Improved self-oscillating class D amplifier device

Abstract

The present disclosure relates to an amplifier device for amplifying an input signal Sx at a device input X into an output signal Sy at a device output Y for driving a loudspeaker or comparable reactive load, comprising a self-oscillating class D power amplifier stage with a transfer function KA (s). The amplifier stage includes a comparator, a power stage, and a second order low-pass output filter including an inductor and a capacitor. A time delay unit provides a delay determined by one or more of the comparator, the power stage, and the first feedback filter. The output signal Sy is fed back to the comparator via a first feedback path to establish an oscillation and provide a basic low frequency feedback. A correction stage is provided that includes a second feedback path that feeds an output signal Sy to a virtual ground input of a transimpedance amplifier via a scaling resistor. This enables pre-distortion of the amplifier input SKin by subtracting the filtered distortion component derived from the tracking filter. To further enhance the correction, a third feedback path is included. It comprises a passive network applying a first transfer function H2 to the output signal Sy and a second transfer function H3 different from the first transfer function H2 to the input signal Sx. The resulting signal is summed as a current into a transimpedance amplifier, which outputs a correction reference signal defining the signal Sr as a correction stage.

Description

TECHNICAL FIELD

[0001] The present invention relates to an amplifier device, in particular to an amplifier device with improved audio quality. BACKGROUND

[0002] In recent years, the field of audio amplification has made significant progress with the development of high-performance class-D amplifiers. Class-D amplifiers, also known as switching amplifiers, are electronic amplifiers in which the power stage semiconductors, typically transistors such as MOSFETs, operate as electronic switches rather than as linear gain devices as in older, traditional amplifiers. They work by generating a series of rectangular pulses of fixed amplitude but varying duty cycle, representing the amplitude variations of an analog audio input signal. The output of the modulator is then used to alternately turn on and off the output transistors. Since the transistors are either fully on or fully off, they consume very little power, which is a typical advantage of class-D amplifiers. A low-pass filter consisting of an inductor and a capacitor provides a path for the audible low frequencies of the audio signal and removes most of the high-frequency content from the audio signal.

[0003] Class-D amplifiers have been known for a long time and have multiple advantages over other types of amplifiers, among which they always provide better efficiency and they are relatively compact in size compared to class-A or class-AB amplifiers. However, they also face certain challenges, such as possible EMC-related issues.

[0004] The complexity of the control circuit in a class-D amplifier can present challenges in design and implementation. Achieving a balance between efficiency, size, and maintaining low distortion requires careful engineering and often trade-offs need to be made. Furthermore, the sensitivity of some class-D amplifiers to load variations can be an issue. Reactive loads can impact performance, making it important to design the amplifier to perform well under widely varying load conditions.

[0005] However, modern high-performance class-D amplifiers perform well in most applications and typically have high-fidelity characteristics, where many of the above-mentioned shortcomings of typical class-D amplifiers have been overcome to some extent.

[0006] But even modern class-D amplifiers are not always without compromise, and there is often still room for improvement.

[0007] Over the years, numerous improvements have been made to enhance the fidelity and robustness of class-D amplifiers, including feedback control techniques and improved component integration. Despite these advances, there is still room for further improvement in many commercial class-D designs, particularly in maintaining high audio quality under real-world dynamic conditions and with reactive loads.

[0008] In particular, the trade-off between efficiency, distortion suppression and control loop stability remains a design challenge, especially when higher fidelity demands are placed on these amplifiers in high-end audio applications.

[0009] It is therefore an object of the present invention to provide an improved self-oscillating class-D amplifier device that provides enhanced audio performance, reduced distortion and improved stability, while maintaining the inherent efficiency advantages of class-D operation. SUMMARY

[0010] In a first aspect of the present disclosure, the above object is at least partly achieved by an amplifier device for amplifying an input signal Sx at a device input X to an output signal Sy at a device output Y for driving a loudspeaker or a comparable reactive load, said amplifier device comprising:

[0011] - a self-oscillating class-D power amplifier stage 10 having a first transfer function K A (s) comprising:

[0012] - a comparator 01 comparing an output signal Sy of said comparator with a local input voltage SKin via a first feedback filter Hosc 06;

[0013] - a class-D power stage 03 coupled to and driven by the output of said comparator 01, converting the output voltage of said comparator to a higher voltage level with high current capability;

[0014] - a second order low pass output filter 04 comprising an inductor and a capacitor 04 coupled between the power stage 03 and the loudspeaker or comparable load, for removing at least part of the high frequency switching noise from the output signal of said class-D power stage;

[0015] - a time delay unit 02 configured to provide a time delay in dependence of one or more of said comparator, Hosc and said power stage; and

[0016] - a first feedback path from the device output Y coupling the output signal Sy back to the comparator 01 via Hosc 06, said first feedback path providing control of the oscillation frequency and at least partly providing low frequency feedback to reduce distortion of the signal in the audio frequency band; and

[0017] - a correction stage comprising:

