FIELD

[0001] This application relates to electronic circuits, and more particularly, to N-path filter circuits.

BACKGROUND

[0002] Radio receivers typically include one or more filters for performing front-end frequency band selection to suppress out-of-band interferers. Traditionally such band-pass filtering has been performed by one or more off-chip SAW (surface acoustic wave) filters and/or one or more bulk acoustic wave (BAW) filters. However, in addition to being expensive to implement, SAW filters and BAW filters are not tunable. Accordingly, in response to the desire to achieve cheaper tunable filtering, there has been a recent trend to replace the off-chip SAW filters with on-chip integrated N-path filters (which may also be referred to as channel-selection filters).

[0003] An N-path filter includes N identical parallel signal paths wherein N is an integer greater than or equal to two. Each path comprises an input modulator which down-converts the input signal to baseband, a low pass filter circuit which filters the baseband signal to generate a filtered baseband signal, and an output modulator which up-converts the filtered baseband signal to the original band of the input signal. At any given time the low pass filter circuitry is connected between the input and output through a single path. The low pass filtering performed at baseband translates to band-pass filtering once up- converted. The centre frequency of the filter is determined by the mixing frequency. N-path filters have proven to provide band-pass filters having high Q-factors and a wide centre-frequency tuning range.

[0004] While N-path filters provide a number of advantages over SAW filters and BAW filters, conventional N-path filters have a number of drawbacks including limited out-of-band rejection.

[0005] The embodiments described below are provided by way of example only and are not limiting of implementations which solve any or all of the disadvantages of known N-path filters.

SUMMARY

[0006] This summary is provided to introduce a selection of concepts that are further described below in the detailed description. This summary is not intended to identify key features or essential features of the claimed subject matter, nor is it intended to be used to limit the scope of the claimed subject matter. [0007] Described herein are N-path filters wherein each path comprises: a low pass filter circuit; a first switching circuit which, when activated, connects the input port of the filter to the low pass filter circuit; and a second switching circuit which, when activated, connects the output port of the filter to the low pass filter circuit. The first switching circuits are activated in sequence and the second switching circuits are activated in the same sequence.

[0008] A first aspect provides an N-path filter comprising: an input port for receiving an input signal; an output port for outputting a filtered version of the input signal; and a plurality of paths, each path comprising: a low pass filter circuit; a first switching circuit which, when activated, connects the input port to the low pass filter circuit; and a second switching circuit which, when activated, connects the output port to the low pass filter circuit; wherein the first switching circuits of the plurality of paths are activated in sequence and the second switching circuits of the plurality of paths are activated in the same sequence.

[0009] By having two switching circuits in each path improved out-of-band rejection can be achieved without significantly increasing the power consumption, or the area, of the filter compared to conventional N-path filters. The N-path filter of the first aspect effectively increases the out-of-band rejection when used in the feedback path of an amplifier, such as a low noise amplifier (LNA).

[0010] The first switching circuit of at least one path may comprise at least one switch.

[0011] The second switching circuit of the at least one path may comprise at least one switch.

[0012] The at least one switch of the second switching circuit may be smaller than the at least one switch of the first switching circuit.

[0013] The at least one switch of the second switching circuit may be an order of magnitude smaller than the at least one switch of the first switching circuit.

[0014] Reducing the size of the switches of the second switching circuits reduces the area to implement the filter and/or ease the load on a downstream component, such as an amplifier.

[0015] The first switching circuit and the second switching circuit of the same path may be activated at the same time.

[0016] This simplifies the circuitry that generates the control signals as the same control signal can be used to activate the first and second switching circuits of the same path.

[0017] The first switching circuit and the second switching circuit of the same path may be activated at different times. [0018] This allows the control signals used to active the first switching circuits to be reused. The control signal used to activate the first switching circuit of the first path can also be be used to activate the second switching circuit of the fourth path, the control signal used to activate the first switching circuit of the second path to also be used to activate the second switching circuit of the first path and so on.

[0019] The low pass filter circuit of at least one path may comprise a capacitor.

[0020] Periodic activation of a first switching circuit of a path may down-convert the input signal into a baseband signal which may be converted to a filtered baseband signal by the low pass filter circuit of that path.

