Patent Application
20100194413
The present invention relates to determining the impedance function of a loudspeaker.
Knowledge of the impedance function of a loudspeaker connected to an amplifier can be used for several purposes, and hence determination thereof is desirable. When knowing the impedance function it is possible to, e.g., perform compensating equalization, adjust limiters, avoid providing damaging power to the loudspeaker, etc. Furthermore, live monitoring of the impedance function can be used to track temperature changes in the loudspeaker components, monitor the wear and aging of the loudspeaker, etc.
Conventional methods disclosed in the prior art comprises measuring the voltage and current at the power output of the amplifier, and calculating the impedance function from these two measurements. An amplifier comprising such measuring and calculating means is described in U.S. Pat. No. 5,719,526, where voltage and current are measured at the power output signal, converted into digital representations, and an impedance function calculated by a digital signal processor.
In some amplifier implementations it may however be a problem or at least an unnecessary cost to provide high-quality A/D-converters in order to be able to process the measurements in the digital processing means. On the other hand, it is impossible to implement contemporary, fast and high-resolution impedance function calculation and analysis thereof in the analog domain. Another issue is that because of the delays in the forward path of contemporary amplifiers, it may in some cases be impossible to react in time on an extreme measurement performed at the end of the path, because of the parts of the signal that has already been provided to the amplifier path from the processing means.
An object of the present invention may therefore be to reduce the amount of analog components, e.g. A/D-converters, needed in order to calculate an impedance function of a load connected to an amplifier.
An object of the present invention may be to improve the centralization in a contemporary amplifier comprising both digital and analog components.
An object of the present invention may be to estimate a representation of the output of an amplifier prior to its actual production and sufficiently early to perform critical actions on the basis thereof.
The present invention relates to a method of determining an impedance function IF of a load LS driven by an amplifier AM, said method comprising the steps of providing a digital audio signal DAS to said amplifier AM,
measuring one of either a current signal representation CSR of current provided to said load LS by said amplifier AM or a voltage signal representation VSR of voltage provided to said load LS by said amplifier AM,
determining a digital signal representation DSR on the basis of said digital audio signal DAS, and
determining said impedance function IF of said load LS on the basis of said digital signal representation DSR and said measured one of either said current signal representation CSR or said voltage signal representation VSR.
According to the present invention is provided a method whereby impedance function calculation of a load can be performed by only one analog measurement and converter, i.e. the current measurement for a traditional amplifier or a voltage measurement for a current amplifier, thereby saving an expensive high-quality and fast A/D-converter solution.
Furthermore the present invention facilitates knowledge of the output signal amplitude, or current for a current amplifier, before it actually happens at the output, because there is a considerable delay in the amplifier means, in particular if it comprises an A/D-converter, causing the digital signal representation DSR to be established up to e.g. 1 ms before its corresponding power amplified analog representation is produced at the amplifier output.
Moreover, the present invention facilitates a direct, digital connection between the digital audio signal and the signal processing applied to that, and the digital impedance calculating circuitry. Thereby is facilitated using a single digital signal processor or other suitable digital processing means for both purposes. If distributed processing is desired, the present invention facilitates avoiding input from the analog domain in an even higher degree than previously known.
When said method is carried out during operation of said amplifier AM, an advantageous embodiment of the present invention is obtained.
According to a very preferred embodiment of the present invention, the impedance function of the load can be determined at any time, even during normal use with an arbitrary input signal just fulfilling a few criteria regarding its frequency spectrum.
When said step of determining said impedance function IF is performed by digital signal processing means MM, DSP, an advantageous embodiment of the present invention is obtained
When said measured one of either said current signal representation CSR or said voltage signal representation VSR is converted into a digital representation by means of an analog-to-digital converter ADC, an advantageous embodiment of the present invention is obtained.
When said method comprises adding a delay to said digital signal representation DSR in order to establish synchrony between said determined digital signal representation and said measured current signal representation, an advantageous embodiment of the present invention is obtained.
In a very preferred embodiment of the present invention a delay is applied to the digital signal representation in order to avoid determining an impedance function on the basis of signals that are not synchronized and would therefore cause the result to be invalid. It is noted that the delay, within the scope of the present invention, may be applied at any suitable step in the processing chain, e.g. at the digital reading point, in the impedance calculation circuit ICP, etc. If implemented in the impedance calculation circuit, it could merely be established by means of a buffer of a suitable length on the digital signal representation input.
