WO2016056951A1

WO2016056951A1 - Power amplifier for amplification of an input signal into an output signal

Info

Publication number: WO2016056951A1
Application number: PCT/SE2014/051155
Authority: WO
Inventors: Richard Hellberg
Original assignee: Telefonaktiebolaget L M Ericsson (Publ)
Priority date: 2014-10-06
Filing date: 2014-10-06
Publication date: 2016-04-14

Abstract

A power amplifier (100) comprising a first and a second sub-amplifier (111, 12) for amplification of an input signal into an output signal. An input network (120) of the power amplifier (100) comprises a first input branch (101) configured to provide a first signal, derived from the input signal. An average rate of amplitude change of the first signal versus amplitude change of the input signal in a first amplitude range is less than an average rate of amplitude change of the first signal versus amplitude change of the input signal in a second amplitude range. The input network (120) further comprises a second input branch (102) configured to provide a second signal, derived from the input signal. The second signal has a triangularly shaped amplitude envelope which increases in the first amplitude range and which decreases in the second amplitude range. The power amplifier (100) is configured to provide the first signal to the first and second sub-amplifiers (111, 112), and provide the second signal to the first sub-amplifier (111).

Description

POWER AMPLIFIER FOR AMPLIFICATION OF AN INPUT SIGNAL INTO AN

OUTPUT SIGNAL

TECHNICAL FIELD

Embodiments herein relate to power amplifiers for wireless communication systems, such as telecommunication systems. In particular, a power amplifier for amplification of an input signal into an output signal is disclosed. Furthermore, a radio network node, comprising the power amplifier, and a user equipment, comprising the power amplifier, are disclosed.

BACKGROUND

Power amplifiers are widely used in communication systems, for example in radio base stations and cellular phones of a cellular radio network. In such cellular radio network, power amplifiers typically amplify signals of high frequencies for providing a radio transmission signal. A consideration in the design of power amplifiers is the efficiency thereof. High efficiency is generally desirable so as to reduce the amount of power that is dissipated as heat. Moreover, in many applications, such as in a satellite or a cellular phone, the amount of power that is available may be limited due to powering by a battery, included in e.g. the satellite. An increase in efficiency of the power amplifier would allow an increase of operational time between charging of the battery.

A conventional Power Amplifier (PA), such as class B, AB, F, has a fixed Radio Frequency (RF) load resistance and a fixed voltage supply. Class B or AB bias causes the output current to have a form close to that of a pulse train of half wave rectified sinusoid current pulses. The Direct Current (DC), and hence DC power, is largely proportional to the RF output current amplitude, and voltage. The output power, however, is proportional to the RF output current squared. An efficiency of the conventional power amplifier, i.e. output power divided by DC power, is therefore also proportional to the output amplitude. The average efficiency is consequentially low when amplifying signals that on average have a low output amplitude, or power, compared to the maximum required output amplitude.

Known RF power amplifiers include both Doherty and Chireix type power amplifiers. These kinds of RF PAs are generally more efficient than the conventional amplifier described above for amplitude-modulated signals with high Peak-to- Average Ratio (PAR), since they have a lower average sum of output currents from the transistors. Reduced average output current means high average efficiency.

The reduced average output current is obtained by using two transistors that influence each other's output voltages and currents through a reactive output network, which is coupled to a load. By driving the constituent transistors with the right amplitudes and phases, the sum of RF output currents is reduced at all levels except the maximum. Also for these power amplifiers the RF voltage at one or both transistor outputs is increased.

Generally, RF power amplifier can be driven in a so called backed off operation. This means that the power amplifier is operated a certain level, e.g.

expressed as a number of decibels (dBs), under its maximum output power. Backed off operation may also refer to that an instantaneous output power is relatively low.

WO03/0611 1 discloses a composite power amplifier including a first and a second power amplifier connected to an input signal over an input network and to a load over an output network. The composite power amplifier may be operated, typically over a 3 to 1 bandwidth, in Doherty mode, in Chireix mode or in other intermediate modes between the Doherty and Chireix modes. Thus, the 3 to 1 bandwidth of high efficiency is achieved by devising an output network 13 that has both suitable impedance transformation characteristics and full power output capacity over the bandwidth.

This kind of composite power amplifiers, sometimes referred to as multi-stage power amplifiers, are usually driven by complex digital adaptive control systems for providing multiple drive signals to be fed to sub-amplifiers of the composite power amplifier. The drive signals are typically created by individual digital-to-analog converters.