[0018] - a second feedback path coupling said output signal Sy from the device output Y via H1 104 to a feedback input of said correction stage, further coupled to a summing node S2 which is a virtual ground input of a transimpedance amplifier HZ1 102 for further correction of errors of said output signal Sy generated by said amplifier stage, wherein said further correction is performed by pre-distorting the power amplifier input signal SKin of said amplifier stage 10 with the inverse of the isolated and filtered distortion components caused by the power amplifier stage, output filter and loudspeaker or comparable load, wherein the isolated distortion components are caused by passing the incoming signal Sr through a tracking filter K T 101 running, and the result of subtracting said feedback signal from the output of H1 104, as a scaled version of the output signal Sy;

[0019] The correction stage further comprises:

[0020] - a third feedback path coupling said output signal Sy via H2+H3 and through HZ2 of said correction stage to an additional feedback input for further enhancing said correction of said output signal Sy by said second feedback path, wherein said third feedback path is configured as a higher order additional for enhancing the selectivity of the feedback path and for providing additional feedback gain, and wherein said third feedback path comprises a network implementing a first transfer function H2 applied to the output signal Sy and a second transfer function H3 applied to the input signal Sx, the second transfer function H3 being different from the first transfer function H2, said H2 and H3 being combined within a single passive network comprising resistors and capacitors, and wherein the respective outputs of H2 and H3 are summed and supplied as a current input to a transimpedance amplifier HZ2, the output of which defines the input signal Sr.

[0021] An important factor in the research leading to the potential new amplifier design and ultimately to the amplifier design of the present invention was the actual perception of audio quality in comparison to previous designs. It was recognized that traditional Fourier transform based distortion measurements are by no means the only way to judge amplifier performance.

[0022] Throughout the design phase, the inventors were aware of updating the known amplifier design and conducting extensive listening tests. The inventors' research led to a different understanding of feedback loop design in which the highest possible overall feedback loop gain was not the only goal, but also the audible effects of the phase response and gain distribution in the loop system were evaluated.

[0023] Critical components such as operational amplifiers have been extensively tested, resulting in minor differences in perceived audio quality that cannot be clearly explained from the relevant datasheets. This has resulted, among other reasons, in the suggestion to use discrete operational amplifiers in an example of the invention.

[0024] According to this first aspect, the amplifier device as Figure 2 indicated with reference numeral 300 can be represented with three different blocks, identified as:

[0025] - a self-oscillating power amplifier device 10;

[0026] - a power amplifier enhanced by a first generation correction loop 100;

[0027] - a power amplifier further enhanced by a second generation correction loop system 200 of higher order.

[0028] The first block 10, i.e. the self-oscillating class-D power amplifier stage has a first transfer function K A (s) and comprises:

[0029] - a comparator 01 comparing its output signal Sy with a local input voltage SKin via a first feedback filter Hosc 06;

[0030] - a class-D power stage 03 coupled to and driven by the output of the comparator 01, converting the output voltage of the comparator to a higher voltage level with high current capability;

[0031] - a second order low pass output filter 04 comprising an inductor and a capacitor 04 coupled between the power stage 03 and a loudspeaker or comparable load, for removing at least part of the high frequency switching noise from the output signal of the class-D power stage;

[0032] - a time delay unit 02 configured to provide a time delay depending on one or more of the comparator, Hosc and the power stage;

[0033] - a first feedback path from the device output Y coupling the output signal Sy back to the comparator 01 via Hosc 06, the first feedback path providing control over the oscillation frequency and at least partly providing low frequency feedback to reduce distortion of the signal in the audio band.

[0034] In one example, the first feedback path can periodically update the output level of the comparator 01 and force the output filter 04 to integrate Sy closer to its exact value it should have at the time of switching. Those skilled in the art will appreciate that since the actual output level of the output of 03 is only updated at certain times defined by the switching frequency, the constant integrated voltage Sy at the device output Y will always have a small error.

[0035] Please note that 10 represents the first generation of self-oscillating class-D amplifier produced and sold by the applicant under the trademark UcD and disclosed in EP 3 721 553 B1 and is therefore incorporated by reference into the present disclosure to form part of the invention.

[0036] In the second generation of these self-oscillating amplifiers, the performance is greatly improved by adding an additional error correction in the audio frequency range with an additional feedback loop, adding an additional error correction in the audio band with a relatively high loop gain while leaving the oscillation feedback effectively intact.

[0037] To achieve this goal, the following is added to the first block 10:

[0038] - a correction stage comprising:

[0039] - a second feedback path coupling said output signal Sy from the device output Y via H1 104 to a feedback input of said correction stage, to a summing node S2, which is a virtual ground input of a transimpedance amplifier HZ1 102, for further correction of errors of said output signal Sy generated by said amplifier stage, wherein said further correction is performed by pre-distorting the power amplifier input signal SKin of said amplifier stage 10 with the inverse of the isolated and filtered distortion components caused by the power amplifier stage, the output filter, and the loudspeaker or comparable load, wherein the isolated distortion components are obtained by passing the incoming signal Sr through the tracking filter K T 101 in operation and subsequently subtracting the result from the output of H1 104 as a scaled version of the output signal Sy.