[0021] Periodic activation of a second switching circuit of a path may upconvert the filtered baseband signal to a signal in a same band as the input signal.

[0022] The input signal may be a radio frequency signal.

[0023] A second aspect provides a filter circuit comprising: an amplifier; and the N-path filter of the first aspect in a feedback path of the amplifier.

[0024] In such a filter circuit the filter rejection is not-limited by the on-resistance of the switching circuits nor is it limited by the bandwidth of the amplifier, thereby providing for and improved out-of- band rejection. Also, due to the Miller effect the size of the low pass filter components can be smaller, thereby allowing a reduction of the overall dimension of the circuit.

[0025] The output port of the N-path filter may be coupled to an input port of the amplifier and the low pass filter circuits may be coupled to an output port of the amplifier.

[0026] The filter circuit may further comprise a matching resistor in a second feedback path of the amplifier.

[0027] The amplifier may be a low-noise amplifier.

[0028] A third aspect provides a method of filtering an input signal comprising: sequentially connecting the input signal to a plurality of low pass filters circuits via first switching circuits associated with the low pass filter circuits; and generating a filtered output signal by sequentially outputting signals generated by the plurality of low pass filter circuits via second switching circuits associated with the low pass filter circuits.

[0029] The above features may be combined as appropriate, as would be apparent to a skilled person, and may be combined with any of the aspects of the examples described herein. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Examples will now be described in detail with reference to the accompanying drawings in which:

[0031] FIG. 1 is a circuit diagram of a N-path filter;

[0032] FIG. 2 is a schematic diagram illustrating example control signals for the N-path filter of FIG. 1;

[0033] FIG. 3 is a circuit diagram of a filter circuit comprising an amplifier and the N-path filter of FIG. 1 in the feedback path of the amplifier;

[0034] FIG. 4 is a circuit diagram of a filter circuit comprising two parallel N-path filters of FIG. 1;

[0035] FIG. 5 is a circuit diagram of an example improved N-path filter;

[0036] FIG. 6 is a circuit diagram of an example filter circuit comprising an amplifier and the N-path filter of FIG. 5 in the feedback path of the amplifier;

[0037] FIG. 7 is a graph illustrating input and output transfer functions of an amplifier and the filter circuit of FIG. 3;

[0038] FIG. 8 is a graph illustrating input and output transfer functions of the filter circuits of FIGS. 3 and 6; and

[0039] FIG. 9 is a flow diagram of an example method for filtering an input signal.

[0040] The accompanying drawings illustrate various examples. The skilled person will appreciate that the illustrated element boundaries (e.g., boxes, groups of boxes, or other shapes) in the drawings represent one example of the boundaries. It may be that in some examples, one element may be designed as multiple elements or that multiple elements may be designed as one element. Common reference numerals are used throughout the figures, where appropriate, to indicate similar features.

DETAILED DESCRIPTION

[0041] The following description is presented by way of example to enable a person skilled in the art to make and use the invention. The present invention is not limited to the embodiments described herein and various modifications to the disclosed embodiments will be apparent to those skilled in the art. Embodiments are described by way of example only. [0042] Described herein are N-path filters with improved out-of-band rejection compared to conventional N-path filters. Conventional N-path filters typically comprise a single switching circuit in each path which connects the input port and the output port to the corresponding low pass filter circuit (i.e. the single switching circuit acts as input modulator and output modulator). Improved out-of-band rejection can be achieved by having two switching circuits in each path. Specifically, each path has a first switching circuit which connects the input port of the filter to the corresponding low pass filter circuit, and a second switching circuit which connects the output port of the filter to the corresponding low pass filter circuit. N-path filters with two switching circuits in each path have proven to significantly improve the out-of-band rejection without significantly increasing the power consumption, or the area, of the filter compared to conventional N-path filters. N-path filters with two switching circuits per path have proven to be particularly effective at increasing the out-of-band rejection when used in the feedback path of an amplifier, such as a low noise amplifier (LNA).

[0043] To more clearly explain the N-path filter with improved out of band rejection, reference is first made to FIG. 1 which illustrates an example of an N-path filter 100. The N-path filter 100 comprises an input port 102 for receiving an input signal (V_IN) and an output port 104 for outputting a filtered version of the input signal (V_OUT). In some cases, the input signal (V_IN) and the output signal (V_OUT) are radio frequency signals (RF) signals. However, in other cases the input signal (V_IN) and the output signal (V_OUT) may be in a different band.

