This invention relates to electronic circuits, and more specifically to a mixed-signal filter.
In the field of communications, particularly wireless voice communications, there is a constant effort to manufacture consumer electronics that allow a clearer signal, such that participants in a communication session can better understand each other. As a result, an incoming communication signal can be filtered to remove unwanted signal power within a given frequency band. As an example, electronic noise and/or interferer signals at frequencies both known and unknown can contribute to such unwanted signal power. Thus, the incoming communication signal can be filtered to attempt to remove the unwanted signal power at specific frequencies in the frequency spectrum of the communication signal.
As the incoming communication signal is first received in the analog domain, some typical filter systems attempt to filter the signal in the analog domain. However, to filter more specific frequencies in the frequency spectrum of a communication signal, analog filter systems can be expensive and bulky, and can involve extremely complicated algorithms to allow them to be tunable to the specific frequency ranges in the frequency spectrum. As incoming communication signals are often converted to the digital domain, some typical filter systems attempt to filter the system in the digital domain. However, while digital filters are typically inexpensive and often involve much simpler algorithms, digital filters may be required to implement extremely high resolution to maintain signal quality of an incoming signal, and can also require an extremely high dynamic range of an upstream analog-to-digital converter (ADC) to be able to receive the incoming signal without losing modulated information. A high dynamic range ADC using existing design techniques is prohibitively expensive, power hungry and bulky, particularly for low-cost and low-power applications, such as mobile stations.
Another manner of filtering that is typically implemented is an adaptive filter algorithm. An adaptive filter algorithm is an algorithm that continuously monitors an output of the adaptive filter to estimate a location of interferer signals. The interferer signals can thus be cancelled based on subtracting power at the estimated locations as the signal is subsequently received. Adaptive filter algorithms, however, are typically very complicated, and can suffer from time delays in filtering the interferers and/or noise, and can be unstable based on the variations and reactions of the filter output. In addition, estimation errors can result in filtering portions of the signal of interest, or possibly undesirably boosting noise and/or interferers based on a lack of convergence of the filter algorithm. As a result, the continuously varying transfer function of the adaptive filter algorithm may not result in accurate filtering of unwanted signal power from a frequency spectrum of an incoming communication signal.
One embodiment of the present invention includes a mixed-signal filter. The mixed-signal filter comprises an analog signal path configured to process a first analog signal and an analog-to-digital converter (ADC) configured to convert the processed first analog signal to a digital signal. The mixed-signal filter also comprises a programmable digital feedback filter configured to filter the digital signal and a digital-to-analog converter (DAC) configured to convert the filtered digital signal to a second analog signal. The mixed-signal filter further comprises a signal combiner configured to combine an analog input signal of the mixed-signal filter and the second analog signal to generate the first analog signal.
Another embodiment of the present invention includes a method for filtering a first analog signal. The method comprises combining the first analog signal with a second analog signal to generate a combined analog signal and processing the combined analog signal in the analog domain. The method also comprises converting the processed combined analog signal to a digital signal and filtering the digital signal in the digital domain based on a predetermined digital filter transfer function. The method further comprises converting the filtered digital signal to the analog domain to generate the second analog signal.
Another embodiment of the present invention includes a mixed-signal filter. The mixed-signal filter comprises means for combining an analog input signal and a feedback signal to generate an analog combined signal. The mixed-signal filter also includes means for processing the analog combined signal to generate an output signal and means for filtering the output signal in the digital domain to generate the feedback signal based on a predetermined transfer function. The mixed-signal filter further includes means for programming the means for filtering to generate the predetermined transfer function.
By way of further example, in the above embodiments, the mixed-signal filter can be considered a mixed-signal servo loop with negative feedback. The overall closed loop response of the servo loop depends on transfer functions of analog and digital components, as well as signal and noise transfer functions of A/D and D/A converters. The combiner output in such embodiments corresponds to the error signal of the servo loop.
The present invention relates to electronic circuits, and more specifically to a mixed-signal filter. The mixed-signal filter receives an analog input signal, such as a downconverted and/or demodulated radio frequency (RF) signal received at an antenna. The analog signal is combined at a signal combiner with a filtered feedback signal. The combined analog signal is provided to an analog signal path that processes the combined signal. The analog signal path can include an analog filter and/or an amplifier to implement the processing of the combined signal. The processed analog signal is converted to the digital domain and input to a programmable digital feedback filter. The programmable digital feedback filter can be set to adjust zeros, poles, and/or gain of the frequency spectrum of the digital signal, such that a transfer function of the mixed-signal filter can be predetermined. The filtered digital signal is converted to the analog domain and is combined with the analog signal upstream from the analog signal path. As a result, a simple analog filter can be combined with a simple digital filter algorithm to generate a static, predetermined transfer function of the filter. Therefore, the mixed-signal filter can be implemented in an affordable and space-saving manner, and does not suffer the shortcomings of adaptive filtering algorithms, such as problems associated with time-delays, convergence, and stability.