[0040] Thus, the signal from S2 is filtered in the transimpedance amplifier HZ1, 102. This amplifier serves to greatly amplify low-frequency errors while leaving the oscillation feedback intact.

[0041] The skilled person will understand that this improvement of the class-D amplifier 100 with this second feedback path included in block 100 represents the second generation of self-oscillating class-D amplifier produced and sold by Hypex Electronics under the trademark NCore and disclosed in EP 2 221 964 B1 and is thus also part of the present disclosure and incorporated by reference in the present disclosure in the form of its disclosure or in a modification thereof. In later parts of this document, in the summary and the detailed description, further details are mentioned that lead to the modification of the invention and its effects. In the various figures and block diagrams, this improvement is marked as 100.

[0042] While the block 100 proved to be a huge improvement in both measured and perceived audio reproduction quality, research and testing led to the conclusion that there was still room for improvement in the audio quality. Further research showed that this further improvement was achieved by adding extra low frequency gain in the feedback loop, which was achieved by a higher order transfer function, which was achieved by:

[0043] Said correction stage further comprises:

[0044] - added to the feedback loop structure in the form of an additional signal path via H2+H3 and HZ2 into the correction stage.

[0045] H2 and H3 are two different transfer functions combined in a single network, H2 being a feedback transfer from the device output Y, while H3 is a feedforward transfer from the device input X, in an example from the input buffer 301. The summed output of both signals Sy and Sx via transfer functions H2 and H3 respectively are fed as a current to the transimpedance amplifier HZ2 to form the signal Sr, which is the input signal to the block 100. The skilled person will understand that the extended loop structure formed by H2+H3 and HZ1 should not be seen as a separate loop, but as a higher order extension of the correction stage.

[0046] Hence, this is reflected in the features of the first aspect of the invention, further comprising:

[0047] - a third feedback path that couples the output signal Sy via H2+H3 and through the correction stage's HZ2 to an additional feedback input to further enhance the correction of the output signal Sy by the second feedback path, wherein the third feedback path is configured as a higher order additional for enhancing the selectivity of the feedback path and for providing an additional feedback gain, and wherein the third feedback path comprises a network implementing a first transfer function H2 applied to the output signal Sy and a second transfer function H3 applied to the input signal Sx, the first transfer function H2 being different from the second transfer function H3, the H2 and H3 being combined within a single passive network comprising resistors and capacitors, and wherein the respective outputs of H2 and H3 are summed and supplied as a current input to a transimpedance amplifier HZ2, the output of which defines the input signal Sr.

[0048] The amplifier device 200 is set to amplify an input signal Sx, such as an analog audio signal at the device input X, into an output signal Sy, for example an amplified audio signal representing an analog audio signal, which is forwarded to a load 05 at the device's output, also referred to as device output Y.

[0049] At the core of this device is a class-D amplifier stage 10. The amplifier stage is composed of several components; a comparator 01, a power stage 03, a low-pass output filter 04 and a first feedback path that feeds back the device output signal Sy via Hosc 06 to the input of the comparator. The intrinsic delay 02 is in fact a distributed function, composed of the design choices of all parts of 10.

[0050] The amplifier part K, 10 is set to generate an amplified output signal. This is an amplified representation of the local input signal SKin of the amplifier part. The amplification can be defined according to a first transfer function K A (s).

[0051] The output filter is a low-pass filter composed of an inductor and a capacitor 04, this LC filter is located between the output of the power stage 02 and the actual output Y of the device. The LC filter removes most of the high frequency switching noise from 03 and thus provides a filtered output of the amplified representation of the input signal SKin of the amplifier part 10. This representation is the actual output signal Sy at the device output Y.

[0052] The first feedback path via 06 is mainly responsible for the oscillation frequency and also for the main feedback of the LF audio, despite having a limited loop gain. The second and third feedback paths are specifically aimed at greatly increasing the loop gain in the audio range, while maintaining the positive feedback required for the power stage to oscillate.

[0053] The operation of the feedback path is based on the principle of feedback error correction. It compares the input signal of the amplifier section with a scaled version of the output signal Sy. Any difference between these two signals is considered an error. This error is then isolated, filtered and amplified and used to correct the output signal Sy, ensuring that it is an accurate representation of the input signal.

[0054] Error isolation can be achieved by subtracting a scaled version of the output signal Sy from the local input signal Sr. In the case of a perfect power amplifier, the result of the subtraction should be 0. This scenario works if the signal to be subtracted at the summing point S2 is first processed via the same transfer function. Since the output signal Sy will be multiplied with the transfer function of 10, the reference signal from Sy will have to be multiplied with a copy of the same transfer function. This copy is made in the tracking filter K T H1 is not used for the purpose of attenuating the output signal Sy, so now we know that any signal from the summing point S2 must be distorted, at least within the audio band.