[0044] The N-path filter also comprises N identical signal paths 106-1 to 106-N wherein N is an integer greater than or equal to 2. Common values for N are 4 and 8, but it will be evident to a person of skill in the art that these are examples only and other values of N may be used. Each path 106-i comprises a low pass filter circuit (R+Ci) and a switching circuit (Si) in series. In the example shown in FIG. 1 each low pass filter circuit is formed by a capacitor Ci and a resistor R (which represents the resistance between the input signal source and the input port 102). However, it will be evident to a person of skill in the art that this is only an example low pass filter circuit and that other low pass filter circuits may be used. In the example shown in FIG. 1 each switching circuit comprises a single switch Si, however, in other examples one or more of the switching circuits may comprise more than one switch.

[0045] The switching circuit (Si) of a path i is situated between the input port and the corresponding low pass filter circuit such that when the switching circuit (Si) is activated (i.e. closed) the input port 102 is connected to the corresponding low pass filter circuit (R+Ci). Since the input port 102 and the output port 104 are short-circuited in the N-path filter 100 of FIG. 1 activating a switching circuit (Si) also connects the output port 104 to the corresponding low pass filter circuit (R+Ci). In this configuration periodic activation of a switching circuit (Si) of a path i down-converts the input signal to baseband. The low pass filter circuit (R+Ci) then filters the baseband signal to generate a filtered baseband signal, and the same switching circuit (Si) up-converts the filtered baseband signal to the original band of the input signal which is provided to the output port 104. The components of the low pass filter circuits (e.g. Cl, C2 ... CN) are selected to provide a desired channel filtering according to the input signal bandwidth.

[0046] Typically, the switching circuits (S I, S2 ... SN) are activated in a sequence such that only one switching circuit (Si) is active at a time and each switching circuit (SI, S2, ... SN) is active for the same amount of time. For example, the switching circuits may be activated in the sequence SI, S2, ... SN. In some cases, each switching circuit (Si) is activated by a corresponding control signal (Pi). Specifically, control signal PI controls the activation of the first switching circuit (SI), control signal P2 controls the activation of the second switching circuit (S2) and so on. In some cases, the control signals are based on a local oscillator (LO) signal. FIG. 2 illustrates an example set of control signals (PI, P2, ... PN) for the N-path filter 100 of FIG. 1 wherein each control signal represents a phase-shifted version of the LO signal. Specifically, the I ^th control signal Pi represents a ((i-l)*360/N) degree phase shifted version of the LO signal. For example, where N=4 there will be four control signals (PI, P2, P3 and P4) wherein PI represents a 0 degree phase shifted version of the LO signal, P2 represents a 90 degree phase-shifted version of the LO signal, P3 represents a 180 degree phase shifted version of the LO signal and P4 represents a 270 degree phase shifted version of the LO signal. In this example, each control signal has a duty cycle equal to 1/N of the LO period. This results in each switching circuit (Si) being activated for 1/Nth of the LO period (T_LO). The LO signal may be set to the centre frequency of the input signal which makes the N-path filter 100 of FIG. 1 well suited for direct conversion receivers.

[0047] The low pass filtering performed at baseband by the plurality of paths 106-1 to 106-N translates to a band-pass filtering effect once up-converted.

[0048] Each switching circuit (Si) has an“on-resistance” (r) when it is activated. Since the on- resistance (r) of each switching circuit (Si) is in the signal path and has current flowing through it, the on-resistance (r) of each switching circuit (Si) limits the amount of out-of-band rejection of the N-path filter 100. Specifically, due to a potential divider effect, the out-of-band gain is approximately r/(R+r).

[0049] One technique to address this issue may be to incorporate the N-path filter 100 into the feedback path of a gain component, such as an LNA (low-noise amplifier). Reference is now made to FIG. 3 which illustrates an example filter circuit 300 that comprises an amplifier 302 with the N-path filter 100 of FIG. 1 in the feedback path of the amplifier 302. Specifically, in the example of FIG. 3 the output port 104 of the N-path filter 100 is coupled to the input of the amplifier 302 and the other side of the low pass filter circuits (Cl, C2, .... CN) (i.e. the side of the low pass filter circuits not coupled to the corresponding switching circuit (Si)) are coupled to the output of the amplifier 304. Due to the Miller effect this allows the size of the low pass filter circuit capacitors (Cl, C2, ... CN) to be reduced by the gain factor.

