Digital power amplifier based on a multi-primary-winding transformer

Abstract

The invention relates to a new and improved technique for digital power amplification that uses a special transformer for signal conversion. The device consists of a transformer with a secondary winding for connection to a load, which may be a loudspeaker, and n primary windings. The primary windings are individually controlled by switches, which may be MOSFET transistors, activated by the corresponding bits of an input digital signal. When a bit is set to "1", the associated switch is activated, allowing a constant current to flow through the corresponding primary winding. Each primary winding has a number of turns such that the integral of the magnetomotive force over a sampling step doubles that of the preceding primary, starting with a minimal number of turns for the least significant bit.A sign bit allows the current in each primary winding to be reversed, via H-bridges, according to whether the input digital signal is positive or negative. The current induced in the secondary winding results from the sum of the currents in the activated primaries, thus creating an analog signal proportional to the value of the input digital signal. The device according to the invention is particularly suited to the audio industry as a high-resolution digital power amplification stage, depending on the number of bits and current values ​​chosen, and with reduced distortion because the digital-to-analog power conversion is performed primarily by magnetic rather than electronic means. Figure to be published for the abstract: [Fig 1].

Description

Title of the invention: Digital power amplifier based on a transformer with multiple primary windings

[0001] The amplification of a sound signal is conventionally achieved using an electronic device that increases the power of an audio signal. This device takes a weak sound signal and makes it loud enough to be heard clearly through loudspeakers. Prior art

[0002] There are several types of amplification devices; currently used amplification devices include:

[0003] Vacuum tube amplifiers:

[0004] Amplification is achieved using vacuum tubes, also called electronic tubes, such as triodes, tetrodes, and pentodes. Electronic amplification of the audio signal provides good power and sound quality.

[0005] Transistor amplifiers:

[0006] Amplification is achieved using transistors, which are semiconductors, either bipolar or field-effect. Transistors allow for more compact and portable designs.

[0007] These devices (tube or transistor) are based on similar electronic principle diagrams and typically use 4 classes of amplification:

[0008] Class A offers good sound quality with low distortion, but consumes a lot of energy because the active elements (tubes or transistors) are always conducting.

[0009] Class B is more efficient, but introduces crossover distortion at the level where the active elements pass from one half of the signal to the other.

[0010] Class AB combines the advantages of classes A and B to reduce distortion while improving energy efficiency.

[0011] Class D amplifiers employ pulse-width modulation (PWM) and use active elements that switch at high frequencies to amplify the audio signal. Class D amplifiers are very efficient and dissipate much less heat than previous classes.

[0012] Other types of amplification devices:

[0013] Acoustic horns have also been used, mainly in phonographs and gramophones. They operated without electricity, simply using the shape of the horn to concentrate and direct sound waves, thus amplifying the mechanical sound produced by the vibrations of a needle in a disc. These amplifying devices used in the last century are now obsolete.

[0014] However, in the prior art described above, it can be seen that in all these devices, the digital input signal is first converted into an analog signal and then amplified to a higher level. To date, these amplification devices use an amplification technique that is inherently analog. Even Class D, which uses switching techniques reminiscent of 1-bit digital systems, remains fundamentally an analog amplification method. These devices have the disadvantage of exhibiting distortion and phase shifts inherent in closed-loop amplification, and the analog input signal is more fragile and prone to noise than a purely digital signal.

[0015] The present invention aims to introduce a new class and an amplification device having a direct conversion of the digital input signal into an analog power signal, without passing through an intermediate stage of fragile and noise-prone analog signal, in open loop, making it possible to overcome the problems mentioned above.

[0016] This amplification device comprises a transformer having n primary windings and one secondary winding, said transformer providing both digital-to-analog conversion and power amplification of the input signal in its secondary circuit; n bits + 1 sign bit being the maximum bit resolution of the digital power amplifier. It also comprises n controlled current sources and n polarity-reversing devices. Each primary winding of the transformer is supplied by an associated current source that delivers pulses. The peak current of these pulses is adjusted so that the integral of the magnetomotive force over a sampling interval is in a ratio of 2 between each successive winding. Each current source can be turned ON or OFF at each sampling interval by one bit of the digital input signal.The direction of the current in each winding can be reversed by a sign bit, using a polarity inverter, depending on whether the input digital signal is positive or negative. Finally, the gain of the digital amplifier is determined by a table containing the peak current values ​​of each of the n current sources. Brief description of the figures

[0017] Fig. 1 represents the block diagram of a new class amplifier device with direct conversion of the input digital signal into an analog power signal using common electronic components.

