Amplifiers – Modulator-demodulator-type amplifier
Reexamination Certificate
2001-04-27
2002-04-16
Pascal, Robert (Department: 2817)
Amplifiers
Modulator-demodulator-type amplifier
C330S20700P
Reexamination Certificate
active
06373336
ABSTRACT:
FIELD OF THE INVENTION
The invention relates to a method of attenuating noise and zero crossing distortion in an amplifier consisting of two pulse width modulators in which an analogue or digital signal is pulse-width modulated to provide pulse-width modulated small-signals, wherein the outputs of the pulse width modulators represent a non-inverted analogue or digital input signal and an inverted analogue or digital input signal, respectively, controlling two sets of switches which, by means of a voltage supply, feed a load with two pulse-width modulated great-signals, which are proportional to the pulse-width modulated small-signals so as to generate a pulse-width modulated great-signal of the class BD type.
BACKGROUND OF THE INVENTION
The invention moreover relates to an amplifier having means for attenuating noise and zero crossing distortion and of the type comprising two pulse-width modulators adapted to pulse-width modulate an inverted and non-inverted analogue or digital signal to provide two pulse-width modulated small-signals, said pulse-width modulated small-signals being fed to two sets of switches adapted to connect and disconnect a voltage supply to provide a pulse-width modulated great-signal for a load.
The invention finally relates to uses of the amplifier.
Traditional amplifiers for audio use are used for amplifying audio signals and forming sound images in the loudspeaker. These amplifiers, however, do not have very high power efficiencies. Therefore, amplifiers are being constructed according to other principles.
These high efficiency amplifiers include e.g. those based on pulse width modulation, which are also called class D amplifiers.
Such an amplifier consists of a pulse width modulator, one or more sets of switches and a low-pass filter.
The principle of a class D amplifier is that one or more sets of switches are switched to conduct and non-conduct, respectively, depending on the amplitude of a signal, such as an audio signal. The information of the audio signal is hereby converted into a number of pulses which carefully correspond to the information of the audio signal.
Pulse-width modulated amplifiers are theoretically very linear and thus have a very low distortion, but practical realisations have shown non-linearities which have caused them to be unsuitable for use in High Fidelity amplifiers, if a strong negative feedback is not established.
However, establishment of feedback systems in pulse-modulated amplifiers is not an easy task, since negative feedback performed prior to the low-pass filtering adds much noise to the system.
Additionally, the load impedance, which is a loudspeaker, is incorporated in the feedback signal, and as it may vary depending on loudspeaker selection, the design of feedback systems is made difficult.
Pulse with modulation can be divided into several modulation classes. The most used are class AD and class BD modulation. Class AD is a modulation form where only two discrete levels are used, e.g., 1 and −1. In a class AD modulated amplifier, either plus the power supply voltage or minus the power supply voltage is applied to the load.
Class BD modulation can be viewed as a differential coupling of two class AD modulators, i.e., both a non-inverted as well as an inverted version of the signal input is pulse width modulated. The load is then connected differentially to the outputs of two output stages controlled of the two class AD modulators.
Class BD operation gives in many respects a number of benefits, i.e. many noise contributions are cancelled by the differential operation and the even-ordered non-linearities are generally cancelled.
However, it has been seen in prior art that class BD amplifiers suffer from cross-over distortion and higher noise around the zero-transition of the input signal.
A great part of the artefacts in pulse-with modulated amplifiers of the class BD type occurs because of interference between the pulse with modulators and the half-bridges in the output module, which results in cross-over distortion (zero crossing distortion) of and noise on the audio signal. The zero crossing distortion and noise occur because the waveforms and especially the puls-edges of the two differentially pulse-with modulated signals are interacting, especially when the edges are nearly coinciding in time during the zero transitions of the audio signal.
It is concretely seen in that the first-occuring edge of two closely spaced flanks interferes with the generation of the second edge so that this is slightly delayed or advanced temporally. It should be noted that the phenomenon occurs primarily when both modulators and half-bridges are switched within a time period which is smaller than a time constant determined by the circuit.
As mentioned, the known methods of linearising pulse-width modulated amplifiers require the use of strong negative feedback. These methods are particularly unsuitable for consumer amplifiers where design criteria such as low complexity and general applicability in a wide range of load are important.
SUMMARY OF THE INVENTION
Accordingly, an object of the invention is to provide a method which is capable of linearising class D type amplifiers by taking the interference between the half bridges and their respective control circuits into account, and by further minimising such interference. Thereby making it possible to produce pulse-width modulated amplifiers of low complexity.
The object of the invention is achieved by the method in which when the analogue or digital signal is close to the value 0, the switching times of the pulses fed to the switches will be shifted relative in time. This can be done by applying a constant delay to the switching times of the control signals so they will coincide at a resulting audio signal value slightly different from zero.
When, the separation is performed by delaying either the pulses representing the non-inverted part of the analogue or digital signal relative to the inverted part of the analogue or digital signal, or the method is performed by delaying the pulses representing the inverted part of the analogue or digital signal relative to the non-inverted part of the analogue or digital signal, the circuit-technical advantage is achieved that the non-linear range of the amplifier is shifted from the zero crossing of the audio signal to a higher signal level. The non-linear range of the amplifier is thus of influence only when the resulting audio signal exceeds the signal level where the non-linear range is present. The advantage of this is that the non-linearity of the amplifier, seen in relation to the level of the resulting audio signal, may be reduced to a level of minimum significance. Also achieved is the advantage that a significantly lower noise level for the amplifier in idle operation is achieved and at low volumes than by a pure class AD or BD operation. The modulation form is hereafter named class ABD, since the modulation shares features of both class AD and class BD modulation.
When, the separation is performed by either temporally delaying the carrier wave signal in the pulse width modulator modulating the non-inverted signal relative to the carrier wave signal in the pulse width modulator modulating the inverted signal, or by temporally delaying the carrier wave signal in the pulse width modulator modulating the inverted signal relative to the carrier wave signal in the pulse width modulator modulating the non-inverted signal, the circuit-technical advantage is achieved that the non-linear range of the amplifier is shifted away from the zero crossing of the audio signal to a higher signal level. The non-linear range of the amplifier is thus of influence only when the resulting audio signal exceeds the signal level where the non-linear range is present. The advantage of this is that the non-linearity of the amplifier, seen in relation to the level of the resulting audio signal, may be reduced to a level of minimum significance. Another achieved advantage is that the relatively simple circuit modification results in a significantly lower n
Anderskouv Niels
Risbo Lars
Kempler Willam B.
Telecky , Jr. Frederick J.
Texas Instruments, Copenhagen ApS
LandOfFree
Method of attenuating zero crossing distortion and noise in... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method of attenuating zero crossing distortion and noise in..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method of attenuating zero crossing distortion and noise in... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2902613