[0050] Furthermore, in this configuration the out-of-band rejection is improved, relative to the N-path filter 100 of FIG. 1, by the gain factor of the amplifier 302. This is because the input signal is amplified at the desired frequency. Such a configuration is most effective at improving the out-of-band rejection when the amplifier has a substantial gain. The maximum gain of an LNA is typically around 20 dB. While higher gains can be achieved, this becomes progressively more difficult to achieve at higher frequencies and generally comes at the expense of higher power consumption. Furthermore, the on- resistance (r) of the switching circuits remains the limiting factor in the amount of out-of-band rejection.

[0051] In some cases, the filter circuit 300 may also include an impedance matching circuit. For example, if the input port 102 is coupled to a circuit/component (e.g. RF antenna) that presents an impedance to the filter circuit 300, for impedance matching the filter circuit 302 may comprise an impedance matching circuit that presents a corresponding impedance to the component (e.g. RF antenna). In some cases, as shown in FIG. 3, the impedance matching circuit may be implemented as a resistor (RF) in a feedback path of the amplifier 302. However, in other cases, no impedance matching may be required or the impedance matching may be performed or implemented in another manner.

[0052] Another technique to address this issue has been to use two N-path filters with different centre frequencies. As described above, the centre frequency of an N-path filter is set by the frequency of the LO signal used to control the switching circuits of the N-path filter. Accordingly N-path filters with different centre frequencies will use different LO signals to control the switching circuits. Reference is now made to FIG. 3 which illustrates an example filter circuit 400 that comprises a first N-path filter 402 (such as the N-path filter 100 of FIG. 1) with a first centre frequency and a second N-path filter 404 (such as the N-path filter 100 of FIG. 1) with a second, different, centre frequency. The output of the filter circuit 400 (V_OUT) is then calculated as the difference between the outputs of the two N-path filters 402, 404.

[0053] Where the second frequency is only slightly different from the first frequency then the phases (f and y _{2 )} °f the outputs V_OUTl and V_OUT2 of the two N-path filters is approximately f, =—f ₂ in the pass-band of the filter, and therefore due to the subtraction, they will add up thus increasing the gain in the pass-band. In contrast, for frequencies far out of the pass-band region of the filter, the output of the two N-path filters are almost in-phase f , = f ₂ and will cancel each other in the subtraction thus increasing the out-of-band rejection. This technique is described in more detail in Milad Darvishi; Ronan van der Zee; Eric A. M. Klumperink; Brain Nauta,“Widely Tunable 4 ^th Order Switched G _m - C Band-pass Filter Based on N-path Filters”, IEEE journal of solid-state circuits, 47(12), 3105- 3119. However, this technique requires more complicated circuitry to generate all the control signals for the switching circuits which increases the power consumption of such a filter circuit compared to a conventional N-path filter. Furthermore, since there are two N-path filters the area to implement such a filter circuit is significantly larger than the area to implement a conventional N-path filter. Such a filter also requires a complicated calibration mechanism to optimize its performance.

[0054] Accordingly, described herein are N-path filters which have improved out-of-band rejection relative to conventional N-path filters yet have similar power consumption and area requirements to conventional N-path filters. The improved out-of-band rejection is achieved by adding an extra switching circuit to each path of the N-path filter which connects the corresponding low pass filter circuit to the output port of the filter. Periodic activation of the extra switching circuit of a path up-converts the filtered baseband signal generated by the corresponding low pass filter circuit to the original band of the input signal and provides the upconverted signal to the output port. In this configuration there will be substantially no current flowing through the extra switching circuits when activated (i.e. closed) thus the on-resistance of the extra switching circuits do not limit the level of the out-of-band rejection.

