WO2012040879A1

WO2012040879A1 - Multi-band switched-mode power amplifier

Info

Publication number: WO2012040879A1
Application number: PCT/CN2010/001533
Authority: WO
Inventors: Linsheng Liu
Original assignee: Telefonaktiebolaget Lm Ericsson (Publ)
Priority date: 2010-09-30
Filing date: 2010-09-30
Publication date: 2012-04-05

Abstract

A power amplifier (20) for operation in at least a first and a second frequency band is disclosed. The power amplifier (20) comprises a transistor (35) for amplifying an input signal of the power amplifier (20). Furthermore, the power amplifier (20) comprises an output coupling network (45) for connecting the power amplifier to a resistive load (55), wherein the output coupling network (45) is operatively connected to an output terminal (40) of the transistor (35), has an output terminal (50) for connection to said resistive load (55), and is configured such that, when the power amplifier (20) is connected to said resistive load (55), the power amplifier (20) is arranged to operate in class E for frequencies in one of the first and the second frequency bands, and to operate in inverse class E for frequencies in the other one of the first and the second frequency band.

Description

MULTI-BAND SWITCHED-MODE POWER AMPLIFIER

Technical Field

The present invention relates to a power amplifier for operation in multiple frequency bands.

Background

Various types of equipment for use in radio communication systems, such as cellular radio communication systems, may be provided with capability of transmitting data in a plurality of different frequency bands. Examples of such equipment are e.g. mobile phones and radio base stations. It is desirable that such equipment is capable of switching smoothly between operation in different frequency bands and in accordance with different standards and specifications.

One component that is important in this aspect is the power amplifier, which is used for feeding one or more transmitting antennas with an electric signal having a relatively high power level. One solution for enabling transmission in different frequency bands is to use multiple single-band power amplifiers and corresponding transmission filters, and select which one of the single-band power amplifiers and corresponding transmission filter that is currently used in accordance with the frequency band to be used using switches. However, such a solution is relatively costly, e.g. in terms of required circuit area. Hence, a simpler and less costly solution is desirable.

Summary

An object of the present invention is to provide a power amplifier capable of operation in a plurality of frequency bands.

According to a first aspect, there is provided a power amplifier for operation in at least a first and a second frequency band, wherein a center frequency s of the second frequency band is higher than a center frequency i of the first frequency band. The power amplifier comprises a transistor for amplifying an input signal of the power amplifier. The power amplifier further comprises an output coupling network for connecting the power amplifier to a resistive load. The output coupling network is operatively connected to an output terminal of the transistor. Furthermore, the output coupling network has an output terminal for connection to said resistive load. Moreover, the output coupling network is configured such that, when the power amplifier is connected to said resistive load, the power amplifier is arranged to operate in class E for frequencies in one of the first and the second frequency bands, and to operate in inverse class E for frequencies in the other one of the first and the second frequency band.

According to some embodiments, neither the second harmonic frequency band related to the second frequency band nor the second frequency band coincides or overlaps with any of the second and third harmonic frequency bands related to the first frequency band.

The output coupling network may be configured such that the load seen by the transistor is essentially capacitive for at least the second and third harmonic frequency bands related to said one of the first and the second frequency bands for which the power amplifier, when connected to said resistive load, is arranged to operate in class E.

The output coupling network may be configured such that the load seen by the transistor is essentially inductive for at least the second and third harmonic frequency bands related to said other one of the first and the second frequency bands for which the power amplifier, when connected to said resistive load, is arranged to operate in inverse class E.

The output coupling network may comprise transmission-line circuits arranged to provide impedance transformations to facilitate said class E and inverse class E operation in the respective ones of the first and the second frequency band.

In addition, the power amplifier may further be arranged for operation in at least one additional frequency band. Said at least one additional frequency band may be a plurality of additional frequency bands.

The output coupling network may be configured such that, when the power amplifier is connected to said resistive load, the power amplifier is arranged to operate in class F in every other one of said plurality of additional frequency bands, and to operate in inverse class F in the remaining ones of the plurality of additional frequency bands. The center frequencies fai- fa* '—- fax of the additional frequency bands, wherein N is the number of such additional frequency bands, may be related according to fa - ^{15 1} faU-t) for / = 2 . - . W . According to a second aspect, there is provided a radio-transmitter circuit that comprises the power amplifier according to the first aspect.