[0055] The function of HZ1 is to significantly increase the gain in the audio band without disturbing the feedback required for the power stage oscillation. This can be achieved with a filter of the second or third order. Some alternative solutions can be seen in the attached Figure 6A , 6B These alternative solutions differ from known T-integrators with different characteristics, each with its own sound profile.

[0056] In a class-D amplifier according to the present disclosure, the feedback system plays a key role in the operational characteristics of the amplifier. The feedback system function in a self-oscillating amplifier, such as the amplifier of the present invention, is dual, which determines the HF switching frequency and at the same time significantly reduces the LF distortion compared to a classic class-D amplifier.

[0057] With the amplifier stage 10 known per se, a simple but stable amplifier device is implemented that is able to provide a high-quality and high-fidelity audio signal, but due to its basic configuration via the internal feedback path of Hocs, it also has a relatively low loop gain, resulting in non-optimal distortion cancellation.

[0058] To this end, the amplifier stage is provided with a further correction stage.

[0059] This correction stage provides a further second feedback path, which is designed to further enhance the accuracy of the output signal Sy. It does this by subtracting the input signal from a scaled version of the output signal to isolate the non-linear distortion components in the output signal. The transfer function K T (s) is inserted into the input signal to replicate the linear transfer function K A(s). This means that, in the case of a perfectly distortionless power level, the result of the subtraction will be equal to 0, or in the case of a real-world power level with a load, the subtraction result contains only error or distortion components.

[0060] The distortion component is then filtered in HZ1 and subtracted from the input signal Sr of the power amplifier. This input signal is thus "pre-distorted" against the distortion components caused by the output stage and the (unloaded) load. HZ1 is a transimpedance amplifier with high gain at frequencies in the audible range.

[0061] The correction stage thus comprises a feedback path that couples the scaled output signal Sy via H1, 104 to the input of the correction stage. This allows the system to continuously monitor the output signal and make the necessary adjustments to further correct any errors generated by the amplifier stage K.

[0062] With the second feedback path, a substantial improvement has been achieved over an amplifier device with only an amplifier power stage, however, it has been determined through listening tests that such a design still leaves room for further improvement.

[0063] The amplifier device according to the present disclosure thus comprises a third feedback path to enhance the correction stage and further improve performance, reduce distortion and improve loop gain compared to known designs based on an amplifier stage and a first correction stage.

[0064] This enhancement is provided in addition to the other layer of error correction to further improve the accuracy of the output signal Sy. This stage comprises a further feedback path that couples the output signal Sy to the input of the correction stage. This further feedback path (third feedback path) is provided between the output Y of the device and the input stage, so that, seen from input to output, the amplifier comprises a correction stage with the third feedback path and the second feedback path, and an amplifier stage with the first feedback path followed by an LC filter.

[0065] The third feedback path is in fact an enhancement of the second feedback path discussed earlier. The high order transfer function of the combined outer loop enables the loop structure to operate with a higher loop gain compared to a low order solution. Only a single correction signal can be considered to be derived from the loop structure. The overall system can first appear to have several completely independent loops, but in fact operates as a single higher order loop, thereby increasing the loop gain through a strictly tailored loop response and thus obtaining a more accurate correction signal. The third feedback path thus provides an additional boost or improvement of the error correction, further enhancing the accuracy and fidelity of the output signal. This makes it a distinctive feature of the amplifier device, particularly in applications where high precision and signal quality are required. The higher order loop filter configuration of the further feedback path allows for a higher loop gain and more complex error correction, making this amplifier device highly effective and accurate.

[0066] In an example, the HZ1 filter is characterized by at least three capacitors configured to implement a higher order transfer function compared to earlier designs, thereby enabling a higher loop gain compared to lower order configurations with fewer capacitors in the feedback network, and improving distortion cancellation and fidelity of the output signal. As a result, the output signal more reliably represents the input signal than earlier versions, thereby enhancing the overall audio quality of the amplifier device.

[0067] In an example, the third feedback path is characterized by a passive second order network at the amplifier inputs H2 and H3, preferably comprising two capacitors and four resistors, feeding current to a first order transimpedance amplifier HZ2. This passive network has two different transfer functions: one seen from the input X of the amplifier and the other seen from the feedback signal Sy. This improves the stability of the overall feedback structure and maximizes the audio quality.

[0068] In an example, the input signal Sx is fed by an input buffer with very low source impedance, wherein said low source impedance buffer is comprised in said amplifier device.

[0069] This arrangement ensures that the input signal is fed into the amplifier device with minimal impedance. This is necessary because the input X is characterized by passive filter networks H2 and H3, whose transfer functions depend on the source impedance from their signal source, which should be close to 0 Ohms. The H2 and H3 networks are added to improve the loop gain and can be considered a significant feature for this purpose.

[0070] In an example, the third feedback path further comprises a transimpedance circuit HZ2 comprising an operational amplifier, a capacitor and one or two resistors. The transimpedance circuit works as a key part of the loop structure together with the passive networks H2 and H3 at the input X, thereby enabling a higher loop gain than previous designs. It effectively enhances the high order transfer function of the feedback system and thus makes the feedback path more effective in distortion cancellation. As a result, the arrangement enhances the amplifier's ability to accurately reproduce the input signal with minimal distortion, making it suitable for demanding audio applications where high quality sound reproduction is essential.