[0055] Reference is now made to FIG. 5 which illustrates an example N-path filter 500 in accordance with an embodiment. Like the N-path filter 100 of FIG. 1, the N-path filter 500 comprise an input port 502 for receiving an input signal (V_IN), an output port 504 for outputting a filtered version of the input signal (V_OUT), and N identical 506-1 to 506-N paths wherein N is an integer greater than or equal to 2. Each path 506-i comprises a low pass filter circuit 508-i, a first switching circuit 510-i and a second switching circuit 512-i. In the example shown in FIG. 5 each low pass filter circuit 508-i is formed by a capacitor Ci and a resistor R (which represents the resistance between the input signal source and the input port 502). However, it will be evident to a person of skill in the art that this is only an example low pass filter circuit and that other low pass filter circuits may be used. In the example shown in FIG. 5 each of the first and second switching circuits 510-i and 512-i comprises a single switch Si or SEi, however, in other examples one or more of the switching circuits 510-1 to 510-N, 512-1 to 512-N may comprise more than one switch. The components of the low pass filter circuits (e.g. Cl, C2 ... CN) are selected to provide a desired channel filtering according to the input signal bandwidth.

[0056] Each first switching circuit 510-1 to 510-N is situated between the input port 502 and the corresponding low pass filter circuit 508-1 to 508-N such that when the first switching 510-i circuit of a path i is activated (i.e. closed) the input port 502 is connected to the corresponding low pass filter circuit 508-i. In this configuration, periodic activation of the first switching circuit 510-i of a path i down-converts the input signal to baseband and provides the baseband signal to the low pass filter circuit 508-i. The low pass filter circuit 508-i then generates a filtered baseband signal from the received baseband signal.

[0057] Each second switching circuit 512-1 to 512-N is situated between the output port 504 and the corresponding low pass filter circuit 508-1 to 508-N such that when a second switching circuit 512-i of a path i is activated (i.e. closed) the output port 504 is connected to the corresponding low pass filter circuit 508-i. In the example shown in FIG. 5 one side of each second switching circuit 512-1 to 512- N is connected to the output port 504 and the other side of the switching circuit 512-1 to 512-N is connected to the wire or trace between the corresponding first switching circuit and the corresponding low pass filter circuit. When the second switching circuits 512-1 to 512-N are situated between the output port and the corresponding low pass filter circuit, periodic activation of a second switching circuit 512-i of a path i up-converts the filtered baseband signal generated by the corresponding low pass filter circuit and provides the up-converted signal to the output port 504. In this arrangement, when the second switching circuit 512-i of a path i is activated (i.e. closed) there will be substantially no current flowing through the second switching circuit 512-i thus the on-resistance of the second switching circuit 512-i does not limit the level of the out-of-band rejection.

[0058] Furthermore, since there will be substantially no current flowing through the second switching circuits 512-1 to 512-N the switches of the second switching circuits 512-1 to 512-N can be smaller than the switches of the first switching circuits 510-1 to 510-N. In some cases, the switches of the second switching circuits may be an order of magnitude smaller than the switches of the first switching circuits. The size of a switch may be the physical size of the switch which may be defined by the length (F) and width (W) of the switch. For example, the size of a switch may be defined by the area of the switch which is equal to the product of the length and the width (F x W). In some cases, the minimum size of the switches of the second switching circuits may be limited by the maximum acceptable noise. For example, the switches of the second switching circuits may be reduced to any size so long as their contribution to the overall noise is acceptable. Generally, the lower the frequency, the smaller the switches can be. Reducing the size of the switches of the second switching circuits may reduce the area to implement the filter and/or ease the load on a downstream component, such as an amplifier, connected to the output port 502 of the N-path filter 500.

[0059] In addition, isolating the input port 502 from the output port 504 achieves extra filtering at the output port 504 which protects any component, such as an amplifier, connected to the output port 504 from being exposed to the power of any unwanted interferers.

