Technical field

[0001 ] The present disclosure relates to a bandpass filter, and in particular to a bandpass filter comprising two double-resonance circuits.

Background

[0002] Bandpass filters are generally used in e.g. signal processing solutions, for filtering signals so that only signals having a frequency within a specific part of a spectrum may pass through the bandpass filter. The bandpass filters are commonly referred to as being perfect, or ideal, meaning that only signal frequencies of the specific spectrum range are passed through, and that signal content having a frequency other than within the specific spectrum range is blocked. However, in reality, this is not entirely true. The difference between the perfect bandpass filter and the reality is illustrated in figure 1a, in which the dashed box is the perfect bandpass filter and the curve is the bandpass filter in reality.

[0003] A bandpass filter may be realised or implemented in different ways. A first example is illustrated in figure 1 b and a so called T-equivalent is illustrated in figure 1c. The frequencies that are allowed to pass through the bandpass filter are referred to as the passband of the bandpass filter.

[0004] For a lower edge of a microwave frequency band, it is always a hard decision whether to use lumped element or distributed realisation. Lumped element filters may be of smaller dimension, but readability limits are the minimum inductance value and series resonant frequency of capacitors. For nearly critical coupling of the resonators, a very small coupling reactance is necessary. This could be realised by a small inductance or a large capacitance. For both, serious readability limits exist.

[0005] For microwaves below 3 GHz, it is common using a distributed filter realisation because parasitics of lumped elements above 500 MHz become significant. For that reason, usually it is suggested to use lumped element filters rather below 500 MHz. Realisability limits are that the minimum inductance value is about 1 nH, and the series resonance frequency of the capacitors are maximum a few GHz that limits the usable maximum capacitance value.

[0006] Entering into the forbidden region of realizability for lumped element filter design above 500MHz-1 GHz, usually results in severe degradation of filter performance and loss of tuneability. High filter steepness requires very high filter orders, therefore introduces higher losses and requires many components.

Summary

[0007] The object is to obviate at least some of the problems outlined above. In particular, it is an object to provide a bandpass filter having steep slopes at the passband edges. These objects and others may be obtained by providing a bandpass filter according to the independent claims attached below.

[0008] According to an aspect a bandpass filter is provided. The bandpass filter comprises a first double-resonance circuit having a first resonance frequency, FR1 , and a first passband P1 of bandwidth, B1 , where the first double-resonance circuit allows signals comprised in the passband P1 of bandwidth B1 , i.e. the range of frequencies around the resonant frequency FR1 at which they will resonate, to pass through, the first double-resonance circuit being connected in series to a coupling circuit, which in turn is connected in series to a second double-resonance circuit having a second resonance frequency, FR2, and a second passband P2 of bandwidth, B2, where the second double-resonance circuit allows signals comprised in the passband P2 of bandwidth B2, i.e. the range of frequencies around the resonant frequency FR2 at which they will resonate, to pass through, wherein the frequency spacing between FR1 and FR2 is such that P1 and P2 at least partly overlap when FR1 is not equal to FR2.

[0009] The bandpass filter may have several advantages. A first possible advantage is that parasitic inductance, L, of capacitors, C, may be treated together with an intentionally placed serial inductance, and thus L and C values are within the realisability limits. A second possible advantage is that due to the introduction of additional transmission zeroes introduced by serial LC circuits in shunt configuration, high slope steepness can be achieved. A third possible advantage is that the frequencies of the transmission zeroes are well tunable. A fourth possible advantage is that the lumped element bandpass filter is smaller in size and weight than a corresponding distributed one, in the range below 3 GHz. Due to the fast transition from pass to stop band, the filter design may allow for signal selectivity and in Sub-Carrier-Modulation, SCM, applications for dense carrier allocation.

Brief description of drawings

[00010] Embodiments will now be described in more detail in relation to the accompanying drawings, in which:

[0001 1 ] Figure 1 a is a schematic illustration of a transfer function of an ideal bandpass filter and a real bandpass filter.

[00012] Figure 1 b is a schematic illustration of an inductively coupled bandpass example.

[00013] Figure 1 c is a schematic illustration of a T-equivalent of the inductively coupled bandpass example in figure 1 b.

