JPS6139621A - Demodulating filter of pulse width demodulating type power amplifier - Google Patents

Abstract

PURPOSE:To improve efficiency of a PWM type D class amplifier by providing a filter that removes carrier component of PWM signals in front stage of a low pass filter that demodulates analog signals. CONSTITUTION:In a PWM type D class amplifier that pulse width modulates carrier by an input analong signal, and after amplifying the modulated pulse signal by a power amplifying section 8, is demodulated by a low pass fiter LPf, a band elimination filter BEF that removes carrier component is provided in front stage of a low pass filter LPF to remove carrier component. By this way, output current at the time of no signal is made small and high efficiency is attained.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a demodulation filter for a PWM (pulse width modulation) type power amplifier, which is a type of class D amplifier.

[Conventional technology]

A PWM type power amplifier, which achieves high efficiency because the amplification operation can be performed by turning on and off switching elements, has a configuration roughly as shown in FIG. In the figure, 1 is a square wave oscillator, whose output (square wave) is integrated by a mirror integrator 2, and further amplified by a triangular wave amplifier 3, thereby generating a triangular wave T with a constant period as shown in FIG. can get. This triangular wave T is applied to one input of a comparator 4 for pulse width modulation. This comparator 4
The other input is an audio input signal Vin which is amplified by the audio preamplifier section 5 and whose amplitude is further limited by the limiter section 6.
It is. Therefore, the output P of the comparator 4 becomes a pulse train having a pulse width converted according to the magnitude relationship between the two human forces,
If the comparator 4 uses the audio input signal Vin as an inverted input and the triangular wave T as a non-inverted input, the pulse train P becomes a PWM signal wave as shown in FIG.

An amplification operation is performed on this PWM signal P.

7 is a pulse drive section for this purpose, and 8 is a pulse power amplification section. The drive section 7 and the amplification section 8 are composed of switching elements, and the amplified output thereof is a form in which only the amplitude of the PWM signal P is amplified by turning on and off the switching elements. A part of the output of this amplifying section 8 is transferred to a feedback section 9
Negative feedback may also be provided to the inverting input of the comparator 4 through the inverter. A filter section 10 is provided to demodulate the audio output signal Vout from the amplified output. The basic form of this filter section 10 has an inductance (
Coil) Lo and capacitance (capacitor) C.

This is a low-pass filter (LPF) consisting of a
is demodulated.

[Problem that the invention seeks to solve]

Generally, when an LPF is constructed of T, C, it is connected in an inverted 7-type configuration as shown in FIG. With this, a cutoff characteristic of -12 dB10 ct can be obtained, but if a steeper characteristic is desired to be obtained, a method of cascading these is used. However, when such an LPF is directly connected to the output stage of a PWM type power amplifier, the following problems occur. In other words, in the circuit of FIG. 7, the impedance characteristic (inductance L') of the speaker 11 and the filter IO
Due to its transfer function, it exhibits a cutoff characteristic as shown in Fig. 8 fa). The cutoff frequency fc at this time is set around 20 KH2 in a normal full-band speaker system. Also, if the PWM sampling frequency (repetition frequency of the triangular wave T) is fs, then normally 50KHz<f s<
It is set at about 300KHz.

As a result, the carrier component is removed and the signal is demodulated, but as it is, the LPF becomes a very heavy load relative to the carrier frequency, and a large output current io flows even when there is no signal.

If you increase the value of L to avoid this point, the impedance for high frequencies will rise by a bit (in will decrease), but the overall frequency characteristics at both ends of the speaker will also be affected, and the
As shown in figure (bl), the high frequency range is attenuated.For this reason, the efficiency of the PWM power amplifier, which is theoretically 100%, is not fully utilized.The present invention is based on a band-rejection filter ( The aim is to solve this problem by using the PWM power amplifier (BEF) and to improve the efficiency of the PWM power amplifier.

[Means for solving problems]

The present invention is a pulse width modulation type power amplifier (1
The m-filter is characterized in that a band-removal filter for removing carrier components is connected upstream of the low-pass filter.

[Effect]

The band-removal filter for carrier component removal significantly attenuates the carrier component that reaches the low-pass filter at the subsequent stage. However, since the signal band (20 Hz to 20 KHz) is passed, there is no problem with demodulation using the low-pass filter. This makes it possible to reduce the output current in when there is no signal, thereby improving the efficiency of the PWM power amplifier. This will be explained in detail below with reference to illustrated embodiments.

〔Example〕

FIG. 1 is a circuit diagram showing one embodiment of the present invention, in which a filter section (demodulation filter) 10 has a configuration in which a band-rejection filter BEFI for carrier components is disposed at the front stage, and a low-pass filter LPF for demodulation is disposed at the rear stage. ing. Filter B E F I is a parallel resonant circuit of inductance L1 and capacitance CI, and its resonance point is the carrier (sampling).
The frequency is set to fs. The filter LPF has an inductance Lo and a capacitance Co 'f! Connected in an L shape,
It is used for demodulating analog signals of 20 to 20 KHz.