[0060] The first switching circuits 510-1 to 510-N are activated in a sequence such that only one switching circuit is active at a time and each first switching circuit 510- 1 to 510-N is active for the same amount of time. For example, the first switching circuits may be activated in the sequence 510-1, 510- 2 ... 510-N. In some cases, each of the first switching circuits 510-1 to 510-N is activated by a corresponding control signal that is a phase-shifted version of a local oscillator signal (LO) wherein each control signal has a duty cycle equal to 1/N of the LO period (T_LO). For example, the first switching circuits 510-1 to 510-N may be controlled by the example control signals P0 to PN respectively illustrated in FIG. 2. As described above, in the example shown in FIG. 2 the i ^th control signal Pi represents a ((i-l)*360/N) degree phase shifted version of the LO signal. For example, where N=4 there will be four control signals (PI, P2, P3 and P4) wherein PI represents a 0 degree phase shifted version of the LO signal, P2 represents a 90 degree phase-shifted version of the LO signal, P3 represents a 180 degree phase shifted version of the LO signal and P4 represents a 270 degree phase shifted version of the LO signal. It will be evident to a person of skill in the art that this is an example only and the control signals for the first switching circuits 510- 1 to 510-N may represent different phase shifts of the LO signal, however, they are typically 360/N degrees apart.

[0061] The second switching circuits 512-1 to 512-N are activated in the same sequence as the corresponding first switching circuits 510-1 to 510-N. For example, if the first switching circuits 510- 1 to 510-N are activated in the sequence 510-1, 510-2, ... , 510-N, then the second switching circuits 512-1 to 512-N are activated in the sequence 512-1, 512-2, ...., 512-N.

[0062] In some cases, the first and second switching circuits 510-i and 512-i of the same path i are activated at the same time. Specifically, in these cases the first and second switching circuits 510-1 and 512-1 of the first path are activated at the same time, the first and second switching circuits 510-2 and 512-2 of the second path are activated at the same time and so on. This may be advantageous where a circuit coupled to the output port 504 of the N-path filter 500 provides impedance matching to the input port 502. It can also simplify the circuitry that generates the control signals as the same control signal can be used to activate the first and second switching circuits 510-i and 512-i of the same path i. For example as shown in Table 1, control signal PI of FIG. 2 could be used to active both the first and second switching circuits 510-1 and 512-1 of the first path, the control signal P2 could be used to activate both the first and second switching circuits 510-2 and 512-2 of the second path etc. Table 1

[0063] However, in other cases, the first and second switching circuits 510-i and 512-i of the same path i are not activated at the same time. In particular, the second switching circuit may be activated by a control signal that is phase shifted by a predetermined amount with respect to the control signal used to activate the corresponding first switching circuit. For example, the second switching circuit of a path may be activated by a control signal that is phase shifted by 90 degrees with respect to the control signal used to active the corresponding first switching circuit. Where N is equal to 4 this may result in control signals for the first and second switching circuits being phase-shifted versions of the LO as shown in Table 2.

Table 2

[0064] Where the first and second switching circuits 510-i and 512-i of the same path i are not activated at the same, it may be advantageous for the phases of their control signals to differ by a factor of 360/N. This allows the control signals used to active the first switching circuits to be reused. For example, in Table 2 N=4 so 360/4=90 and the controls signals for the first and second switching circuits of the same path have a phase difference of 90 degrees. This allows the control signal used to activate the first switching circuit of the first path to also be used to activate the second switching circuit of the fourth path, the control signal used to activate the first switching circuit of the second path to also be used to activate the second switching circuit of the first path and so on.

[0065] The N-path filter 500 of FIG. 5 provides a filter with substantially improved out-of-band rejection compared to an N-path filter, such as the N-path filter 100 of FIG. 1, without additional power consumption, very little impact on area, little effect on noise, no extra complicated circuitry and no need for fine tuning.

[0066] In some cases, a filter with further improved out-of-band rejection may be achieved by placing the N-path filter described herein in the feedback path of an amplifier. This combines the out-of-band rejection improvements achieved with the N-path filter 500 of FIG. 5 with the out-of-band rejection improvements achieved by having an N-path filter in the feedback path of an amplifier.

[0067] Reference is now made to FIG. 6 which illustrates a filter circuit 600 that comprises an amplifier 602 and the N-path filter 500 of FIG. 5 in the feedback path of the amplifier 602. Specifically, in the example shown in FIG. 6 the output port 504 of the N-path filter 500 is electrically coupled to the input port of the amplifier and the low pass filters 508-1 to 508-N are electrically coupled to the output of the amplifier 602. In some cases the amplifier 602 may be a low noise amplifier (LNA). Such a filter circuit 600 has improved out-of-band rejection over the filter circuit 300 of FIG. 3 because the filter rejection is not-limited by the on-resistance of the switching circuits nor is it limited by the bandwidth of the amplifier. Also, due to the Miller effect the size of the low pass filter components (e.g. Cl, C2, ... CN) can be smaller.