When said delay comprises a delay corresponding to a delay of said amplifier AM and a delay of said analog-to-digital converter ADC, an advantageous embodiment of the present invention is obtained.
When said method comprises a step of performing compensation signal processing of said digital signal representation DSR, an advantageous embodiment of the present invention is obtained.
In a preferred embodiment of the present invention, compensation processing is applied to the digital signal representation, in order to establish a signal that better resembles the output signal of the amplifier in embodiments where the amplifier applies an error to the output signal in an area significant to the impedance function calculation.
When said compensation signal processing is performed in accordance with an amplification means model AMM comprising information about said amplifier AM, an advantageous embodiment of the present invention is obtained.
When said amplification means model AMM comprises information about the delay of said amplifier AM, an advantageous embodiment of the present invention is obtained.
When said amplification means model AMM comprises information about the DC gain of said amplifier AM, an advantageous embodiment of the present invention is obtained.
When said amplification means model AMM comprises information about the frequency-dependent delay of said amplifier AM, an advantageous embodiment of the present invention is obtained.
When said compensation signal processing is performed in accordance with an amplification means model AMM comprising information about an output impedance of said amplifier AM, an advantageous embodiment of the present invention is obtained.
According to an embodiment of the present invention, information about the amplifier's output impedance may e.g. comprise information about an output filter of said amplifier.
When said amplification means model AMM comprises information about the transfer function of said amplifier AM, an advantageous embodiment of the present invention is obtained.
When said amplification means model AMM comprises information about said amplifier AM for a predefined frequency band, preferably the audio band, an advantageous embodiment of the present invention is obtained.
When compensation signal processing is performed in accordance with an amplification means model AMM comprising information, e.g. DC resistance, impedance, etc., about a cable connecting said amplifier AM with said load LS, an advantageous embodiment of the present invention is obtained.
When said amplification means model AMM is calibrated on a regular basis, an advantageous embodiment of the present invention is obtained.
In a preferred embodiment, the amplification means model AMM is calibrated or verified on a regular basis in order to reflect variations and fluctuations of amplifier parameters in the impedance function calculation procedure. Any way of determining the variations or fluctuations, i.e. manually, semi-automatically or fully automatically, by means of test signals or “live” signals, etc., are within the scope of the present invention.
When said calibration of said amplification means model AMM is performed on the basis of voltage or current measurements at the output of said amplifier AM of a reproduced test signal at each start-up and/or at user-specified times, an advantageous embodiment of the present invention is obtained.
When said step of determining said digital signal representation DSR on the basis of said digital audio signal DAS comprises reading a digital value from a register or buffer, an advantageous embodiment of the present invention is obtained.
When said amplifier AM comprises a voltage amplifier, an advantageous embodiment of the present invention is obtained.
When said amplifier AM comprises a current amplifier, an advantageous embodiment of the present invention is obtained.
The present invention further relates to a load monitoring amplifier comprising amplification means AM comprising an amplifier input AI for receiving a digital audio signal DAS and an amplifier output AO for delivering an amplified signal to a load LS and an analog reading point AR establishing one of either a current signal representation CSR by measuring the current of said amplified signal delivered to said load LS or a voltage signal representation VSR by measuring the voltage of said amplified signal delivered to said load LS, said load monitoring amplifier further comprising a digital reading point DR for determining a digital signal representation DSR on the basis of said digital audio signal DAS and a monitoring means MM for determining an impedance function IF of said load LS on the basis of said digital signal representation DSR and said one of either said current signal representation CSR or said voltage signal representation VSR.
The present invention provides an advantageous amplifier that is able to determine the impedance function of a connected load on the basis of only one analog measurement. This is particularly advantageous in contemporary amplifiers with digital processing and often even so-called digital amplification (class-D amplifiers), as all of the information and processing means are available in the digital domain of the amplifier, except from a single analog measurement. Hence, it is possible with the amplifier of the present invention to only measure the output current for traditional voltage amplifiers, or the output voltage for current amplifiers, in order to determine the impedance function and all the information derivable when the impedance function is determined.
It should be noted that even though the amplifier input AI is said to receive a digital audio signal DAS, a product comprising an amplifier according to an embodiment of the present invention can evidently within the scope of the invention comprise inputs for analog signals, e.g. RCA-connectors, BNC-connectors, etc., followed by suitable A/D conversion means in order to establish a digital audio signal DAS. This is further described with reference to
When said monitoring means MM comprises an analog-to-digital converter ADC to convert said one of said current signal representation CSR or said voltage signal representation VSR into a digital representation and an impedance calculation circuit ICP for determining said impedance function IF, an advantageous embodiment of the present invention is obtained.