[00014] Figure 2a is a schematic illustration of a bandpass filter according to an exemplifying embodiment.

[00015] Figure 2b is a schematic illustration of the individual transfer functions of the two double-resonance circuits of the bandpass filter of figure 2a and a resulting transfer function of the bandpass filter.

[00016] Figure 2c is a schematic illustration of a bandpass filter according to another exemplifying embodiment.

[00017] Figure 3a is an illustration of transfer functions for different coupling.

[00018] Figure 3b is a schematic illustration of an inductively coupled bandpass example where the inductance L3 in figure 1 a is replaced by a general coupling impedance. [00019] Figure 3c is a schematic illustration of an inductively coupled bandpass example where the general coupling impedance of figure 3b is replaced by a serial LC circuit, i.e. a serial inductance and capacitance circuit.

[00020] Figure 3d is an illustration of the transfer function of the circuit in figure 3c.

[00021 ] Figure 4a is an illustration of an example of a bandpass filter according to an exemplifying embodiment having specific values for the different capacitances and inductances.

[00022] Figure 4b is an illustration of the transfer function of the circuit in figure 4a.

Detailed description

[00023] Briefly described, a bandpass filter comprising two double-resonance circuits is provided. The bandpass filter exploits imperfections of circuits and components, especially parasitic inductances of capacitors in order to achieve a bandpass filter having steep slopes at the edges of the passband of the bandpass filters. In other words, the bandpass filter exploits imperfections of real capacitors in order to achieve a bandpass filter being close to the ideal bandpass filter illustrated in figure 1 a. Exploiting these imperfections allow for operating the bandpass filter beyond the component realisability limits in higher microwave frequencies.

[00024] Embodiments of such a bandpass filter will now be described with reference to figure 2a. Figure 2a illustrates the bandpass filter 200 comprising a first double-resonance circuit 210 having a first resonance frequency, FR1 , and a first passband P1 of bandwidth, B1 , where the first double-resonance circuit 210 allows signals comprised in the passband P1 of bandwidth B1 , i.e. the range of frequencies around the resonant frequency FR1 at which they will resonate, to pass through, the first double-resonance circuit 210 being connected in series to a coupling circuit 230, which in turn is connected in series to a second double- resonance circuit 220 having a second resonance frequency, FR2, and a second passband P2 of bandwidth, B2, where the second double-resonance circuit allows signals comprised in the passband P2 of bandwidth B2, i.e. the range of frequencies around the resonant frequency FR2 at which they will resonate, to pass through, wherein the frequency spacing between FR1 and FR2 is such that P1 and P2 at least partly overlap when FR1 is not equal to FR2.

[00025] The first and the second double-resonance circuit 210 and 220 may have the same components as illustrated in figure 2a, however, the different values of the inductances and capacitances may be different for the two circuits.

[00026] The different values of the inductances and the capacitances determine the first resonant frequency FR1 and the second resonant frequency FR2. It shall be pointed out that the two resonant frequencies FR1 and FR2 may be the same. Each double-resonance circuit 210 and 220 has a respective passband, which as explained above with reference to figurela is not a perfect rectangle but rather a curve. Figure 2b illustrates that the first and the second double-resonance circuit 210 and 220 have a respective passband and that the passband for the first double-resonance circuit 210 has a steep slope to the left of the resonance frequency FR1 and a more tilted or inclined slope to the right of FR1. Likewise, figure 2b illustrates that the second double-resonance circuit 220 has a passband having a steep slope to the right of the resonance frequency FR2 and a more tilted or inclined slope to the left of FR2.

[00027] Since the first and the second resonant frequencies FR1 and FR2 may be the same or at least relatively close such that P1 and P2 at least partly overlap when FR1 is not equal to FR2, the resulting passband of the bandpass filter 200 has steep slopes at both the lower and the upper frequencies of the passband.