In the example of
The combined signal 16 is input to an analog signal path 18 that is configured to process the combined signal 16. In the example of
The processed analog signal 24 is provided to an analog-to-digital converter (ADC) 26 configured to convert the processed analog signal 24 to the digital domain. As an example, the ADC 26 can be implemented as a flash ADC, a pipelined ADC, a sigma-delta ADC, or any of a variety of different types of ADC. Thus, the ADC 26 generates a digital signal DIG_OUT that is a digital version of the processed analog signal 24. The digital signal DIG_OUT is output from the mixed-signal filter 10, such that it can be provided to any of a variety of additional processing devices, such as down-sampling, synchronization, and/or additional filtering devices. The digital signal DIG_OUT is also provided to a digital feedback filter 28. The digital feedback filter 28 can be configured as a programmable digital filter that is operative to provide further filtering of the digital signal DIG_OUT, in addition to that already performed by the analog filter 20.
The mixed-signal filter 10 also includes a filter controller 30 configured to provide programming commands to the digital feedback filter 28. The filter controller can be implemented as including a register or other storage that can store filter coefficients. Additionally or alternatively, the filter controller can include digital logic, a digital signal processor or an application specific integrated circuit configured to program the digital feedback filter.
As an example, the digital feedback filter 28 can be programmed by the filter controller 30 to set poles, zeros, and/or gain associated with the frequency spectrum of the digital signal DIG_OUT. As a result, interferer signals at a frequency range that is close to the frequency range of the signal of interest in the digital signal DIG_OUT can be removed based on the programming commands provided by the filter controller 30. The filter controller 30 can be configured to set the poles, zeros, and/or gain, such as based on preset commands from associated control circuitry.
The poles, zeros, and/or gain can be defined for a given application by a user or be defined based on a specific known communication format or standard. As an example, upon the wireless communication device that includes the mixed-signal filter 10 communicating in the Global System for Mobile Communications (GSM) standard, the filter controller 30 can be configured to set a zero at a specific frequency based on a known interferer in the GSM frequency band. As another example, the filter controller 30 can be configured to program the digital feedback filter 28 dynamically, such as in response to detecting an interferer signal during run-time operation of the wireless communication device in which the mixed-signal filter 10 is included. Accordingly, the digital feedback filter 28 can be programmed to have a predetermined transfer function corresponding to the frequency spectrum associated with the digital signal DIG_OUT.
The digital feedback filter 28 provides a digitally filtered output signal 32 to a digital-to-analog converter (DAC) 34, which converts the digitally filtered output signal 32 to the analog domain. The DAC 34 can be configured as any of a variety of types of DAC, such as a sigma-delta DAC. The DAC 34 outputs the analog feedback signal 14, which is thus the analog version of the digitally filtered output signal 32. Accordingly, the analog version of the digitally filtered output signal 32 is combined with the analog input signal AN_IN to implement mixed-signal filtering of the analog input signal AN_IN. Specifically, the digital signal DIG_OUT is output from the mixed-signal filter 10 having both analog and digital filtering performed upon it.
The transfer function demonstrated by the graph 50 includes a power spectral density (PSD) function that plots a magnitude of the PSD of the digital output signal DIG_OUT in decibels (dB) over frequency in 105 Hz. The graph 50 demonstrates a first curve 52, indicated in the example of
The graph 50 also demonstrates a second curve 54, indicated in the example of
The transfer function of the second curve 54 can also be express in a zero/pole/gain (zpk) format based on Equation 1 in Equation 2 below:
Where i=√{square root over (−1)}.