[0071] In an example, the amplifier device further comprises a clamp circuit having one input coupled to an input of HZ1 and another input coupled to an input of HZ2, and an output coupled to an output of HZ1.

[0072] The soft limiter 103 block compares the isolated error signal from 102 to a predetermined fixed voltage. Since under normal operation the error signal can be expected to be small, this is a viable way to assess whether the amplifier is working within normal operating limits. For example, the error signal can exceed the expected maximum level due to clipping of the amplifier or severe underloading. In these cases the feedback gain of the correction loop will be scaled. By hearing tests, a more aggressive approach of disabling the loop under these abnormal operating conditions was evaluated and found to be the best solution.

[0073] In another aspect, there is provided an amplifier device for amplifying an input signal Sx at a device input X to an output signal Sy at a device output Y for driving a loudspeaker or a comparable reactive load, the amplifier device comprising:

[0074] - a first transfer function K A a self-oscillating class-D power amplifier stage 10 of (s) comprising:

[0075] - a comparator 01 configured to compare a scaled version of the output signal Sy received via a first feedback path comprising a filter Hosc 06 to a local input signal SKin;

[0076] - a class-D power stage 03 coupled to and driven by an output of the comparator 01, set to convert the comparator output to a higher voltage switched output;

[0077] - a second order low pass output filter 04 comprising an inductor and a capacitor coupled between the power stage 03 and the load, configured to attenuate high frequency switching components in the output signal Sy;

[0078] - a delay unit 02 configured to implement a delay associated with one or more of the comparator 01, the power stage 03 and the filter Hosc 06, said delay being a characteristic of the self-oscillating behavior; and

[0079] - a first feedback path from the device output Y coupling the output signal Sy via Hosc 06 to the comparator 01 for determining the oscillation frequency and providing at least low frequency feedback for reducing distortion; and

[0080] - a correction stage 100 comprising:

[0081] - a second feedback path coupling the output signal Sy via a scaling element H1 104 to a feedback input of a correction stage, further coupled to a summing node S2, which is a virtual ground input of a transimpedance amplifier HZ1 102, the correction stage being configured to pre-distort the input signal SKin based on a distortion signal obtained by:

[0082] - applying a tracking filter K T 101 to the reference signal Sr; and

[0083] - subtracting the filtered reference signal from the scaled output signal from H1 104; the correction stage 100 further comprises 200:

[0084] - a third feedback path comprising:

[0085] - a passive network comprising at least a resistor and a capacitor, configured to apply:

[0086] - a first transfer function H2 from the device output Y to the output signal Sy; and

[0087] - a second transfer function H3 from the device input X to the input signal Sx; wherein the first transfer function H2 is different from the second transfer function H3,

[0088] and wherein respective outputs of H2 and H3 are added in a current domain and supplied to a virtual ground input of a transimpedance amplifier HZ2, said transimpedance amplifier being configured to generate the reference signal Sr.

[0089] In examples of the first or second aspect, the third feedback path comprises a combined feedback network configured to apply:

[0090] - a first transfer function H2 from the device output Y to the output signal Sy, and

[0091] - a second transfer function H3 from the device input X to the input signal Sx, wherein the first transfer function H2 is different from the second transfer function H3.

[0092] In examples of the first or second aspect, the transfer functions H2 and H3 are defined to be different in amplitude and / or phase at least in a portion of an audio frequency range.

[0093] In examples of the first or second aspect, the feedback network is set up to transmit a current representing a sum of signals processed by the transfer functions H2 and H3 to a virtual ground input of a transimpedance amplifier HZ2.

[0094] In examples of the first or second aspect, an output of the transimpedance amplifier HZ2 is coupled to a tracking filter KT The tracking filter K T is configured to model the transfer function of an amplifier stage.

[0095] In examples of the first or second aspect, the feedback network implementing H2 and H3 comprises passive components comprising at least a resistor and a capacitor.

[0096] In examples of the first or second aspect, the feedback network comprises four resistors and two capacitors arranged to define the transfer functions H2 and H3.

[0097] In examples of the first or second aspect, the transfer function H2 comprises a first RC filter defined by a first resistor and a first capacitor connected to receive the output signal Sy, and the transfer function H3 comprises a second RC filter defined by a second resistor and a second capacitor connected to receive the input signal Sx, and wherein the outputs of the first and second RC filters are added in the current domain and provided to the virtual ground input of the transimpedance amplifier HZ2. BRIEF DESCRIPTION OF DRAWINGS

[0098] The application will be further described with respect to specific, non-limiting examples in conjunction with the accompanying drawings, wherein:

[0099] Figure 1 A self-oscillating class-D amplifier device known from the prior art is schematically disclosed;

[0100] Figure 2 A self-oscillating class-D amplifier device according to an aspect of the disclosure is schematically disclosed;