[0018] Figure 2A represents an H-bridge in its rest position. The four switches are deactivated.

[0019] Fig. 2B represents an H-bridge when SI and S4 are closed.

[0020] Fig. 2C represents an H-bridge when S2 and S3 are closed.

[0021] Fig. 2D represents the electronic diagram of an H-bridge.

[0022] Figure 3 shows the shape of the current pulses in the windings primary, which are in soft switching (trapezoidal shape).

[0023] Fig. 4 represents the electronic diagram of a current source. Description of the implementation methods

[0024] The sub-assemblies constituting the amplifier device for direct conversion of the input digital signal into an analog power signal are explained below.

[0025] First sub-assembly: transformer.

[0026] Reference 1 in [Fig.1] represents a transformer having a secondary winding 4 and n primary windings 5. Each primary winding is connected to an independent current generator which delivers constant current pulses.

[0027] The transformation ratio in terms of currents is: L _ A p

[0028] The contribution of the winding i: — A[p (ij jy( f )

[0029] The product 1 pii)TVp(i) is also called magnetomotive force = fmm(i).

[0030] By application of the superposition theorem: y _ (i) JV (i) with ' • Np(i) is the number of turns of primary i • Ns is the number of turns in the secondary winding • Ip(i) is the current of primary i • Is is the secondary current • i is between 0 and n - 1

[0031] If the currents Ip and the number of turns Np are calibrated in such a way that I(P) ( i ) TVp (i) = 21 p ( i - 1 ) .Np ( i - 1 ) or fmm(i) = 2.fmm(i - 1 ) (the product of the current by the number of turns of a primary winding to the next is in geometric progression of common ratio 2), then the current induced in the secondary (which results from the addition of the currents of the activated primaries) is an analog signal proportional to the value of the binary digital input signal, n bits + 1 sign bit represent the desired precision in number of bits of the digital amplification.

[0032] Second subset: H-bridge.

[0033] Reference point 2 in [Fig. 1] represents an H-bridge allowing the sign of the current in the associated primary winding 5 to be reversed according to the positive or negative alternation of the digital input signal. There are as many H-bridges as there are primary windings on transformer 1.

[0034] When the H-bridge is in the rest position, as shown in [Fig.2A], the four switches are deactivated and the primary winding is not energized.

[0035] When SI and S4 are closed, as shown in [Fig.2B], the positive supply goes to the left of the primary winding and the negative supply to the right. The primary operates in positive alternation.

[0036] When SI and S3 are closed, as shown in [Fig.2C], the positive supply goes to the right of the primary winding and the negative supply to the left. The primary then operates in negative alternation.

[0037] Fig. 2D represents the electronic diagram of such an H-bridge based on N- and P-channel MOSFET transistors. An implementation based on vacuum tubes is also possible.

[0038] Third sub-assembly: controlled current source.

[0039] Reference point 3 in [Fig. 1] represents a current source. Each current source is controlled by one bit of the input digital signal. There are as many controlled current sources as there are primary windings 5 ​​on the transformer.

[0040] As shown in [Fig. 3], in order to prevent reverse oscillation during current switching in the primary winding, which could cause unwanted overvoltages, the rise and fall of the current source control signal follow a controlled slope. This ensures smooth switching and trapezoidal current pulses for each bit where the generator is activated.

[0041] With such a trapezoidal impulse, the magnetomotive force fmm is no longer constant over the time interval Te (or sampling period). The integral of the magnetomotive force, which is the fmm applied over a time interval, is then considered.

[0042] The integral of the magnetomotive force is found by integrating the mmf with respect to time: = = / ^(^. / ( / )}dt aVeC : • J is the integral of the magnetomotive force in ampere-revolutions-seconds (Ats), • fmm = N x I is the magnetomotive force in ampere-revolutions (A-t), • N is the number of turns, • I is the current in amperes (A), • Te is the sampling period in seconds (s).

[0043] The resultant of the powers of the individual bits is obtained by adding the primary fluxes in the transformer's magnetic circuit. In order to recover the linearity of the digital-to-analog conversion, the weighting rule must be that, from one winding i to the next, the integral of the magnetomotive force over the period The sampling rate must be in geometric progression with a ratio of 2. That is to say, the peak current must be calibrated, according to the number of turns, so that J(i) = 2.J(il).