According to a third aspect, there is provided a radio-communication apparatus that comprises the radio transmitter circuit according to the second aspect. The radio- communication apparatus may e.g. be, but is not limited to, a mobile phone or a radio base station.

Further embodiments of the invention are defined in the dependent claims.

It should be emphasized that the term "comprises/comprising" when used in this specification is taken to specify the presence of stated features, integers, steps, or components, but does not preclude the presence or addition of one or more other features, integers, steps, components, or groups thereof.

Brief Description of the Drawings

Further objects, features and advantages of embodiments of the invention will appear from the following detailed description, reference being made to the

accompanying drawings, in which:

Fig. 1 schematically illustrates a mobile terminal in communication with a radio base station;

Fig. 2 is a block diagram of a radio-transmitter circuit according to an embodiment of the present invention;

Fig. 3 is a block diagram of a power amplifier according to an embodiment of the present invention;

Fig. 4 is a schematic circuit diagram of an output coupling network according to an elucidating embodiment of the present invention;

Figs.5-7 show plotted simulation results for the embodiment in Fig. 4;

Fig. 8 is a schematic circuit diagram of an output coupling network according to an elucidating embodiment of the present invention; and

Figs.9-12 show plotted simulation results for the embodiment in Fig. 4.

Detailed Description

Fig. 1 illustrates schematically an environment where embodiments of the present invention may be employed. A mobile terminal (MT) 1 , in Fig. 1 depicted as a mobile phone, communicates wirelessly via radio signals with a radio base station (BS) , e.g. in a cellular communication network. The MT 1 and the BS 2 are non-limiting examples of what is generically referred to below as a "radio-communication apparatus".

Fig. 2 is a block diagram of a radio-transmitter circuit 10 according to an embodiment of the present invention. The radio-transmitter circuit 10 may be comprised in a radio-communication apparatus, such as the MT 1 or BS 2. According to the embodiment, the radio-transmitter circuit 10 comprises a signal-generation unit 15 and a power amplifier (PA) 20. The signal generation unit 15 is adapted to generate an input signal to the PA 20 and supply the input signal to an input terminal 22 of the PA 20. An output terminal 24 of the PA 20 is operatively connected to an antenna 30, which is fed by the PA 20. The antenna 30 may be a dedicated transmit antenna. Alternatively, the antenna 30 may be a combined transmit and receive antenna which is shared between the radio-transmitter circuit 10 and a radio-receiver circuit of the same radio- communication apparatus.

Fig. 3 is a block diagram of the PA 20 in accordance with an embodiment of the present invention. According to the embodiment, the PA 20 is adapted for operation in at least a first and a second frequency band. A center frequency f* of the second frequency band is higher than a center frequency f of the first frequency band.

Furthermore, according to the embodiment, the PA 20 comprises a transistor 35 for amplifying the input signal of the PA 20. In Fig. 3, the transistor is depicted as an

NMOS transistor. This is, however, only an example. Other types of transistors may be used in other embodiments. Moreover, according to the embodiment illustrated in Fig. 3, the PA 20 comprises an output coupling network 45 for connecting the power amplifier to a resistive load 55. The output coupling network 45 is operatively connected to an output terminal 40 of the transistor 35. For the particular example depicted in Fig. 3 using an NMOS transistor, the output terminal 40 is the drain terminal of the NMOS transistor. Furthermore, the output coupling network has an output terminal 50 for connection to said resistive load 55. Moreover, the output coupling network is configured such that, when the PA 20 is connected to said resistive load 55, the PA 20 is arranged to operate in class E for frequencies in one of the first and the second frequency bands, and operate in inverse class E (also referred to as class IE) for frequencies in the other one of the first and the second frequency band. Thereby, a single PA 20 capable of operating in multiple frequency bands is obtained. This is advantageous over the solution using multiple single-band PAs discussed in the background section since a reduction of an overall cost, such as circuit area, is facilitated. As illustrated in Fig. 3, the PA 20 may further comprise a biasing unit 60 arranged to bias the PA 20 at a suitable operating point.

Class E and class IE amplifiers belong to the group of so called switched-mode amplifiers, where the transistor is operated as a switch having two states; on (i.e.

saturated) and off (i.e. deeply pinched off). An advantage of class E and class IE amplifiers, e.g. compared with class B amplifiers, is that they are capable of providing a higher power-added efficiency (P AE) with essentially the same saturated output power.