[0101] Figure 3 Details of H2 / H3 according to an aspect of the disclosure are disclosed;

[0102] Figure 4A and Figure 4B Details of two alternatives of HZ2 according to an aspect of the disclosure are disclosed;

[0103] Figure 5 Details of H1 according to an aspect of the disclosure are disclosed;

[0104] Figure 6A and Figure 6B Details of two alternatives of HZ1 according to an aspect of the disclosure are disclosed;

[0105] Figure 7A and Figure 7B Details of two alternatives of a soft limiter are disclosed. DETAILED DESCRIPTION

[0106] Figure 1 The basic configuration of a self-oscillating class-D amplifier device known from the prior art and marketed by the applicant of the present disclosure as NCore is shown. The amplifier device comprises three main components, denoted as transfer functions K(s), H(s) and K'(s). K(s) is the transfer function of the self-oscillating output stage, K'(s) is the transfer function of a linear tracking filter having a copy of said transfer function K(s) of the output stage K but without the nonlinear distortion of K.

[0107] The input signal Sx is input at node X and is guided via a direct feedforward path to a summing node W, where the signal is combined with a signal from filter H(s) and provided to the output stage K. The output stage has a transfer function K(s) and adds a (typically small) perturbation ε to the actual output signal Sy. The actual device output signal Sy exits the control loop from output node Y to a first summing unit provided between K' and HZ, i.e. from the device output Y back to said first summing unit. This control loop together with the feedforward filter H(s) significantly improves the (low frequency) loop gain of the amplification device.

[0108] The lead-lag compensation together with the zero-crossing detector and the output filter Hout create a self-oscillating amplification unit with well-defined gain and transfer function K(s). The skilled person will understand that the internal configuration and operation of the power amplifier stage 10 is known per se and does not form part of the inventive concept described herein. In the block 10 a schematic representation is provided, while more details can be found in the amplifier system demonstrated in US 7,113,038 B2 and EP 2 221 964 B1. Figure 2 The lead-lag compensation together with the zero-crossing detector and the output filter Hout create a self-oscillating amplification unit with well-defined gain and transfer function K(s). The skilled person will understand that the internal configuration and operation of the power amplifier stage 10 is known per se and does not form part of the inventive concept described herein. In the block 10 a schematic representation is provided, while more details can be found in the amplifier system demonstrated in US 7,113,038 B2 and EP 2 221 964 B1.

[0109] In the shown amplifier device, the tracking filter K' with transfer function K'(s) and the subsequent summing unit serve to isolate the distortion component ε originating from the power stage K and its load.

[0110] K' receives the input signal Sx from the device input node X. Since the transfer function K'(s) emulates the transfer function K(s) of the power stage K, the output of the tracking filter K'(s) is an idealized version of Sy without the error term ε. The first subtraction unit on the far left subtracts the output signal Sy from the idealized version from the detection filter of the tracking filter received via the control loop. Thus, the feedforward filter H(s) receives only (inverted) perturbations ε, i.e. the difference between the signal received via the control loop and the signal received from the detection filter. In normal conditions, the perturbations ε are considered small, so the output of the feedforward filter will also be small. H(s) is a low pass function and an amplifier precisely configured to maximize the gain in the audio band and at the same time maintain the correct conditions for oscillation. In the summation node W, the isolated and inverted error is added to the input signal of K(s), thus pre-distorting the input signal with the distortion component of K and the inverse of its load. In embodiments, the tracking filter K' is implemented as a direct RC filter.

[0111] Figure 2 An amplifier device 300 according to embodiments of the present disclosure is shown. The device 300 can be represented in three different blocks, identified as:

[0112] - a self-oscillating power amplifier device 10;

[0113] - a power amplifier enhanced by a first generation correction loop 100;

[0114] - a power amplifier further enhanced by a second generation correction loop system 200 of higher order.

[0115] The figure can be best explained and described from right to left, as this best illustrates the evolution of the technology. Three red boxes can be identified, each with a title that roughly covers a specific stage of development of the technology. These are UcD 10, NCore 100 and Nilai 200. The new generations are constantly developed on the basis of the previous one, each stage of technology being a subset of the latest development.

[0116] Regarding the amplifier stage UcD, this is not really a high-end amplifier yet, but can provide significantly good performance with relatively simple technology. The whole amplifier can be considered as a comparator capable of delivering a large amount of current, followed by a simple LC low-pass filter and a feedback mechanism. The LC filter 04 greatly attenuates the high-frequency components of the pulse-width signal from 03, leaving the low-frequency signal for the loudspeaker. This LF signal (via Hosc) is fed back to the input of the comparator and scaled before comparison with the input signal. Based on the result of this comparison, the state of the comparator is adjusted to integrate the output signal in the correct direction, thereby reducing the difference between the scaled output signal Sy and the local input signal SKin. With some imagination, the operation of the comparator 01, the LC filter 04 and the feedback can be considered as a linear amplifier with feedback. The output LC filter (with load), the feedback network and the comparator itself introduce a certain delay in the described control loop. This delay determines the switching frequency of the self-oscillating system, which is approximately 500 kHz in the absence of an input signal. In addition, the switching frequency will vary depending on the actual state of the audio signal, decreasing when the excursion is higher. The advantages of the UcD 10 are mainly simplicity and robustness, but the simple control loop has a limited gain and therefore does not compromise on sound quality.