[0044] Similarly, J(i) (and consequently the energy injected into winding i) must be constant for each period during which the current source is controlled. In particular, it must not be temperature-dependent.

[0045] Figure 4 shows the electronic diagram of such a current source based on a MOSFET transistor. A vacuum tube-based implementation is also possible.

[0046] Fourth subset: digital signal generator.

[0047] Reference figure 8 in [Fig. 1] represents a digital signal generator that synchronously delivers, at each sampling period of the input digital signal, the logic levels of the n bits + 1 sign bit. The logic circuits must be chosen so that the jitter on the data and the sign bit is as low as possible. Real-time software or microcode makes this possible. An input audio file in PCM format is well suited to generating the digital signal because the samples and the sign bit are easily extracted from it. Many modern audio sources (streaming, digital files, etc.) are already in digital format.

[0048] Regarding output filtering, if the input signal exhibits stepping, high-quality audio transformers have a bandwidth that can reach several tens of kilohertz, but rarely beyond. The high-frequency components associated with the stepping of the input digital signal will be strongly attenuated by the transformer due to its limited bandwidth; this will have the effect of smoothing out the abrupt discontinuities in the input signal. However, this smoothing will not be perfect. In order to improve the smoothing, an oversampling circuit with interpolation is used. For example, if the input signal is sampled at 44.1 kHz, it is possible to oversample this signal to a much higher frequency, such as 88.2 kHz or 176.4 kHz. Thus, the constraint of a low-pass filter at the output of -60 dB / octave becomes -30 dB / octave or -15 dB / octave, which will be effectively filtered by the transformer.The implementation of an oversampler with interpolation in the digital signal generator 8 can be done with a digital signal processor (DSP) or integrated circuits specialized in oversampling.

[0049] In this magnetic conversion system, there is no feedback loop from the secondary circuit to the input; the gain is fixed by the peak current setpoint values. As a result, the inherent defects of differential amplifiers are not present: • The phase rotations of the signal are zero. • The rise time is very short because the future output signal does not depend on the input signal at time t.

[0050] Also, current-mode amplification, because electroacoustic transducers (loudspeakers) are current-controlled devices, allows us to overcome their variable impedance depending on the frequency; this therefore allows for better linearity.

[0051] Furthermore, the input signal is immune to electronic noise, and the direct conversion improves signal purity by reducing quantization errors and artifacts associated with multiple conversions. This contributes to creating a very high-fidelity audio amplifier that accurately reproduces recordings for audiophile music lovers without the need for expensive cables.

[0052] As explained above, this device is intended for music lovers seeking high-quality sound reproduction. The present invention can also be used in the film industry as an amplifier with unparalleled clarity and sonic transparency. In the film industry, there is a strong demand for equipping cinemas with high-fidelity sound reproduction systems for projecting films with immersive sound quality.

[0053] Other uses are possible in auditoriums and amphitheaters for conferences and shows and in music and theater halls for rehearsals and artistic performances.

[0054] This device, and this new class of amplification, capable of directly converting the digital signal into an analog power signal without going through an intermediate DAC conversion, has many advantages which make it a perfect solution for audiophiles and applications where sound quality and efficiency are paramount.

Claims

Demands

1. Power amplifier device and new class of purely digital amplification, without passing through a fragile and noise-prone intermediate analog signal stage, characterized in that it comprises a transformer (1), having n primary windings (5) and a secondary winding (4), said transformer performs both digital-to-analog conversion and power amplification of the input signal in its secondary circuit; n bits + 1 sign bit being the maximum bit resolution of the digital power amplifier.

2. Digital power amplifier device according to claim 1 characterized in that it comprises n controlled current sources (3).

3. Digital power amplifier device according to claim 1 characterized in that it comprises n polarity reversing devices (2).

4. The device according to claim 1, characterized in that each primary winding of the transformer (1) is supplied by a current source (3) which delivers pulses. The peak current of these pulses is adjusted such that the integral of the magnetomotive force over a sampling interval is in a ratio of 2 between each successive winding.

5. Device according to claim 1 characterized in that each current source can be ON or OFF at each sampling interval by a bit (7) of the digital input signal.

6. Device according to claim 1 characterized in that the direction of the current in each winding can be reversed by a sign bit (6), using a polarity reversing device (2), according to the positive or negative alternation of the digital input signal.

7. Device according to claim 1 characterized in that the gain of the digital amplifier is determined by a table containing the values ​​of the peak currents of each of the n current sources (3).

Images