The output of the power amplifier 8 when there is no signal is only the carrier component, and the amplitude of the pulse train f (tl, this pulse train f ft) with a duty of 50% as shown in FIG.
When A is expanded into a Fourier series, f ft1 = 8 fsA (sint bow - 5in 3
t+-A-sin 5j+・-・・）・・・・・・(1
) =8fsA E ↓sin ntn=1,3,5
It becomes n. Figure 2 (bl is the frequency characteristic of filter BEF I, and its resonant frequency fo is set to . Therefore, f of this filter BEF I is
The impedance to s becomes infinite and does not allow the fundamental wave component of the carrier represented by the first term of equation (1) to pass. Therefore, when there is no signal, the current flowing through the filter LPF (
Since to) in FIG. 7 is significantly reduced, efficiency is improved. Of course, if modulation is applied, the duty of the pulse train f (tl will be a value other than 50%, so a component that passes through the filter BEF1 will appear. Since this is a modulation component, it can be amplified by giving it to the filter L P F. The analog signal OUT thus obtained is demodulated.

FIG. 3 shows another embodiment of the present invention in which band-rejection filters are connected in multiple stages to further improve efficiency. As shown in equation (1), the pulse train f (tl is the fundamental wave component 5 int
In addition, 3rd harmonic 5in3t, 5th harmonic 5in5
Including t,... Therefore, in order to improve efficiency, it is also necessary to remove harmonic components. However, each amplitude is 1/
3.115. . . . Therefore, the fundamental wave component is set as 1, and the ratio of the total of the third, fifth, . . . odd harmonic components is determined as follows. From this formula, the sum up to n=5 is 0°87
It becomes 8. Therefore, if harmonics up to n=5 are removed, the amount of current wasted is reduced by the amount shown by .

FIG. 3 shows an example that realizes this, where BEF+ is a band-elimination filter for the fundamental wave component element s, BEF2 is a band-pass filter for the third-order high tIII wave 3fs, and BE
F3 is a band rejection filter for the fifth harmonic 5fs. ■, 1 to L3 and C1 to C3 are each filter BE
The inductance and capacitance forming FI to BEF 3 have the following relationship.

■ FIG. 4 is a specific example of FIG. 3, with inductance L.

One feature of ~L3 is that a series of coils 101 are used separately. The second feature is that the capacitances 01 to C3 should all be the same value (Cl=C2=C3), and the resonance frequency f s,
3fs, 5fs is inductance L+~L
This point is set by making 3 different. One way of thinking is to keep Ll-L3 constant and vary Cl-C5, but as in this example, the method of dividing the series of coils 101 and varying L+~I,3 is simply to change the number of turns T1~T3.
This method is easier to implement in terms of element selection than the method of changing C+ to C3. From this point of view, in the example of Fig. 4, L 1-3L2 = 5L2 (Tl-
An intermediate tap is provided at 3T2-5 T2), and CI to C3 of the same value are connected in parallel.

〔Effect of the invention〕

As described above, according to the present invention, since the filter that removes the carrier component of the PWM signal is provided before the low-pass filter, it becomes difficult for the current due to the carrier component to flow through the low-pass filter that demodulates the analog signal. P
This has the advantage that the efficiency of the WM type power amplifier is improved.

[Brief explanation of the drawing]

Fig. 1 is a circuit diagram showing one embodiment of the present invention, Fig. 2 is an explanatory diagram of band-rejection filter characteristics, Fig. 3 is a circuit diagram showing another embodiment of the invention, and Fig. 4 is a specific example thereof. times to show? Figure 8, Figure 5 & J: A block diagram showing an example of a PWM type power amplifier, Figure 6 is its operating waveform diagram, Figure 7 is a circuit diagram showing an example of a conventional demodulation filter, and Figure 8 is its It is a characteristic diagram. In the figure, 1 is a square wave oscillation section, 3 is a triangular wave amplification section, and 4 is a PW
M comparison section, 8 a pulse power amplification section, 10 a filter section (demodulation filter), 11 a speaker, 101 a coil,
L P F is a low pass filter, BEFI-BEF3 is a band rejection filter, Co-C3 is a capacitance, L o -L
3 is inductance.

Claims (3)

[Claims] (1) In the demodulation filter of a pulse width modulation power amplifier that pulse width modulates a carrier with an input analog signal, power amplifies the modulated pulse signal, and then demodulates it with a low pass filter, a band rejection filter that removes the carrier component is used. A demodulation filter for a pulse width modulation type power amplifier, characterized in that the demodulation filter is connected before the low-pass filter. (2) A demodulation filter for a pulse width modulation type power amplifier according to claim 1, characterized in that a series of coils are used separately for the inductance of the band-rejection filter and the inductance of the low-pass filter. (3) The capacitance of each band-elimination filter for fundamental waves and harmonics is set to the same value, and the values of each inductance forming a parallel resonant circuit are set to be different from each other. A demodulation filter for the pulse width modulation type power amplifier described in .