When said monitoring means MM comprises delay means DM for adding a delay to said digital signal representation DSR, an advantageous embodiment of the present invention is obtained.
When said delay means comprises a delay corresponding to a delay of said amplification means AM and said analog-to-digital converter ADC, an advantageous embodiment of the present invention is obtained.
When said monitoring means MM comprises an amplification means model AMM in accordance with which the digital signal representation DSR is processed before used for impedance function determination, an advantageous embodiment of the present invention is obtained.
When said amplification means model AMM comprises information about said amplification means AM, such as delay, frequency dependent delay, DC-gain, frequency dependent gain, non-linearities, output impedance, transfer function, etc., an advantageous embodiment of the present invention is obtained.
When said amplification means model AMM comprises information, e.g. DC resistance, impedance, etc., about a cable connecting said amplification means AM with said load LS, an advantageous embodiment of the present invention is obtained.
When said load monitoring amplifier further comprises a signal processor SP for processing said digital audio signal DAS, and wherein said signal processor SP and said impedance calculation circuit ICP is comprised in a digital signal processor DSP or other digital processing means, an advantageous embodiment of the present invention is obtained.
When said load monitoring amplifier further comprises a digital register or buffer from which said digital signal representation DSR may be read on the basis of said digital audio signal DAS, an advantageous embodiment of the present invention is obtained.
When said amplifier AM comprises a voltage amplifier, an advantageous embodiment of the present invention is obtained.
When said amplifier AM comprises a current amplifier, an advantageous embodiment of the present invention is obtained.
An invention further relates to an amplifier compensation circuit AC comprising a filter with a transfer function resembling the reverse of the difference between the transfer function of a subsequent amplification means and a predefined transfer function.
When said filter is adjustable, an advantageous embodiment of the present invention is obtained.
An invention further relates to an amplifier comprising an amplification compensation circuit AC according to the above and an amplification means AM.
When said amplifier comprises monitoring means MM for determining an impedance function of a load LS connected to said amplifier and means for adjusting said filter on the basis of said impedance function, an advantageous embodiment of the present invention is obtained.
When said amplifier comprises means for determining the class of a load LS connected to said amplifier and means for adjusting said filter on the basis of information related to said class of said load LS, an advantageous embodiment of the present invention is obtained.
When said amplification means AM comprises an output filter OF, an advantageous embodiment of the present invention is obtained.
When said impedance function is determined according to a method of determining an impedance function according to any of the above or by means of a load monitoring amplifier according to any of the above, an advantageous embodiment of the present invention is obtained.
The invention will in the following be described with reference to the drawings where
To establish an impedance function associated with the loudspeaker is in principle needed measurements of the voltage and the current supplied to the loudspeaker. As described above, it is well-known to simply measure these representations at the amplifier output AO, convert them to digital signals, and calculate the impedance function by digital processing means. The present invention, however, requires with the embodiment of
Even without knowing the gain of the amplification means, it is thereby possible to determine a relative or normalized impedance function on the basis of the digital signal representation DSR and an analog-to-digital converted version of the current signal representation CSR. A normalized impedance function suffices for several purposes, e.g. for frequency dependent impedance function analysis, recognition of impedance function characteristics and feature extraction, etc., which may, e.g., be used for identifying the type or model of loudspeaker, determining the temperature of internal loudspeaker components, etc.
In a more advanced embodiment of the present invention, the gain of the amplification means is known by the monitoring means MM, and it is thereby possible to determine the absolute impedance function of the loudspeaker. The absolute impedance function may be used for the same purposes as described above, and for further purposes requiring information about absolute impedances, e.g. for determining the number of loudspeakers coupled to the amplifier output AO in parallel.
In real amplifiers, the amplification means comprises not only a gain, but also a delay and a transfer function often causing less gain at in particular very low and very high frequencies. Also non-linear distortion exists to some, however low, degree in the amplification means. Hence, the presumption that a normalized or absolute impedance function can be calculated from the digital signal representation derived prior to the amplification means, is not true if a very accurate impedance function for in particular low and high frequencies is desired. In such cases, and depending on the degree of accuracy desired or required, the digital signal representation DSR may be processed before use in the impedance calculation to compensate for some of the above errors. Embodiments of the present invention covering this aspect are described in more detail below.