[00028] The bandpass filter may have several possible advantages. A first possible advantage is that parasitic inductance, L, of capacitors, C, may be treated together with an intentionally placed serial inductance, and thus L and C values are within the readability limits. A second possible advantage is that due to the introduction of additional transmission zeroes introduced by serial LC circuits in shunt configuration, high slope steepness can be achieved. A third possible advantage is that the frequencies of the transmission zeroes are well tunable. A fourth possible advantage is that the lumped element bandpass filter is smaller in size and weight than a corresponding distributed one, in the range below 3 GHz. Due to the fast transition from pass to stop band, the filter design may allow for signal selectivity and in Sub-Carrier-Modulation, SCM, applications for dense carrier allocation.

[00029] SCM is a modulation scheme where several sub-signals (carriers) are modulated as one composed signal on one optical carrier. If filters are very selective (steep slopes), it is possible to place sub-carriers closer, spectrally more efficient, since guard spectrum (to get space between signal spectra) may be made small.

[00030] According to an embodiment, the bandpass filter 200 according to claim 1 , wherein the first double-resonance circuit 210 has a notch, N1 , at a frequency lower than FR1 where the transfer through the first double-resonance circuit 210 is minimal, wherein the second double-resonance circuit 220 has a notch, N2, at a frequency higher than FR2 where the transfer through the first double-resonance circuit 210 is minimal, wherein the frequency of FR1 is lower than the frequency of FR2.

[00031 ] Assuming the capacitors, e.g. 214/224 (which will be described in more detail below), are ideal, then there is no notch. However, due to the fact that capacitors in real are not ideal, there is a notch. In practice, capacitors are nearly ideal but inductors have significant parasitic series loss resistors. Their effect is that the transfer function at notches N1 and N2 are not exactly zero but a small value.

[00032] The two notches N1 and N2 are a result of resonance between respective capacitor B 214, 224 and inductance 216, 226 of the first and the second double-resonance circuit 210 and 220 as will be described below in more detail. The steepness of the slope close to notches N1 and N2 of each double- resonator 210, 220 is very sharp compared to the steepness of the slope on the opposite side of the resonant frequency FR1 and FR2 respectively, and each notch N1 and N2 respectively, occur at a frequency where the transfer through the respective double-resonance circuit 210, 220 is minimal.

[00033] According to yet an embodiment, an input of the first double-resonance circuit 210 is coupled to an input 201 of the bandpass filter and an output of the second double-resonance circuit is coupled to an output 202 of the bandpass filter.

[00034] The bandpass filter is illustrated in figure 2a by a dotted box having an input, IN, 201 coupled to the input (not specifically shown) of the first double- resonance circuit 210 and the output, OUT, 202, coupled to the output (not specifically shown) of the second double-resonance circuit 220. This means that signals of different frequencies may be inputted to the bandpass filter 200 by means of the input, IN, 201 . The inputted signals may then pass through first the first double-resonance circuit 210 if the frequencies are within the passband of the first double-resonance circuit 210, and then through the coupling circuit 230. After the coupling circuit 230, the signals may pass through the second double- resonance circuit 220 if the frequencies of the signals are within the passband of the second double-resonance circuit 220 and outputted from the bandpass filter 200 by means of the output OUT 202. Consequently, only the frequency content of signals that is within the passband of both the first and the second double- resonance circuit 210 and 220 may pass through the bandpass filter 200. Also, frequency content of signals that is too far from FR1 and FR2 and thus outside the range of both resonance frequencies is highly attenuated and thus blocked from passing.

[00035] According to still an embodiment, illustrated in figure 2c, the bandpass filter 200 further comprises a first capacitor 203 arranged between the input of the first double-resonance circuit 210 and the input 201 of the bandpass filter 200 and a second capacitor 204 arranged between the output of the second double- resonance circuit 220 and the output 202 of the bandpass filter 200.

[00036] By these two capacitors, 203 and 204, e.g. 50 ohm impedance of terminations connected to the bandpass filter 200 are not a significant load to the resonators so that their quality factors may be lower. The two capacitors further enable the passband of the bandpass filter 200 may be larger than without the two capacitors, 203 and 204.

