frequency limits of typical content to about 300Hz, followed by a relatively steady demand for power from about 300Hz to the upper frequency limits of typical content. This distribution indicates that most demand for power occurs below about 300 Hz. This suggests that efficiency of a multi-band loudspeaker system may be improved by choosing a suitable cross over point to a common channel. Fig. 1 also shows that there exists a steep fall in demand for power as frequency rises, while demand for power above about 300 Hz is relatively low and steady. According to one aspect of the present invention the cross over point may be chosen such that it may correlate with a transition point in demand for power based on the spectral content. According to another aspect of the present invention the cross over point to a common channel may be substantially greater than 100

Hz and preferably about 300 Hz. By increasing the cross over point to a frequency substantially above 100 Hz, overall power requirements for a given sound pressure level with typical spectral content may be reduced substantially while still being able to achieve a substantially flat frequency response.

The problem of localisation of monophonic frequencies above 100 Hz may be addressed according to the present invention by reducing directional components in the monophonic frequencies. The present invention recognizes that the problem is not that frequencies above 100 Hz are inherently directional, but that subwoofers typically generate harmonics that are directional. Pure sine waves on the other hand are relatively non-directional at frequencies up to about 300Hz.

According to one aspect of the present invention there is provided a multi-band loudspeaker system for reproducing sound having spectral content, said system comprising: a first frequency band; a second frequency band; a third frequency band; and a respective power amplifier for driving each frequency band, wherein power and transition to each frequency band are arranged such that the power available to each band is at least sufficient to avoid clipping and/or compression having regard to demand for said power based on said spectral content, and wherein power in excess of said demand in each band is minimized.

The transition to each frequency band may be arranged such that for each frequency band a ratio of amplifier power to required power based on average demand of the reproduced sound is substantially the same in each frequency band.

The power required to achieve a substantially flat frequency response at rated output may be reduced to each successive band substantially by an order of

magnitude, at least for a loudspeaker system having a relatively small enclosure size.

The first frequency band may be below substantially 100 Hz. The second frequency band may be between substantially 100 to 300 Hz. The third frequency band may be above substantially 300 Hz.

The required power to each successive frequency band may be in the ratio substantially 100:10:1 . In one embodiment the power to the first, second and third frequency bands may be 100 watts, 10 watts and 1 watt respectively. The first frequency band may be monophonic. The second frequency band may be monophonic. The first and second frequency bands may be provided in a single enclosure.

The frequency at which crossover to the third frequency band occurs may be chosen on the basis that it may be the highest frequency that may be reasonably obtained from a loudspeaker without causing discernable localisation when the system is in operation.

Since directional information in pure sine tones is not readily discernible below 300 Hz it may be ignored in a multi-channel loudspeaker system. Therefore there may be little advantage in providing signals below approximately 300 Hz to stereo or surround sound channels if monophonic low frequency channels in the loudspeaker system are substantially free of harmonic frequencies. According to the present invention discernable directional content from monophonic low frequency channels may be substantially avoided by means including one or more low pass filters, use of low distortion loudspeaker components and low distortion loudspeaker enclosure designs.

In one embodiment the present invention may provide a 3 band system wherein each band is separately amplified at an appropriate power level. Although multiple separately amplified bands have been employed in prior art systems they fail to arrange crossover frequencies and amplifier power ratings to match power demand of typical program content. For example it is not

uncommon to see crossover frequencies set as high as 2 kHz in three band systems. Accordingly frequencies between 300 Hz and the crossover point are inefficiently driven by a high powered amplifier or alternatively are driven by an amplifier that is inadequate to meet demands of the woofer band. Another example is that power ratings of amplifiers in active multi band systems are typically matched to a flat frequency response instead of being matched to power demand based on spectral content of typical program material. The deficiencies of prior art arrangements are particularly apparent where Class A amplifiers are used.

Use of class A amplifiers for bands above 300Hz is desirable notwithstanding their relatively low efficiency because they are more linear than other designs and because the human ear is more sensitive to non linear distortion at frequencies higher up the audible spectrum (eg. above about 450Hz). If it is desired to deliver class A amplifier performance in a full range stereo system, for example using full range amplifiers, then assuming that amplifiers of 100 watts per channel are used wasted power dissipation from the amplifiers would be at least 200 watts per channel totalling 400 watts. However by using separate amplifiers according to the present invention equivalent performance may be obtained using low dissipation monophonic amplifiers (eg class AB) for bands below 300 Hz where amplifier performance is masked by loudspeaker performance giving losses of approximately 33 watts, and class A amplifiers above 300 Hz. Assuming that power to the first, second and third frequency bands is 100 watts, 10 watts and 1 watt respectively, a waste power dissipation of 2 watts per channel may be achieved for the class A amplifiers giving a total dissipation of less than 40 watts. The relatively low power dissipation may enable the class A amplifiers to be constructed without a heat sink. Overall system efficiency gain may typically be an order of magnitude.