[0117] In the next block Ncore 100, an additional control loop is provided via Hl 104. Blocks 101 and 102 are also added. The strategy here is to isolate the low-frequency distortion component of the UcD part (also called K10 in the block diagram) and subtract it from the input signal (from R) to cancel the distortion. The input signal to the power stage 10 is thus "pre-distorted" and has the inverse of its own distortion. The output signal Y is fed back via Hl 104 to be summed with the inverse of the signal from 101 at summing point S2. Thus, after filtering in 101, the local input signal Sr is subtracted from the scaled output signal Y. 104 is merely a resistor responsible for the correct attenuation of Sy, while 101 is a filter network trying to model the transfer function of the UcD part 10 as accurately as possible. Thus, if 10 provided a perfect transfer, i.e. no distortion, the result of this subtraction would always be 0. In practice, however, there will be an isolated distortion component of 10 here. 102 is a transfer function and gain stage configured precisely to provide a high gain in the audio band while preserving the oscillation condition of 10. We only want to subtract the low-frequency part of the error from the input signal and optimize the loop gain in the audio passband. "HZ" indicates that this is a transimpedance circuit with virtual ground as input, so it can easily be used as a summing point.

[0118] Next, in S1, the input signal to the power stage 10 is pre-distorted by subtracting the error from the input signal Sr, with its inverse isolated and filtered. The degree to which it succeeds in isolating the distortion component depends exactly on the good tracking of 101 with 10, and the loop gain in the outer loop that can be obtained without compromising stability. In theory, this means that the higher the gain the better. Although the 102 function already exists in NCore, this function is modified for Nilai, and becomes slightly more complex than the variant that exists in NCore. See in particular the attached Figure 6A and 6B .

[0119] NCore has been significantly improved over UcD, but further improvements are considered desirable and feasible for those who pursue the high end. This is achieved by further increasing the loop gain of the control loop. For this, 201 and 202 are added. 202 is again a transimpedance amplifier. H2 and H3 201 can be interpreted as separate blocks, however, since both are combined in the same passive network, they are shown as a single block with two inputs; one of the two is the actual amplifier input, the other receives the output signal of the amplifier for feedback. The circuit is characterized by different transfer functions in either signal path.

[0120] The open loop response is that of a finely tuned low pass filter, while the closed loop response in the audio band is all pass, but it does add phase margin in the control loop, allowing for greater gain without the risk of the entire circuit oscillating at undesirable frequencies. The entire structure that now emerges is like a complex control loop that must be designed and evaluated as a whole. The output of 102 is the point at which the correction signal to isolate and invert the distortion component is available. The A1 circuit leading to the loop input of S1 can be considered a single outer loop with a single loop transfer function. Although Figure 2 the block diagram in can suggest a separate second feedback loop via 104 and a third feedback loop via 201, the entire structure can be seen as a single, higher order correction loop, with the summing point S1 as the actual point at which the distortion component of 10 is eliminated. The circuit in 201 and 202 controls the overall feedback in the amplifier device and adds additional feedback gain. Careful distribution of loop gain between 102 and 201 has a large impact on the perceived sound quality, as can be seen from the hearing tests. Determining the size of these blocks is therefore an important part of the design philosophy.

[0121] The amplifier block 200 has a soft limiting circuit 103. In Figure 7A and Figure 7BTwo different embodiments of this level detection circuit can be found in the middle. This circuit is advantageous to prevent the high gain feedback loop system from overreacting to undesired operating conditions such as clipping of the power stage 10 or severe underload.

[0122] Some amplifiers use hard clipping detection at the power amplifier to completely disable the feedback system. In 200, a choice is made to monitor 102 the isolated error at the output. Since in normal operation the signal level at this point is only small, a viable solution is to compare the error signal to a predetermined fixed voltage. If the error voltage suddenly exceeds this expected maximum, the circuit will proportionally limit the gain in HZ1 and HZ2 in a tracking manner. A listening test at high sound levels with this limitation seems to sound more "relaxed" than the hard limiting approach.

[0123] Figure 3 A filter network 201 is shown. This circuit controls the overall feedback in the amplifier together with a transimpedance amplifier 202 and adds extra low frequency loop gain to effectively cancel out distortion components. The network consists of 2 capacitors and 4 resistors and enables 2 transfer functions in the same circuit, one in the feedback loop system and the other from a feedforward from input X. The signal Sx is fed from the input buffer 301 and has a low impedance.

[0124] Figure 4A An embodiment of the transimpedance amplifier 202 is shown. This configuration is a first order low pass filter with a virtual ground current input, intended to work in conjunction with 201.

[0125] Figure 4B Another embodiment of 202 is shown, this time featuring a shelving filter function.