[0068] Furthermore, as described above, since there is no direct short connection between the input port and the input to the amplifier 602 extra filtering is achieved at the input of the amplifier 602 (i.e. there is a higher level of out-of-band rejection at the input of the amplifier 602) which protects the amplifier 602 from being exposed to the power of any large unwanted interferers. This means that the amplifier 602 will not be subjected to the full power of the unwanted interferers and thus may not require large signal handling capacity. This in turn, reduces the risk of signal compression within the amplifier 602.

[0069] The out-of-band linearity of the filter circuit 600 is also greatly improved relative to the filter circuit 300 of FIG. 3 at negligible cost to the noise figure and power consumption. See FIGS. 7 and 8 for example transfer functions illustrating this. [0070] As with the filter circuit 300 of FIG. 3, the filter circuit 600 of FIG. 6 may also comprise an impedance matching circuit. For example, if the input port 502 of the N-path filter 500 is coupled to a circuit/component (e.g. an RF antenna) that presents an impedance to the filter circuit 600, for impedance matching the filter circuit 600 may comprise an impedance matching circuit that presents a corresponding impedance to the circuit/component (e.g. antenna). In some cases, as shown in FIG. 6, the impedance matching circuit may be implemented as a resistor (RF) in a feedback path of the amplifier 602. However, in other cases, no impedance matching may be required or the impedance matching may be performed in another manner.

[0071] Where the filter circuit 600 comprises an impedance matching circuit (e.g. resistor RF) in a feedback path of the amplifier 602 the first and second switching circuits of the same path of the N-path filter 500 may be activated at the same time so that the impedance matching performed by the impedance matching circuit (e.g. resistor RF) performs impedance matching in the same manner as the same impedance matching circuit in the filter circuit 300 of FIG. 3. Specifically, when the first and second switching circuits of the same path are activated at the same time the component (e.g. antenna) coupled to the input port 502 will always see the impedance circuit in response to an input signal. However, instead of there being a direct short connection between the input port and the input of the amplifier as in the filter circuit 300 of FIG. 3 the input port is connected to the input of the amplifier 602 via two switching circuits.

[0072] Reference is now made to FIGS. 7 and 8 which illustrate the improved out-of-band rejection that can be achieved using the filter circuit 600 of FIG. 6 that comprises the N-path filter 500 of FIG. 5 in the feedback loop of an LNA relative to the out-of-band rejection achieved using the filter circuit 300 of FIG. 3 that comprises the N-path filter 100 of FIG. 1 in the feedback loop of an LNA. In these examples, the centre frequency of the N-path filters is 2 GHZ (i.e. the switching circuits of the N-path filters are controlled by a LO signal with a frequency of 2 GHz).

[0073] FIG. 7 illustrates the input and output transfer functions 702 and 704 of a LNA without an N- path filter in the feedback loop, and the input and output transfer functions 706 and 708 of an LNA with an N-path filter in the feedback loop (i.e. the circuit 300 of FIG. 3). Each input transfer function 702, 706 illustrates the magnitude (in dB) of the signal input to the LNA relative to the magnitude of the original input signal in the frequency domain. Similarly, each output transfer function 704, 708 illustrates the magnitude (in dB) of the signal output from the LNA relative to the original input signal in the frequency domain.

[0074] It can be seen from the input and output transfer functions 702 and 704 that when there is no N-path filter in the feedback loop the input to the LNA generally matches the original input signal and the output of the LNA is an amplified version of the input signal. Although it can be seen that there is generally more gain applied to the frequencies below the centre frequency (2 GHz in this example) and less gain applied the frequencies above the centre frequency.