In class E and class IE amplifiers, an output coupling network (corresponding to the output coupling network 45 in Fig. 3) is tuned such that the load impedance seen by the transistor (from the output terminal of the transistor) at one or more harmonic frequencies is essentially (ideally completely) reactive. In the case of a class E amplifier, the load impedance seen by the transistor at said one or more harmonic frequencies is essentially capacitive. On the other hand, in the case of a class IE amplifiers, the load impedance seen by the transistor at said one or more harmonic frequencies is essentially inductive. Class E and class IE amplifiers can be regarded as each other's electrical duals.

In order to get the best possible efficiency in a class E amplifier, the load impedance should be essentially (ideally completely) capacitive for all harmonic frequencies. Likewise, in order to get the best possible efficiency in a class IE amplifier, the load impedance should be essentially (ideally completely) inductive for all harmonic frequencies. However, in practice, e.g. due to problems with parasitic capacitances and inductances, and taking into account the resulting complexity of the output coupling network, most realizations of class E and class IE amplifiers only consider the second harmonic (i.e. at ²f where / is the fundamental frequency) and the third harmonic (i.e. at 3 )_; which are the harmonics that have the largest impact on the efficiency of the PA (and possibly a few more higher order harmonics as well). Accordingly, for class E operation considered in embodiments of the present invention presented in the following, the load impedance as seen by the transistor is essentially (or completely) capacitive at least for the second and third harmonics. Hence, according to some embodiments, the output coupling network 45 is configured such that the load seen by the transistor 35 is essentially capacitive for at least the second and third harmonic frequency bands related to said one of the first and the second frequency bands for which the PA 20, when connected to said resistive load 55, is arranged to operate in class E.

Similarly, for class IE operation considered in embodiments of the present invention presented in the following, the load impedance as seen by the transistor is essentially (or completely) inductive at least for the second and third harmonics. Hence, according to some embodiments, the output coupling network 45 is configured such that the load seen by the transistor 35 is essentially inductive for at least the second and third harmonic frequency bands related to said other one of the first and the second frequency bands for which the PA 20, when connected to said resistive load 55, is arranged to operate in class IE.

The inventor has realized that the combined operation in class E in one band and class IE in another band is facilitated in cases where neither the second harmonic frequency band related to the second frequency band nor the second frequency band coincides or overlaps with any of the second and third harmonic frequency bands related to the first frequency band. This ensures that there are no conflicting

requirements on the load impedance in any of the fundamental frequency bands, nor in any of the second or third harmonic frequency bands. For example, there is no requirement stating that the load impedance in an harmonic frequency band should be capacitive to provide class E operation and, at the same time, be inductive to provide class IE operation (which could e.g. be the case if the second harmonic frequency band relating to the second frequency band had coincided with the third harmonic frequency band relating to the first frequency band). For embodiments where tuning of the load impedance (to be capacitive for class E operation and inductive for class IE operation) is considered in other harmonic frequency bands, in addition to the respective second and third harmonic frequency bands, additional requirements may be set on which frequency bands should not be allowed to coincide or overlap to avoid conflicting requirements in these bands. In essence, neither the second frequency band, nor any of the harmonic frequency bands relating to the second frequency band that are considered for such tuning of the load impedance should be allowed to coincide or overlap with any of the harmonic frequency bands relating to the first frequency band that are considered for such tuning of the load impedance.

The design of an optimal (or close to optimal) class E or IE amplifier can e.g. follow the following basic principles:

- The active device (i.e. transistor) may be chosen and the driver (i.e. the circuitry responsible for generating the input voltage to the transistor) may be designed so as to minimize the switch on voltage and off leakage current, respectively. (These are independent of the design of the load network.)

For class E:

- The drain voltage should be brought back to zero (or close to zero) at the time of transistor turn-on (so called zero voltage switching).

- The slope of the drain voltage should be zero (or close to zero) at the time of turn-on (so called zero voltage derivative switching).

Correspondingly, for class IE:

- The drain current should be brought back to zero at the time of transistor turn-on (so called zero current switching);

- The slope of the drain current should be zero at the time of turn-on (so called zero current derivative switching).

Waveforms for the switch current (i.e. current through the switch or transistor) and the switch voltage (i.e. voltage across the switch or transistor) are sketched in Fig. 4 for an ideal class E amplifier. The labels "ON" and "OFF" on the time axis indicate the time periods for which the switch is on and off, respectively. The corresponding waveforms for a class IE amplifier are sketched in Fig. 5.