[0126] Figure 5 A schematic of H1 104 is shown. This is just a resistor, intended to properly scale the feedback signal from the device output Y to the summing point S2 as a virtual ground.

[0127] Figure 6A An embodiment of the transimpedance amplifier HZ1 102 is shown. This model features an operational amplifier, 3 resistors and 3 capacitors.

[0128] Figure 6B Another embodiment of the same circuit 102 is shown, this time with the intermediate capacitor removed. The impact of this capacitor in the actual circuit is very small and only significant at very high frequencies. Ultimately, it was decided that this part could be completely omitted without significant impact on the overall loop performance. The key is the tuning of the circuits in 201 and 202.

[0129] Figure 7AAn embodiment of the level detector in 103 is shown. If the input signal from the output of 103 exceeds a predetermined voltage level, the circuit will feed current to the virtual ground inputs of 102 and 202 in a proportional manner, thus limiting the gain of the loop.

[0130] Figure 7B A second embodiment of the same circuit 103 is shown, but with the addition of extra resistors in the emitter of the transistors providing more control of the soft limiting behavior of the circuit.

Claims

1. An amplifier device for amplifying an input signal (Sx) of a device input (X) into an output signal (Sy) at a device output (Y) for driving a loudspeaker or a comparable reactive load, said amplifier device comprising: - having a first transfer function K A a self-oscillating class-D power amplifier stage (10) of the type (s) comprising: - a comparator (01) comparing its output signal (Sy) with a local input voltage (SKin) via a first feedback filter Hosc (06); - a class-D power stage (03) coupled to and driven by the output of the comparator (01) converting the output voltage of the comparator into a high voltage level with high current capability; - a second order low pass output filter (04) comprising an inductor and a capacitor (04) coupled between the power stage (03) and the loudspeaker or comparable load for removing at least part of the high frequency switching noise from the output signal of the class-D power stage; - a time delay unit (02) configured to provide a time delay depending on one or more of the comparator, Hosc and the power stage; - a first feedback path from the device output Y coupling the output signal (Sy) back to the comparator (01) via Hosc (06), said first feedback path providing control over the oscillation frequency and at least partly providing low frequency feedback to reduce distortion of the signal in the audio frequency band; and - a correction stage comprising: - a second feedback path coupling the output signal (Sy) from the device output (Y) via H1 (104) to a feedback input of the correction stage, further to a summing node (S2) which is a virtual ground input of a transimpedance amplifier HZ1 (102) for further correction of errors of the output signal (Sy) generated by the amplifier stage, wherein the further correction is performed by pre-distorting the power amplifier input signal (SKin) of the amplifier stage (10) with the inverse of the isolated and filtered distortion components caused by the power amplifier stage, the output filter, and the loudspeaker or the comparable load, wherein the isolated distortion components are obtained by passing an incoming signal (Sr) through a tracking filter K T (101) running, and subsequently subtracting the result from the feedback signal from the output of the H1 (104), as a scaled version of the output signal (Sy); the correction stage further comprising: - a third feedback path coupling the output signal (Sy) to an additional feedback input via a first transfer function (H2) + a second transfer function (H3) and through a transimpedance amplifier HZ2 of the correction stage to further enhance the correction of the output signal (Sy) by the second feedback path, wherein the third feedback path is configured as a higher order additional for enhancing the selectivity of the feedback path and for providing additional feedback gain, and wherein the third feedback path consists of a network implementing the first transfer function (H2) applied to the output signal (Sy) and the second transfer function (H3) applied to the input signal (Sx), the first transfer function (H2) being different from the second transfer function (H3), the first and second transfer functions (H2, H3) being combined within a single passive network comprising resistors and capacitors, and wherein the respective outputs of the first and second transfer functions (H2, H3) are summed and supplied as a current input to the transimpedance amplifier HZ2, the output of which defines the incoming signal (Sr).

2. The amplifier device of claim 1, wherein, the second feedback path comprises a feedback network of a minimum order of two, preferably of three, more preferably of four, capacitors and three, more preferably of four, resistors, said feedback network being configured to enhance error correction and to further increase the low frequency loop gain.

3. The amplifier device of claim 1 or 2, wherein, The third feedback path further comprises an input / feedback summing function via a summing network of a first transfer function (H2) + a second transfer function (H3) of a capacitor and a resistor, the input / feedback summing function configured to add an input signal (Sx) with an output signal (Sy) fed back from the output (Y) in a frequency dependent manner, wherein the summed output is fed as a current to a virtual ground input of a transimpedance amplifier HZ2.

4. The amplifier device of claim 3, wherein, The summing function further comprises a transimpedance circuit HZ2 comprising an operational amplifier, one or two resistors, and a capacitor.

5. The amplifier device according to any of the preceding claims, further comprising a soft limiting circuit having an input coupled to the output of (102), an output coupled to an input of HZ1, and a further output coupled to an input of HZ2, the soft limiting circuit being effective to proportionally limit the gain of the correction loop structure if the voltage level at the output of the HZ1 (102) exceeds a predetermined voltage level.

Images