As demonstrated in the example of
Proper choice of the digital feedback filter 28 can achieve desired signal processing (e.g., enhanced filtering at a certain frequency region) while ensuring stability of the system. For example, to achieve enhanced filtering at a certain frequency region, the digital feedback filter 28 can be configured to provide a gain at that frequency region, such that the overall filter response can include a notch at that frequency region. In addition, proper placement of zeros can add phase compensation, thereby improving loop stability. Furthermore, as described above, the output of the digital feedback filter 28 is converted into an analog signal by the DAC 34. In the example of
As an example, the filter controller 30 can program the digital feedback filter 28 to have a zero at approximately 400 kHz. For example, there may be a known and/or detected interferer signal in the analog input signal AN_IN at approximately 400 kHz. Therefore, upon combination of the feedback signal 14 with the analog input signal AN_IN at the signal combiner 12, the zero programmed at approximately 400 kHz combines with the transfer function demonstrated by the first curve 52 to generate an asymptotic notch 56 at approximately 400 kHz. As demonstrated in the example of
Due to the digital feedback filter 28 being fed back to the input of the mixed-signal filter 10 to implement a closed-loop filtering algorithm, a transfer function of the mixed-signal filter 10, such as the transfer function demonstrated by the second curve 54 in the example of
It is to be understood that the mixed-signal filter 10 is not intended to be limited to the example of
In the example of
The analog input signal AN_IN is provided as the input signal to a first one of the signal combiners 102, demonstrated as SIGNAL COMBINER 1 in the example of
The processed analog signal 112 output from the ANALOG SIGNAL PATH N is provided to an ADC 114 configured to convert the processed analog signal 112 to the digital domain. Thus, the ADC 114 generates a digital signal DIG_OUT that is a digital version of the processed analog signal 112. The digital signal DIG_OUT is output from the mixed-signal filter 100, and is also provided to a plurality N of digital feedback filters 116. Each of the digital feedback filters 116 can be configured as a programmable digital filter that is operative to filter the digital signal DIG_OUT. The digitally filtered output from each of the digital feedback filters 116 is provided to a corresponding DAC 118, which converts each of the digitally filtered output signals to the analog domain to provide the corresponding feedback signals 110.
The mixed-signal filter 100 also includes a filter controller 120 configured to provide separate programming commands to each of the digital feedback filters 116 via separate control lines 122. As an example, each of the digital feedback filters 116 can be individually programmed by the filter controller 120 to set poles, zeros, and/or gain associated with the frequency spectrum of the digital signal DIG_OUT. The filter controller 120 can be configured to set the poles, zeros, and/or gain based on preset commands, such as set by a user or based on a specific known communication format or standard. As a result, each of the digital feedback filters 116 can separately contribute to an overall transfer function of the mixed-signal filter 100. In addition, because of the cascade configuration of the mixed-signal filter 100 (e.g., with regard to the cumulative combination of the feedback signals 110 with the processed combined signals output from the analog signal paths 104), the filter controller 120 can generate any of a variety of transfer functions that can be significantly more complex than that generated by the mixed-signal filter 10 in the example of
It is to be understood that the mixed-signal filter 100 is not intended to be limited to the example of
The mixed-signal filter 150 receives a complex analog input signal, demonstrated in the example of
The signal combiners 152 and 154 each provide combined signal outputs to an I-channel analog signal path 160 and a Q-channel analog signal path 162, respectively. The I-channel analog signal path 160 and Q-channel analog signal path 162 are each configured to process (e.g., filter and/or amplify) the combined signals. The processed outputs of the I-channel analog signal path 160 and the Q-channel analog signal paths 162 are each provided to respective ADCs 164 and 166 that are each configured to convert the processed analog signals to the digital domain. Thus, the ADC 164 generates a digital signal I_OUT and the ADC 166 generates a digital signal Q_OUT. The digital signals I_OUT and Q_OUT are output from the mixed-signal filter 150, such that they can each be provided to any of a variety of additional processing devices. The digital signals I_OUT and Q_OUT are also each provided to a complex digital feedback filter 168. The complex digital feedback filter 168 can be configured as a programmable digital filter that is operative to filter both the in-phase digital signal I_OUT and the quadrature digital signal Q_OUT. Similar to as described above in the examples of
As a result of the filtering of the digital output signals I_OUT and Q_OUT, the complex digital feedback filter can provide a transfer function that is asymmetric about zero frequency, thus providing a complex transfer function of the mixed-signal filter 150.
The transfer function demonstrated by the graph 200 includes a PSD function that plots a magnitude of the PSD of the complex digital output signal components I_OUT and Q_OUT in dB over frequency in 106 Hz. The graph 200 demonstrates a first curve 202, indicated in the example of
The graph 200 also demonstrates a second curve 204, indicated in the example of
As a further example, the transfer function of the second curve 204 can also be expressed in a zero/pole/gain (zpk) format based on Equation 3 as shown below in Equation 4:
Where i=√{square root over (−1)}.