For an actual implementation of a class E amplifier, the slope of the switch voltage at turn-off and the duty cycle of the switch are design parameters.

Correspondingly, for an actual implementation of a class IE amplifier, the slope of the switch current at turn-off and the duty cycle of the switch are design parameters. According to an idealized analysis, high efficiency operation is possible with a variety of values of the turn-off slope (of the switch voltage in class E or switch current in class IE) and duty cycle. The optimum operating point has a 50-percent duty cycle and a turn- off slope of 0. This combination not only produces the peak power output capability for a given device, but also prevents negative switch voltages as well as negative switch currents from arising.

In addition to the first and the second frequency band, the PA 20 may in some embodiments be adapted to operate in one or more additional frequency bands. In each of these additional frequency bands, the PA 20 may be adapted to, when connected to said resistive load, operate in any suitable class. For example, in some embodiments, output coupling network 45 may be configured such that, when the PA 20 is connected to said resistive load 55, the PA 20 is arranged to operate in class F in every other one of a plurality of additional frequency bands and operate in inverse class F (also referred to as class IF) in the remaining ones of the plurality of additional frequency bands.

Class F and class IF amplifiers also belong to the group of so called switched- mode amplifiers. In class F and class IF amplifiers, an output coupling network

(corresponding to the output coupling network 45 in Fig. 3) is tuned to selectively block one or more harmonics while shorting one or more other harmonics, as is described in more detail below. In a class F amplifier, at least the third harmonic (i.e. at ³ , where f is the fundamental frequency) is blocked. Possibly, one or more additional odd harmonics may be blocked as well. At the same time, at least the second harmonic (i.e. at 2/ ) is shorted out. Possibly, one or more additional even harmonics may be shorted out as well. Conversely, in a class IF amplifier, at least the third harmonic, and possibly one or more additional odd harmonics, is shorted out. At the same time, at least the second harmonic, and possibly one or more additional even harmonics, is blocked.

Generally, the more odd harmonics that are blocked (class F) or shorted out (class IF), and the more even harmonics that are shorted out (class F) or blocked (class IF), the higher the PAE becomes. As the number of the manipulated harmonics increases, the efficiency of an ideal class F or class IF PA increases from the 50% toward unity.

However, the non-idealities of the transistor output capacitance and the parasitics make it nearly impossible to make all even and odd harmonics well tuned. Consequently, most practical realizations of microwave class F and class IF amplifiers consider only a few harmonics, mostly the second and third harmonics, which have the most effects on the efficiency and output power enhancement.

The inventor has realized that relatively low circuit complexity can be obtained when the center frequencies fta-fa*>—· ΤαΝ of the additional frequency bands, wherein W is the number of such additional frequency bands, are related according to

faj = 1·5 · fa(j-i) for / = 2,3, .... JV (the ^α in the index of the center frequencies of the additional frequency bands stands for "additional" and should thus be seen as a label and not as a numeric parameter). As an example, in that case, the second harmonic of fat coincides with the third harmonic of fai . Thereby, for embodiments where the PA 20 is adapted to operate in class F for frequencies in the first additional frequency band (center frequency fa ) and in class IF for frequencies in the second additional frequency band (center frequency fai ), the same circuitry that is responsible for blocking the third harmonic of at (in class F operation) can be used for blocking the second harmonic of fai (in class IF operation). Similarly, for embodiments where the PA 20 is adapted to operate in class IF for frequencies in the first additional frequency band and in class F for frequencies in the second additional frequency band, the same circuitry that is responsible for shorting out the third harmonic of

(in class IF operation) can be used for shorting out the second harmonic of fat (in class F operation).

According to some embodiments, the PA 20 is arranged for operation in class F for frequencies in the first additional frequency band, for operation in class IF for frequencies in the second additional frequency band, etc. In these embodiments, the output coupling network 45 may be adapted to (when the PA 20 is connected to the resistive load 55), provide an essentially lossless electrical connection between the output terminal 40 of the transistor 35 and the resistive load 55 at the frequencies fai , fai , ... (the fundamental frequencies of the respective bands), provide an open circuit between the output terminal 40 of the transistor 35 and the resistive load 55 at the frequencies ³ fa = ² fa* , ³ f s— ² fa* , ³faizk-u = ²faak) , etc, (thereby providing the aforementioned blocking at these frequencies), and provide a short circuit between the output terminal 40 of the transistor 35 and a reference-voltage node (e.g. ground) at the frequencies ½i , 2f_aa , ²faisk-i> , etc, and ³f s , ³ fa* , %ί»), etc (thereby providing the aforementioned shorting out at these frequencies).