[0075] It can be seen from the input and output transfer functions 706 and 708 that when a N-path filter is placed in the feedback loop of the LNA (i.e. to generate the circuit 300 of FIG. 3) that both the input and outputs of the LNA are attenuated with respect to the original input signal at frequencies above and below the centre frequency. Specifically, while the inputs and outputs of the circuit 300 of FIG. 3 have substantially the same gain as the inputs and outputs of an LNA without an N-path filter in the feedback loop at the centre frequency - i.e. the input of a LNA without an N-path filter in the feedback loop has a 70.719 mdB gain at the centre frequency and the input to the circuit 300 of FIG. 3 has a 359.82 mdB gain at the centre frequency; and the output of an LNA without an N-path filter in the feedback loop has a 18.057 dB gain at the centre frequency and the output of the circuit 300 of FIG. 3 has a 13.809 dB gain at the centre frequency - both the inputs and outputs of the LNA are significantly attenuated for other frequencies. In other words, when the N-path filter is used in the feedback loop of the LNA there is significant out-of-band rejection in both the input and the output. However, it can be seen from FIG. 7 that the transfer functions 706 and 708 are not symmetric about the centre frequency. Specifically, for both the input and output of the LNA, the gain for frequencies above the centre frequency generally increases with frequency and the gain for frequencies above the centre frequency generally decreases with frequency. This is caused by the bandwidth of the LNA.

[0076] FIG. 8 illustrates the input and output transfer functions 706 and 708 of an LNA with a N-path filter in the feedback loop (i.e. the circuit 300 if FIG. 3) and the input and output transfer functions 802 and 804 with the improved N-path filter 500 of FIG. 5 in the feedback look (i.e. the circuit 600 of FIG. 6). In can be seen from FIG. 8, that for both the inputs and output of the LNA in the circuit 600 of FIG. 6 that the attenuation for the out-of-band frequencies (i.e. the out-of-band rejection) is significantly improved relative to the input and output of the LNA in the circuit 300 of FIG. 3. It also can be seen that, for both the input and output of the LNA in the circuit 600 of FIG. 6, the attenuation for frequencies on either side of the centre frequency does not increase or decrease in the same manner as when an N- path filter is used in the feedback path. Therefore the out-of-band linearity for the circuit 600 of FIG. 6 is improved relative to the circuit 300 of FIG. 3. Accordingly, when the improved N-path filter is used in the feedback loop of the LNA, the out-of-band rejection is not limited, in the same manner, by the LNA bandwidth on either side of the centre frequency.

[0077] The improved out-of-band rejection can ease the linearity of compression components downstream of the LNA/filter circuit. For example, when the filter circuit is used in a radio receiver the LNA/filter circuit may be followed by a mixer and/or baseband circuitry and the improved out-of-band rejection may ease the linearity and compression requirements on the mixer and baseband circuitry.

[0078] Reference is now made to FIG. 9 which illustrates an example method 900 for filtering an input signal. The method 900 begins at both blocks 902 and 904. At block 902, the input signal is sequentially connected to each of a plurality of low pass filter circuits (e.g. low pass filter circuits 508-1 to 508-N). In some cases, the input signal may be connected to a low pass filter circuit (e.g. low pass filter circuit 508-i) via a first switching circuit (e.g. first switching circuit 510-i) associated with that low pass filter circuit. At block 904, the low pass filter circuits (e.g. low pass filter circuits 508-1 to 508-N) are sequentially connected to an output port of the filter to generate a filtered output signal. In some cases, a low pass filter circuit may be connected to the output port via a second switching circuit associated with the low pass filter circuit. The low pass filter circuits are connected to the output port in the same sequence as the low pass filter circuits are connected to the input port. In some cases, the same low pass filter may be connected to the input port and the output port at the same time. In other cases, the same low pass filter may be connected to the input port and the output port at different times. For example, a low pass filter may be connected (via the corresponding first switching circuit) to the input port for a first period and the same low pass filter may be connected (via the corresponding second switching circuit) to the output port for a second, different, period. The first switching circuit may comprise one or more switches. The second switching circuit may comprise one or more switches. In some cases, the switches of the second switching circuits are smaller than the switches of the first switching circuits.

[0079] The applicant hereby discloses in isolation each individual feature described herein and any combination of two or more such features, to the extent that such features or combinations are capable of being carried out based on the present specification as a whole in the light of the common general knowledge of a person skilled in the art, irrespective of whether such features or combinations of features solve any problems disclosed herein, and without limitation to the scope of the claims. The applicant indicates that aspects of the present invention may consist of any such individual feature or combination of features. In view of the foregoing description it will be evident to a person skilled in the art that various modifications may be made within the scope of the invention.