[00037] According to another embodiment, each double-resonance circuit 210, 220 comprises two inductances 21 1 , 212, 221 , 222 connected in series between the input of the double-resonance circuit and an output of the double-resonance circuit, and three capacitors A denoted 213 and 223 in figure 2a, B denoted 214 and 224 in figure 2a, and C denoted 215 and 225 in figure 2a, each having a first and a second connection point, wherein the three capacitors are arranged in parallel with regards to each other such that the first connection point of capacitor A 213, 223 is coupled to the input of the double-resonance circuit 210 ,220, the first connection point of capacitor B 214, 224 is coupled to a point between the two inductances 21 1 , 212, 221 , 222, and the first connection point of capacitor C 215, 225 is coupled to the output of the double-resonance circuit 210 ,220, wherein the second connection point of capacitor B 214, 224 is coupled to a first connection point of a third inductance 216, 226, wherein the respective second connection point of capacitor A 213, 223 and C 215, 225 are further coupled together with a second connection point of the third inductance 216, 226, which in turn may be coupled to ground.

[00038] Each double-resonance circuit 210 and 220 thus has the respective capacitor B denoted 214 and 224 in figure 2a respectively connected in series with the third inductance 216 and 226 respectively. The inductance may be a separate component or an imperfection of the respective capacitor B 214 and 224 respectively. The inductance may further be a property of the connections or of the metallic wire within parts of the double-resonance circuit 210 and 220. Still further, the third inductance may be any combination of a separate component, an imperfection of the respective capacitor B and/or a property of the metallic wire within parts of the double-resonance circuit.

[00039] It shall be pointed out that the capacitance B 214, 224 and the inductance 216, 226 may be implemented or realised such that the inductance 216, 226 is "above" the capacitance B 214, 224 in figure 2a. In other words the capacitance B 214 and inductance 216 may change places as well as the B 224 and inductance 226.

[00040] Consequently, each double-resonance circuit 210, 220 comprises two inductances 21 1 , 212, 221 , 222 connected in series between the input of the double-resonance circuit and an output of the double-resonance circuit, and three capacitors A denoted 213 and 223 in figure 2a, B denoted 214 and 224 in figure 2a, and C denoted 215 and 225 in figure 2a, each having a first and a second connection point, wherein the three capacitors are arranged in parallel with regards to each other such that the first connection point of capacitor A 213, 223 is coupled to the input of the double-resonance circuit 210 ,220, the first connection point of capacitor B 214, 224 is coupled to a second connection point of a third inductance 216, 226, wherein a first connection point of the third inductance 216, 226 is coupled to a point between the two inductances 21 1 , 212, and 221 , 222 respectively, and the first connection point of capacitor C 215, 225 is coupled to the output of the double-resonance circuit 210 ,220, wherein the respective second connection point of capacitor A 213, 223, B 214, 224 and C 215, 225 are further coupled together with a second connection point of the third inductance 216, 226, which in turn may be coupled to ground.

[00041 ] Further, the capacitor E 232 and the inductance 234 may change place in figure 2a. Thus, any pairwise permutation of series capacitance and inductance such as 214<->216, 232 <->234, and 224<->226 is a valid implementation.

[00042] For an inductance and a capacitance connected in series, there is a frequency value where the aggregated impedance is zero, wherein a signal of that frequency may pass nearly non-attenuated through the inductance and the capacitance. The frequency of signals that may pass through the capacitor B 214, 224 and the inductance 216, 226 is dependent upon the size of the capacitance of capacitor B 214, 224 and the inductance 216, 226. For the first double-resonance circuit 210, the frequency for nearly zero impedance of capacitor B 214 together with inductance 216 gives rise to notch N1 and for the second double-resonance circuit 220, the frequency for nearly zero impedance of capacitor B 224 together with inductance 226 gives rise to notch N2. [00043] According to yet embodiment, the coupling circuit 230 comprises three capacitors D 231 , E 232 and F 233, and a inductance 234, wherein a first connection point of capacitor D 231 is coupled to the output of the first double- resonance circuit 210 and a second connection point of capacitor D 231 is coupled to a first connection point of capacitor F 233, wherein a second connection point of capacitor F 233 is coupled to the input of the second double-resonance circuit 220, wherein a first connection point of capacitor E 232 is coupled to the second connection point of capacitor D 231 and the first connection point of capacitor F 233, and a second connection point of capacitor E 232 is coupled to a first connection point of the inductance 234 of the coupling circuit 230, wherein a second connection point of the inductance 234 of the coupling circuit is connected to the second connection point of the respective second connection point of capacitor A 231 , 223 and C 215, 225 of the first and the second double-resonance circuit and the second connection point of the third inductance 216, 226 of the first and the second double-resonance circuit 210, 220, which in turn may be coupled to ground.