DESCRIPTION OF A PREFERRED EMBODIMENT

A preferred embodiment of the present invention will now be described with reference to the accompanying drawings wherein:-

Fig. 1 shows (in broken outline) a graphical representation of power demand relative to frequency for typical program content together with (in solid outline) frequency bands selected according to one embodiment of the present invention; and

Fig. 2 shows a block diagram of a 3 band loudspeaker system according to one embodiment of the present invention.

Since power demand of typical program content varies according to frequency, power to each frequency band may be distributed such that it may more closely match spectral content of the program. One factor that may determine power distribution is that (refer broken outline in Fig. 1 ) demand beyond about

300 Hz tends to flatten out. Experiments conducted by the inventor show that an efficient split of frequency bands (refer solid outline in Fig. 1 ) for a three band system may include:

• a first or subwoofer band A for frequencies 100 Hz and below. The associated loudspeaker may be referred to as a subwoofer.

• a second or woofer band B for frequencies between 100 Hz to 300 Hz. The associated loudspeaker may be referred to as a woofer.

• a third or high frequency band C for frequencies 300 Hz and above.

Acceptable crossover points for the three frequency bands may be 100 Hz +/- 25% and 300 Hz +/- 25%. As shown in Fig. 1 power demand for typical program content declines by an order of magnitude (factor of approximately ten) for each band as frequency increases. It is fortuitous therefore from the point of view of directional content that 300 Hz is also an acceptable cross over frequency for monophonic channels as this may facilitate matching of input power requirements to the monophonic cross over frequency.

In Fig 1 the broken line is an example of power demand associated with typical program content. In order to avoid clipping and/or compression at rated output the amplifier driving each band A, B, C should have a power rating above the broken line at all points. The amplifier may be most efficient if the hatched

areas between its power rating (the solid outline) and power demand (the broken outline) can be minimized. Any significant variation using different crossover frequencies or less bands will enclose greater hatched areas and is likely to be less efficient.

Significant efficiency may be gained for a loudspeaker system according to the present invention by using separate amplifiers to drive sections of a loudspeaker system that is split into three bands with crossover frequencies at approximately 100 Hz and 300 Hz respectively. Further benefits and efficiency gains may be achieved by operating the 100 Hz to 300 Hz band monophonically wherein the content of all channels is combined into a single monophonic channel.

It is possible to achieve greater efficiency by adding more separately amplified bands. However this may only be the case if the bands are selected to minimise the area between amplifier power and a power demand curve for typical program content. The power demand curve for typical program content may vary somewhat according to the content. However it is unlikely to depart substantially from the broken line in Fig 1.

Fig. 2 shows a multi-band loudspeaker system 20 comprising a centre channel (C), front left channel (FL), front right channel (FR), rear left channel (RL), rear right channel (RR) and low frequency effects channel (LFE). System 20 operates with a signal source 21 having spectral content from below 50 Hz to above 10 kHz. System 20 is split into three frequency bands. Each frequency band is serviced by a respective electroacoustic transducer 22, 23, 24. Transducer 24 relates to the RR channel, the latter being essentially similar to the C, FL, FR and RL channels. Therefore only the RR channel will be described below as it may be assumed that the C, FL, FR and RL channels will be similar to the RR channel. Each transducer 22, 23, 24 is driven by a respective power amplifier, 25, 26, 27. Power amplifier 25 is fed via a low pass filter 28 adapted to pass frequencies substantially below 100Hz. Power amplifier 26 is fed via a band pass filter 29 adapted to pass frequencies substantially between 100 to 300 Hz. Power amplifier 27 is fed via a high pass

filter 30 adapted to pass frequencies substantially above 300 Hz. In one embodiment power amplifiers 25, 26, 27 may be rated at 100 watts, 10 watts and 1 watt respectively. Power amplifier 27 at least may be class A.

The bands below 300 Hz may be monophonic. In a practical system the transducer 22 (subwoofer) and transducer 23 (woofer) may be incorporated into a separate enclosure 31 . An active system may include associated amplifiers 25, 26 for the woofer and subwoofer bands. In such a construction the volume of the woofer may be only a fraction of that of the subwoofer. The fraction may vary according to subwoofer cut-in frequency. Filters 28, 29, 30, power amplifier 27 and associated circuitry may be incorporated in a separate enclosure 32. In some embodiments power amplifier 26 may also be incorporated in enclosure 32.

In more elaborate embodiments of the present invention four or more separately amplified bands may be employed to achieve an even more efficient loudspeaker system. The bands may be chosen according to the input power demands of typical program content. However the cost benefits of such systems may depend on their application.

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 invention.