As demonstrated in the example of
As an example, the complex digital feedback filter 168 can be programmed to set zeros at approximately 530 kHz and at approximately −270 kHz, such as in response to known and/or detected interferer signals in the complex analog input signal components I_IN and Q_IN. The zeros at approximately 530 kHz and −270 kHz are thus asymmetric about zero frequency, and thus suitable to generate a complex transfer function. Therefore, upon combination of the feedback signals 156 and 158 with the analog input signal components I_IN and Q_IN, the zeros programmed at approximately 530 kHz and −270 kHz combine with the transfer function demonstrated by the first curve 202 to generate an asymptotic notch 206 at approximately 530 kHz and an asymptotic notch 208 at approximately −270 kHz. Accordingly, the mixed-signal filter 150 can be set with the predetermined complex transfer function demonstrated by the second curve 204 for the digital output signal DIG_OUT to remove the interferer signals at or near approximately 530 kHz and/or −270 kHz.
It is to be understood that the mixed-signal filter 150 is not intended to be limited to the example of
A mixed-signal filter that is configured to provide efficient signal filtering based on a combination of analog and digital algorithms, such as the mixed-signal filters 10, 100, and 150 in the examples of
In the example of
It is to be understood that the mixed-signal filter 262 is not limited to use for analog input signals received at the antenna 252. For example, the mixed-signal filter 262 can be implemented in the transceiver 254 to filter outgoing signals that are to be transmitted from the antenna 252, as well. Furthermore, the mixed-signal filter 262 is not limited to use within the transceiver 254, but can be implemented for any analog filtering application within the wireless communication device 250, or any other of a variety of electronic devices that incorporate analog filtering.
In view of the foregoing structural and functional features described above, certain methods will be better appreciated with reference to
At 308, the combined signal is filtered in the digital domain based on a predetermined digital filter transfer function. The predetermined digital filter transfer function can be programmable to set zeros, poles, and/or gain of the combined signal in the digital domain. The programming of the predetermined digital filter transfer function can occur during run-time operation of the wireless communication device in which the method 300 is implemented. At 310, the digitally filtered combined signal is converted to the analog domain to generate the feedback signal.
What have been described above are examples of the invention. It is, of course, not possible to describe every conceivable combination of components or methodologies for purposes of describing the invention, but one of ordinary skill in the art will recognize that many further combinations and permutations of the invention are possible. Accordingly, the invention is intended to embrace all such alterations, modifications, and variations that fall within the scope of this application, including the appended claims.
This application claims the benefit of U.S. Provisional Patent Application No. 60/868,580, which was filed on Dec. 5, 2006, and entitled “A Method and Apparatus for Realizing Complex Filtering Using Real Analog Filters Together with Mixed-Signal Complex Feedback,” the entire contents of which are incorporated herein by reference.
Number | Name | Date | Kind |
---|---|---|---|
3067996 | Theodosiou | Dec 1962 | A |
3228019 | Visceglia | Jan 1966 | A |
5187482 | Tiemann et al. | Feb 1993 | A |
5729230 | Jensen et al. | Mar 1998 | A |
6018556 | Janesch et al. | Jan 2000 | A |
6121910 | Khoury et al. | Sep 2000 | A |
6282185 | Hakkinen et al. | Aug 2001 | B1 |
6418220 | Fischer et al. | Jul 2002 | B2 |
6674381 | Gomez et al. | Jan 2004 | B1 |
6678709 | Gandhi et al. | Jan 2004 | B1 |
6822592 | Gandolfi et al. | Nov 2004 | B2 |
6864818 | Hezar | Mar 2005 | B1 |
6956517 | Baker et al. | Oct 2005 | B1 |
7034727 | Xu | Apr 2006 | B2 |
7042377 | Oliaei | May 2006 | B2 |
7058464 | Mallinson | Jun 2006 | B2 |
7061989 | Bellaouar et al. | Jun 2006 | B2 |
7091894 | Fudge et al. | Aug 2006 | B2 |
7155022 | Danielsen et al. | Dec 2006 | B2 |
7183957 | Melanson | Feb 2007 | B1 |
7218904 | Staszewski et al. | May 2007 | B2 |
7285538 | Choulika et al. | Oct 2007 | B2 |
7365668 | Mitteregger | Apr 2008 | B2 |
7420494 | Schreier | Sep 2008 | B1 |
7545303 | Silva et al. | Jun 2009 | B1 |
20050116850 | Hezar et al. | Jun 2005 | A1 |
20050275576 | Fudge et al. | Dec 2005 | A1 |
20060238395 | Nagai et al. | Oct 2006 | A1 |
Number | Date | Country | |
---|---|---|---|
20080129569 A1 | Jun 2008 | US |
Number | Date | Country | |
---|---|---|---|
60868580 | Dec 2006 | US |