Furthermore, according to other embodiments, the PA 20 is arranged for operation class IF for frequencies in the first additional frequency band, for operation in class F for frequencies in the second additional frequency band, .... In these

embodiments, the output coupling network 45 may be adapted to (again when the PA 20 is connected to the resistive load 55) provide an essentially lossless electrical connection between the output terminal 40 of the transistor 35 and the resistive load 55 at the frequencies fai , fas , ... (the fundamental frequencies of the respective bands), provide an open circuit between the output terminal 40 of the transistor 35 and the resistive load 55 at the frequencies ²Αι , 2/_{a a} , 2 _o(2k__{0 ? e}t_C) and ½s , ³ m , ³ aC3« , etc, (thereby providing the aforementioned blocking at these frequencies), and provide a short circuit between the output terminal 40 of the transistor 35 and a reference- voltage node (e.g. ground) at the frequencies 3/ai = 2/_{α3 >} 3/_o3 = 2 _{Q4 j} ^"ifa&k-a— ²fat2ia , etc (thereby providing the aforementioned shorting out at these frequencies).

Optimum load impedances (seen by the transistor 35) for different amplifier classes are given below for the fundamental frequency (denoted fo ), the second harmonic (denoted ²/o), and the third harmonic (denoted ³ Ό). V_da denotes the supply voltage and Pout denotes the output power.

The output coupling network 45 may comprise transmission-line circuits arranged to provide impedance transformations to facilitate said class E and inverse class E operation in the respective ones of the first and the second frequency band. An example of such an embodiment of the output coupling network 45 is illustrated in Fig. 6. The embodiment of the output coupling network illustrated in Fig. 6 is suitable TIT TE HEET RULE 26 for use in embodiments of the PA 20 adapted for operation in class E in the first frequency band and in class IE in the second frequency band. In the embodiment of the output coupling network 45 illustrated in Fig. 6, a DC-blocking capacitor 70 is provided in order to prevent the DC level of the transistor output from propagating to the resistive load 55. For the RF (radio frequency) frequencies of interest (Λ , Λ , and their

respective harmonics), the DC-blocking capacitor 70 acts as a short circuit.

Furthermore, the embodiment of the output coupling network 45 illustrated in

λ

Fig. 6 comprises a transmission line 100 having a length of 6 at the frequency ft (where - λ denotes the corresponding wavelength). The transmission line 100 is responsible for tuning the desired reactance at the frequency 3 ₃ for providing the desired essentially inductive load impedance at this frequency.

Moreover, the embodiment of the output coupling network 45 illustrated in

A

Fig. 6 comprises a transmission line 110 having a length of 4 at the frequency ft . The transmission line 110 is responsible for tuning the desired reactance at the frequency

2/i for providing the desired essentially capacitive load impedance at this frequency.

The embodiment of the output coupling network 45 illustrated in Fig. 6 comprises a tuning line (transmission line) 120. Furthermore, the embodiment of the output coupling network 45 illustrated in Fig. 6 comprises a transmission line 130

λ_

having a length of 12 at the frequency i . The transmission line 130, together with the tuning line 120, is responsible for tuning the desired reactance at the frequency 3 j. for providing the desired essentially capacitive load impedance at this frequency.

Moreover, the embodiment of the output coupling network 45 illustrated in

λ

Fig. 6 comprises a transmission line 140 having a length of 4 at the frequency /¾ . The transmission line 140, together with the tuning line 120, is responsible for tuning the desired reactance at the frequency 2/₅ for providing the desired essentially inductive load impedance at this frequency.

In addition, tuning lines 150 and 160 are provided in the embodiment of the output coupling network illustrated in Fig. 6. These tuning lines 150 and 160 are responsible for providing the impedance matching at the fundamental frequencies A

Suitable dimensions of the tuning lines 120, 150, and 160 can be derived through computer simulations to obtain the desired load impedances at the various frequencies.

A corresponding embodiment of the output coupling network 45 suitable for use in an embodiment of the PA 20 adapted for operation in class IE in the first frequency band and in class E in the second frequency band can be obtained by simply interchanging A and A in Fig. 6 and in the above description of Fig. 6.