It should be noted that any other implementation of signal processing and digital reading point, and any combination of the above-described features, is within the scope of the present invention.
It should be noted that any representation or format of the amplifier output AO is within the scope of the present invention, e.g. single-ended or balanced outputs.
DAS into an analog audio signal for the analog power stage AA.
It should be noted that any other amplification means implementation or combination of above-described features is within the scope of the present invention.
In a more advanced embodiment, the delay means DM adds a frequency dependent delay, as the delay added to the audio signal by the amplification means is often frequency-dependent, i.e. is different for different frequencies.
In a preferred embodiment of the invention, the amplification means model AMM is extended to also include a model of the loudspeaker cable, or significant errors related to the loudspeaker cable. In loudspeaker setups with relatively long cables the impedance of the cable, in particular it's DC resistance, becomes significant compared to the loudspeaker impedance, and will thus influence the impedance calculation significantly. A certain loudspeaker cable of 40 meters may for instance add a resistance of 1Ω (Ohm), and as the analog reading point AR in any practical case is located at the amplifier's end of the loudspeaker cable, the impedance function calculated will be an impedance function of the combined loudspeaker cable and loudspeaker. By compensating for the cable impedance in the extended amplification means model AMM, calculation of the loudspeaker impedance is facilitated, even with long, non-ideal cable connections.
The establishment of a cable model or an estimate of the most significant errors introduced by the cable may, e.g., be made by allowing the user to input cable characteristics such as cable length, cross section and resistivity into the processing means by means of a user interface. Alternatively, an amplifier with impedance calculation for example according to the present invention can be used to estimate the cable impedance by shorting the cable at the loudspeaker end during measuring, and subsequently establish a cable model to include in an extended amplification means model AMM from the measurements. Alternatively, as a neglected, significant cable resistivity will typically make a calculated impedance function indicate a very hot loudspeaker, the amplifier may provide a user interface means for providing to the processing means the information that the loudspeaker is definitely not hot, and the impedance features indicating a hot loudspeaker should instead be considered as cable impedance and, e.g., regarded as a cable model for subsequent measurements.
The amplification means model AMM may be established by measurements at the time of manufacture of the amplifier, or it may be configurable or adjustable in order to change with any changes of the amplification means AM over time. In an advanced embodiment, the transfer function, or significant characteristics thereof, of the amplification means is measured at each start-up or at user-defined times, and the result is used to calibrate the amplification means model AMM. For this purpose the amplifier may comprise means for measuring the voltage of the amplifier output signal, and an A/D-converter to provide this signal to the amplification means model AMM for calibration purposes. It is noted, however, that such voltage measurement does not require the same degree of quality, e.g. in regard to the A/D-converter, as if it is used for runtime impedance calculation as described in the prior art, as timing is not an important issue in a calibration situation.
In an advanced embodiment, the monitoring means further comprises means for analysing the digital signal representation, possibly after part of the amplification means model processing has been carried out, but before the delay has been added. Thereby is established a representation of the output signal, or at least the amplitude thereof, a considerable time, e.g. 0.25 or 1 ms, before it actually happens at the output. This time is sufficient to perform some degree of analysis and in case of critical results thereof, e.g. excessive power output, coarse clipping, etc., perform actions to avoid or reduce damage to the loudspeaker or unpleasant sound reproduction. The knowledge about the output signal before it happens could obviously also be used for non-critical purposes such as compensation, fine-tuning the signal processing, etc.
In yet an advanced embodiment, once the impedance function of the loudspeaker is calculated accurately, or when an accurate impedance function can be determined or established beforehand, it is possible to use the monitoring means for calculating the current of the amplifier output signal on the basis of the output voltage estimated from the digital signal representation, and the impedance function determined previously. Hence, it becomes possible to estimate both voltage and current of the power output signal before it actually happens and react accordingly. For this purpose, the current signal representation measurement and associated A/D-converter then become irrelevant.
In a preferred embodiment of the invention, the impedance calculation circuit ICP comprises windowing in the time domain of the input signals, and/or weighted averaging of the calculated impedance in order to establish a good estimate of the impedance function, and in order to avoid impedance functions calculated at uncertain signals or under uncertain conditions, e.g. during clipping, to influence the established impedance function significantly.