[00044] The coupling circuit 230 has the function of a capacitance divider. A capacitor divider blocks any unwanted DC current bias flowing from the first double-resonance circuit 210 to the second double-resonance circuit 220.

[00045] Further, the coupling circuit 230 may affect the signals passing through it from the first double-resonance circuit 210 to the second double-resonance circuit 220 minimally attenuated. The coupling circuit 230 further serves as a resonance and impedance matching between the first and the second double-resonance circuit 210 and 220. The coupling circuit realises a nearly critical coupling between the first and the second double-resonance circuit 210 and 220.

[00046] According to yet embodiment, the bandpass filter is operational with regards to microwave frequencies.

[00047] Radio Frequency is a term that is often used to describe the number of times per second or oscillation of an electromagnetic field. Anything between 30 kHz and 300GHz is generally referred to as radio waves, and they are subdivided or classified depending on the actual frequency. Microwave is the general term used to describe radio waves that start from Ultra High Frequency, UHF, to Extremely High Frequency, EHF, which covers all frequencies between about 300 MHz to 300 GHz, lower frequencies are generally referred to as radio waves while higher frequencies are generally called millimetre waves. At low frequencies, parasitics in capacitors and inductors may usually be ignored, but in high frequency circuits parasitics can be a major problem as discussed above.

[00048] According to still an embodiment, capacitors A 213, 223, B 214, 224, C 215, 225 and E 232 are variable capacitors.

[00049] The higher the frequency, the more severe a parasitic capacitance that occurs in the circuits of the bandpass filter 200. By making the capacitors A 213, 223, B 214, 224, C 215, 225 and E 232 being variable capacitors, the bandpass filter 200 may be trimmed in order to compensate the effects of parasitic capacitances or inductances. Merely as an example, by adjusting the capacitance of capacitor B 214, the frequency of notch N1 may be moved along the frequency scale to somewhat higher or lower frequencies. Likewise for the capacitor B 224, by adjusting its capacitance, the frequency of notch N2 may be moved along the frequency scale to somewhat higher or lower frequencies. It shall be pointed out that the trimmer caps cold theoretically substituted by fixed caps, if possible to realise such fixed caps, especially at lower frequencies.

[00050] Likewise for capacitors A 213, 223, capacitors C 215, 225 and capacitor E 232, by varying their respective capacitance, the parasitics of e.g. the wiring of the bandpass filter may either be compensated for or exploited in order to make the bandpass filter operate as desired, e.g. for the specific passband, for the steepness for the slope at the end frequencies of the passband etc. Thus, by varying capacitor A 213 and capacitor C 215, the first double-resonance circuit 210 may be tuned. By varying capacitor A 223 and capacitor C 225, the second double-resonance circuit 220 may be tuned. By varying capacitor E 232, the coupling circuit 230 may be tuned. [00051 ] According to an embodiment, capacitor B 214, 224 connected in series with the third inductance 216, 226 comprises a series resonance circuit which results in the respective notch N1 and N2, wherein by varying the capacitance of capacitor B 214, 224 the respective frequency of N1 and N2 is changed.

[00052] As described above, the capacitors B 214, 224 connected in series with the third inductance 216, 226 comprise series resonance circuits, wherein at a specific frequency, which is dependent on the capacitance and inductance, the signal of that specific frequency passes with low attenuation through the capacitance and the inductance. Since the series of inductance 216, 226 and capacitance 214, 224 is in shunt configuration, e.g. one connection point of the series of inductance 216, 226 and capacitance 214, 224 is coupled to ground, this means that the respective notch N1 and N2 appear.

[00053] According to yet an embodiment, the third inductance 216, 226 is dependent on parasitic inductance of capacitor B 214, 224.