According to a particular example embodiment of the PA 20, where the output coupling network 45 is implemented as in Fig. 6, the supply voltage V&d— 28 v, the output power = 100 w (50 dBm), fx = 1900 MHz, and = ²⁰⁰⁰ MHz.

Optimum load impedances for the fundamental frequency bands, second harmonic frequency bands, and third harmonic frequency bands can be derived from the table above. Simulation results for this example embodiment, with the tuning lines 120, 150, and 160 tuned to (i.e. given the appropriate dimensions to) provide these optimum load impedances are plotted in Figs. 7-9 to illustrate the functionality. In these simulations, the transistor 35 has been modeled as an ideal switch.

Fig. 7 shows the power-added efficiency (PAE) 170 and the output power (Pout ) 180 as a function of RF frequency. At the frequency 1900 MHz (i.e. i ), the output power is 50.06 dBm, and the PAE is 95.5 %. At the frequency 2000 MHz (i.e. ), the output power is 50.00 dBm, and the PAE is 95.6 %.

Figs. 8 and 9 illustrate the simulated waveforms of the current (190 and 210, respectively) through the switch (transistor 35) and the voltage (200 and 220,

respectively) for operation in class E (i.e. at fi = 190O MHz) and for operation in class IE (i.e. at A = 2000 MHz), respectively.

Another example an embodiment of the output coupling network 45 is illustrated in Fig. 10. The embodiment of the output coupling network illustrated in Fig. 8 is suitable for use in embodiments of the PA 20 adapted for operation in class E in the first frequency band, in class IE in the second frequency band, and in class IF in a (first) additional frequency band (center frequency i ). Similarly to the embodiment of the output coupling network 45 illustrated in Fig. 6, a DC-blocking capacitor 70 is provided in order to prevent the DC level of the transistor output from propagating to the resistive load 55. Again, for the RF (radio frequency) frequencies of interest (A , , f and their respective harmonics), the DC-blocking capacitor 70 acts as a short circuit.

The transmission lines 100-140 have the same purpose as in Fig. 6 and are not further described. In addition, the embodiment of the output coupling network 45 illustrated in Fig. 10 comprises transmission lines 230 and 240. The transmission lines 230 and 240 are responsible for providing the desired open circuit at the frequency ² fax and the desired short circuit at the frequency , thereby facilitating operation in class IF in the first additional frequency band. Moreover, the embodiment of the output coupling network 45 illustrated in Fig. 10 comprises transmission lines 230 and 240. The transmission lines 230 and 240 are responsible for providing the impedance matching at the fundamental frequencies , fs , and f x .Suitable dimensions of the tuning lines 120, 240, 250, and 260 can be derived through computer simulations to obtain the desired load impedances at the various frequencies.

According to a particular example embodiment of the PA 20, where the output coupling network 45 is implemented as in Fig. 10, the supply voltage = ²⁸ V, the output power Po t = 00 W (50 dBm), ft = 1900 MHz, A = 2000 MHz, and

fax - l-660^> MHz. Optimum load impedances for the fundamental frequency bands, second harmonic frequency bands, and third harmonic frequency bands can be derived from the table above. Simulation results for this example embodiment, with the tuning lines 120, 240, 250, and 260 tuned to (i.e. given the appropriate dimensions to) provide these optimum load impedances are plotted in Figs. 1 1-14 to illustrate the functionality. In these simulations, the transistor 35 has, again, been modeled as an ideal switch.

Fig. 11 shows the PAE 270 and ^pous 280 as a function of RF frequency. At the frequency 1660 MHz (i.e. fa ), the output power is 51.6 dBm, and the PAE is 95.1 %. At the frequency 1900 MHz (i.e. f ), the output power is 49.7 dBm, and the PAE is 96.2 %. At the frequency 2000 MHz (i.e. fs ), the output power is 50.07 dBm, and the PAE is 95.9 %.

Figs. 12, 13, and 14 illustrate the simulated waveforms of the current (290, 310, and 330, respectively) through the switch (transistor 35) and the voltage (300, 320, and 340, respectively) for operation in class IF (i.e. at = i660 MHz), class E (i.e. at f = 1900 MHz) and for operation in class IE (i.e. at fi = 2000 MHz), respectively.

In the transmission-line realization of the output coupling network 45 presented in accordance with embodiments above, the output coupling network 45 can be considered as having two parts; one part responsible for harmonic termination and another part responsible for fundamental matching. In these embodiments, the part responsible for harmonic termination is located "first" (i.e. closest to the transistor) and the part responsible for fundamental matching is placed "last" (i.e. closest to the resistive load 55). This configuration has the advantage that the part responsible for the fundamental load does not affect the open/short terminations for the harmonics as seen from the side of the transistor 35.