In a preferred embodiment, the impedance calculation circuit ICP comprises a multirate fast fourier transform FFT algorithm in order to establish impedance functions in relevant time windows, but any method of estimating or calculating an impedance function on the basis of the digital signal representation DSR and a voltage signal representation VSR or a current signal representation CSR is within the scope of the present invention.
The analog reading point AR is in the present embodiment of the invention measuring a voltage signal representation VSR of the voltage provided via the amplifier output AO to the load or loudspeaker LS by the current amplifier amplification means AM. The analog reading point may comprise any suitable means for determining voltage. Numerous methods for voltage measurements are described in the prior art, and any method suitable for use at a sensitive, amplified audio signal, is within the scope of the present invention. The voltage signal representation VSR provided by most of the possible voltage measurement methods is an analog representation, but any representation is within the scope of the present invention.
To establish an impedance function associated with the loudspeaker is in principle needed measurements of the voltage and the current supplied to the loudspeaker. As described above, it is well-known to simply measure these representations at the amplifier output AO, convert them to digital signals, and calculate the impedance function by digital processing means. The present invention, however, requires with the embodiment of
Even without knowing the exact current gain of the current amplification means, it is thereby possible to determine a relative or normalized impedance function on the basis of the digital signal representation DSR and an analog-to-digital converted version of the voltage signal representation VSR. A normalized impedance function suffices for several purposes, e.g. for frequency dependent impedance function analysis, recognition of impedance function characteristics and feature extraction, etc., which may, e.g., be used for identifying the type or model of loudspeaker, determining the temperature of internal loudspeaker components, etc.
In a more advanced embodiment of the present invention, the gain of the amplification means is known by the monitoring means MM, and it is thereby possible to determine the absolute impedance function of the loudspeaker. The absolute impedance function may be used for the same purposes as described above, and for further purposes requiring information about absolute impedances, e.g. for determining the number of loudspeakers coupled to the amplifier output AO in parallel.
In real amplifiers, the current amplification means comprises not only a current gain, but also a delay and a transfer function often causing less current gain at certain frequency bands. Also non-linear distortion exists to some, however low, degree in the current amplification means. Hence, the presumption that a normalized or absolute impedance function can be calculated from the digital signal representation derived prior to the amplification means, is not true if a very accurate impedance function is desired. In such cases, and depending on the degree of accuracy desired or required, the digital signal representation DSR may be processed before use in the impedance calculation to compensate for some of the above errors. As a digital signal processor or likewise digital processing means are inherently available, it is possible to implement more advanced compensation processing for current amplifiers without additional circuitry or logics.
Thus, a typical aim of an amplification means AM is to apply a transfer function as, e.g., illustrated in
In other words, it will typically not be possible to provide an output impedance OF or an amplification means AM that applies the same, ideal low pass transfer function 81 for any load LS. In
The present invention related to
The establishment of a suitable amplifier compensation circuit AC can initially be done by the amplification means manufacturer for a nominal load impedance. In a preferred embodiment, the amplifier compensation circuit AC is, however, adjustable or adaptive in order to automatically, semi-automatically or manually adapt to compensate for changes in the amplification means transfer function due to change in load impedance.
According to a preferred embodiment, a monitoring means MM according to the present invention described above is applied in order to establish an impedance function of the load. When knowing the impedance function of the load, the amplifier compensation circuit AC can be adapted to the resulting transfer function of the amplification means.
In an embodiment of the invention, the load impedance calculation and the adjustment of the amplifier compensating circuit AC have to be done iteratively because the adjusted amplifier compensating circuit AC also changes the impedance calculation regarding the load. In other words, as the transfer function of the amplification means and the impedance measured between the amplifier output AO over the load depends on each other, the adjustment of the compensation has to be done iteratively.
In an embodiment of the invention, the monitoring means MM comprises means for identifying the load or class of load on the basis of the calculated impedance or other load characteristics, or by means of user input, and on the basis of the then known load impedance the compensation circuit AC can be adjusted in one step.
In an embodiment of the invention, the amplifier compensation circuit AC is adapted to be a means by which the amplifier transfer function can be shaped to any desired form.
The present invention moreover facilitates using output filters in the amplification means which are tuned and optimized regarding noise attenuation or other purposes instead of the typical aim of establishing an ideal transfer function, as this problem is by the present invention handled in the amplifier compensation circuit.
| Filing Document | Filing Date | Country | Kind | 371c Date |
|---|---|---|---|---|
| PCT/DK07/50099 | 7/16/2007 | WO | 00 | 1/18/2010 |