[00054] As described above, the different components of the bandpass filter, or generally most electrical components, are non-ideal. As high frequencies, typically at microwave frequencies, the non-ideal components give rise to parasitic capacitance and/or inductance. Also the metallic wire connecting the different components in the circuit, i.e. the bandpass filter, is non-ideal thereby also giving rise to parasitic capacitance and/or inductance, especially at microwave frequencies. Thus, the inductances 216, 226 may comprise only parasitic effects from the capacitors B 214, 224 and/or the metal wiring of the circuits of the bandpass filter, or the third inductance 216, 226 may be a separate component having an inductance which may be at least partly dependent on parasitic effects from e.g. the capacitor B 214, 224 and/or the metal wiring of the circuits of the bandpass filter.

[00055] Different transfer functions of different bandpass filters are illustrated in figure 3a. In figure 3a, the curves denoted by triangles illustrate a tight coupling, the curve denoted by squares illustrates a critical coupling and the curve denoted by diamonds illustrates a loose coupling. Different circuit may be more or less loosely coupled. In electrical circuits, the terms loose, critical and tight coupling are often used. Mutual inductance occurs when the change in current in one inductor induces a voltage in another nearby inductor. It is important as the mechanism by which transformers work, but it can also cause unwanted coupling between conductors in a circuit. The mutual inductance is also a measure of the coupling between two inductors. The mutual inductance also has a relationship with the coupling coefficient. The coupling coefficient is always between 1 and 0, and is a convenient way to specify the relationship between certain orientations of inductors with arbitrary inductance. When two tuned circuits are loosely coupled through mutual inductance, the bandwidth will be narrow. As the amount of mutual inductance increases, the bandwidth continues to grow. When the mutual inductance is increased beyond a critical point, the peak in the response curve begins to drop, and the centre frequency will be attenuated more strongly than its direct sidebands. This is known as over-coupling.

[00056] For narrowband filters where the steepness is high, the coupling is nearly critical. Critical coupling is the boundary where the frequency dependence of the transfer function, say S21 , has one or two maxima. For coupling less than critical, there is one maximum while for coupling greater than critical there are two maxima, see figure 3a. The latter is used for broader passband bandpass filters, while the former is used for narrower passband bandpass filters.

[00057] Critical coupling corresponds to small value of L3 in figure 1c compared to the values of L1 and L2. At microwave frequencies, realisation of coupling becomes difficult so T equivalent has to be used instead. But due to the lower bound of realisable L3 value, readability of this kind of filter is questionable at microwave frequencies.

[00058] In figure 3b, the inductance L3 of figure 1c is replaced by a general coupling impedance, COUPL. IMP. The impedance Z corresponds to w ^* L3, where w is the angular frequency. However, Z may be replaced by a capacitance that shows the same absolute value of the impedance, i.e. C=1/(w ^A 2 ^* L) at a desired angular frequency w. For nearly critical coupling, the capacitance in the place of the impedance Z in figure 3b is also possible as just described. But the required small absolute value of Z results in high capacitance needed. Again there is a realisability limit, because capacitor realisations prohibit a serial resonance frequency due to the parasitic serial inductance in real capacitors.

[00059] Figure 4a is an illustration of an example of a bandpass filter according to an exemplifying embodiment having specific values for the different

capacitances and inductances. In this example, L1 =3.6nH, L2=3.6nH, L3=1 nH, L4=0nH, L5=3.6nH, L6=3.6nH, and L7=1 nH. Further in this example, C1 =0.7pF, C2=4.2pF, C3=15pF, C4=4pF, C5=1 pF, C6=1 pF, C7=5pF, C8=3.8pF, C9=20pF, C10=4pF, and C1 1 =0.7p, Fc=1215 MHz . In the example of figure 4a, the centre frequency is 1215 MHz and the bandwidth is 40MHz.

[00060] By the bandpass filter illustrated in figure 4a, one transmission zero is placed below and the other above the passband of the bandpass filter. Figure 4b illustrates a transfer function of the bandpass filter illustrated in figure 4a. The transmission zeroes are due to the notches described above.

While the embodiments have been described in terms of several embodiments, it is contemplated that alternatives, modifications, permutations and equivalents thereof will become apparent upon reading of the specifications and study of the drawings. It is therefore intended that the following appended claims include such alternatives, modifications, permutations and equivalents as fall within the scope of the embodiments and defined by the pending claims.