Selection, or tuning, of transmission line dimensions, e.g. for the tuning lines, can be performed based on computer simulation to obtain the desired functionality. This is typically an iterative process, as tweaking of the dimensions of one transmission line may require additional tweaking of the dimensions of other transmission lines. Also, nonideal properties of the transistor 35, e.g. in terms of parasitics and other properties that causes the transistor to deviate from an ideal switch, should be taken into account in the design process. This iterative optimization process can be directed towards obtaining either a maximum PAE or a maximum ¾ut as a goal of the optimization.

It should be noted that not only the dimensions of the transmission lines labeled "tuning lines" in the figures, but also the tuning lines indicated in the figures as having a certain length, can be subject to refinement in the iterative design process. Thus, the indicated certain lengths of such transmission lines can be seen as suitable start values.

The present invention has been described above with reference to specific embodiments. However, other embodiments than the above described are possible within the scope of the invention. For example, numerous other structures of the output coupling network 45, both concerning the part responsible for harmonic termination and the part responsible for fundamental matching, are possible within the scope of the invention. The different features of the embodiments may be combined in other combinations than those described. The scope of the invention is only limited by the appended patent claims.

Claims

1. A power amplifier (20) for operation in at least a first and a second frequency band, wherein a center frequency of the second frequency band is higher than a center frequency ft. of the first frequency band, comprising

a transistor (35) for amplifying an input signal of the power amplifier (20); and an output coupling network (45) for connecting the power amplifier to a resistive load (55), wherein the output coupling network (45) is operatively connected to an output terminal (40) of the transistor (35), has an output terminal (50) for connection to said resistive load (55), and is configured such that, when the power amplifier (20) is connected to said resistive load (55), the power amplifier (20) is arranged to

- operate in class E for frequencies in one of the first and the second frequency bands; and

- operate in inverse class E for frequencies in the other one of the first and the second frequency band.

2. The power amplifier (20) according to claim 1, wherein neither the second harmonic frequency band related to the second frequency band nor the second frequency band coincides or overlaps with any of the second and third harmonic frequency bands related to the first frequency band.

3. The power amplifier (20) according to claim 2, wherein the output coupling network (45) is confi ured such that the load seen by the transistor (35) is essentially capacitive for at least the second and third harmonic frequency bands related to said one of the first and the second frequency bands for which the power amplifier (20), when connected to said resistive load (55), is arranged to operate in class E.

4. The power amplifier (20) according to claim 2 or 3, wherein the output coupling network (45) is configured such that the load seen by the transistor (35) is essentially inductive for at least the second and third harmonic frequency bands related to said other one of the first and the second frequency bands for which the power amplifier (20), when connected to said resistive load (55), is arranged to operate in inverse class E.

5. The power amplifier (20) according to any preceding claim, wherein the output coupling network comprises transmission-line circuits arranged to provide impedance transformations to facilitate said class E and inverse class E operation in the respective ones of the first and the second frequency band.

6. The power amplifier (20) according to any preceding claim, further arranged for operation in at least one additional frequency band.

7. The power amplifier (20) according to claim 6, wherein said at least one additional frequency band is a plurality of additional frequency bands.

8. The power amplifier (20) according to claim 7, wherein the output coupling network (45) is configured such that, when the power amplifier (20) is connected to said resistive load (55), the power amplifier (20) is arranged to:

- operate in class F in every other one of said plurality of additional frequency bands; and

- operate in inverse class F in the remaining ones of the plurality of additional frequency bands.

9. The power amplifier (20) according to claim 8, wherein the center frequencies fai- faj - -·· · fax of the additional frequency bands, wherein ^ΛΓ is the number of such additional frequency bands, are related according to f *i ^{~ 1-5 "} f ²i/-¹) for / = 2,3, ... ,iV .

10. A radio-transmitter circuit (10) comprising the power amplifier (20) according to any preceding claim.

11. A radio-communication apparatus (1, 2) comprising the radio transmitter circuit (10) according to claim 10.

12. The radio-communication apparatus (1) according to claim 11, wherein the radio-communication apparatus (1) is a mobile phone.

13. The radio-communication apparatus (2) according to claim 12, wherein the radio-communication apparatus (2) is a radio base station.