US7145387 discloses a system including a 2-stage high-efficiency power amplifier with increased robustness against circuit variations and with radically increased bandwidth of high efficiency. This system is operable in high-efficiency modes, for example at or near a Chireix mode but using Doherty-like drive signals. In more detail, US7145387 discloses a composite power amplifier including a first and a second power amplifier connected to an input signal over an input network and to a load over an output network. The output network includes phase shifting elements for generating different phase shifts from each power amplifier output to the load. The input network includes means for driving both power amplifiers to produce first output current components having an amplitude that increases linearly with increasing output signal amplitude below a transition point and decreases monotonically with increasing output signal amplitude above said point, and second output current components having an amplitude that increases linearly with increasing output signal amplitude both below and above the transition point. A problem with this system is that it is adapted for operation at one frequency only. Additionally, the drive signal control structure described in US7145387 is based on output network

transimpedance relations in 2-stage amplifiers. This can be implemented as an analog RF-in circuit. However, the analog RF-in circuit is difficult to operate automatically, i.e. to obtain the transimpedance relations, or over wide bandwidths. Moreover, a problem is that these transimpedance relations become severely more difficult to obtain for 3-stage amplifiers and further multi-stage amplifiers.

Further examples are found in e.g. C. Andersson et al., "A 44 dBm 1.0-3.0 GHz GaN Power Amplifier with over 45% PAE at 6 dB back-off", Proc. IMS 2013, and C. Andersson et al., "A 1-3 GHz digitally controlled dual-RF input power amplifier design based on a Doherty-Outphasing continuum analysis".

SUMMARY

An object is to provide a power amplifier of the above mentioned kind with improved analogue operation.

According to an aspect, the object is achieved by a power amplifier comprising a first and a second sub-amplifier for amplification of an input signal into an output signal. The first and second sub-amplifiers are connected to an input network for receiving the input signal at an input port of the input network, and the first and second sub-amplifiers are connected to an output network for providing the output signal at an output port of the output network. The input network comprises a first input branch configured to provide a first signal, derived from the input signal. An average rate of amplitude change of the first signal versus amplitude change of the input signal in a first amplitude range is less than an average rate of amplitude change of the first signal versus amplitude change of the input signal in a second amplitude range. The input network further comprises a second input branch configured to provide a second signal, derived from the input signal. The second signal has a triangularly shaped amplitude envelope which increases in the first amplitude range and which decreases in the second amplitude range. The power amplifier is configured to provide the first signal to the first and second sub-amplifiers, and provide the second signal to the first sub-amplifier. According to another aspect, the object is achieved by a radio network node, comprising the power amplifier. According to a further aspect, the object is achieved by a user equipment, comprising the power amplifier.

According to the embodiments herein, a universal way of controlling and driving a high-efficiency multi-stage power amplifier based on passive output combining networks is provided. The power amplifier according to embodiments herein may be used for single-frequency amplifiers as in prior art, but may be especially well suited for so called wideband power amplifiers.

As mentioned above, embodiments of the power amplifier makes use of two nonlinear amplitude shaped signals, i.e. the first signal and the second signal. The second signal rises at low amplitudes, e.g. in the first amplitude range, and decreases at high amplitudes, e.g. in the second amplitude range. The second signal is fed to at least one sub-amplifier, in preferred embodiments to a single sub- amplifier. The second signal is controlled in phase and amplitude in accordance with the transimpedance from a single sub-amplifier to the output. The first signal may be low, e.g. zero, in the first amplitude range and increasing in the second amplitude range. The second signal may be fed to all sub-amplifiers based on an in-phase combining criteria at the output port. In some embodiments, such as for wideband power amplifiers with output networks consisting of transmission lines, or generally, networks that behave like transmission lines within a bandwidth of interest, it is possible to make good wideband analog implementations according to embodiments herein due to the following reasons.

Firstly, in-phase combining may be achieved by a wideband passive distribution network of transmission lines whose length differences are the mirror image of the ones in the output network.

Secondly, the second signal may in preferred embodiments be routed to a single sub-amplifier by passive combining and gain setting/switching.

A wideband approximation to the ideal phase of the first signal (which is generally different from the in-phase combining phase) may in most cases be provided by inserting this signal at a point in the transmission line distribution network at the input side, possibly also with a static phase shift. The input network is therefore passive, except for switching used to select and adjust which sub-amplifier to be operated at low amplitudes, i.e. in the first and second amplitude ranges. BRIEF DESCRIPTION OF THE DRAWINGS

The various aspects of embodiments disclosed herein, including particular features and advantages thereof, will be readily understood from the following detailed description and the accompanying drawings, in which:

Figure 1 is a schematic overview of the power amplifier according to embodiments herein,

Figure 2 is a diagram illustrating amplitude envelopes for the first and second signal,

Figure 3a is an illustration of electrical lengths of transmission lines at center frequency,

Figures 3b-3d are diagrams including graphs over efficiency, transition point amplitude and relative phase versus frequency,

Figure 4 is another schematic overview of an exemplifying power amplifier according to embodiments herein,

Figure 5a is another illustration of electrical lengths of transmission lines at center frequency,

Figures 5b-5d are other diagrams including graphs over efficiency, transition point amplitude and relative phase versus frequency,

Figure 6a is a further illustration of electrical lengths of transmission lines at center frequency,

Figures 6b-6d are further diagrams including graphs over efficiency, transition point amplitude and relative phase versus frequency,

Figure 7 is a further schematic overview of another exemplifying power amplifier according to embodiments herein,

Figure 8 illustrates an exemplifying radio network node according to embodiments herein, and

Figure 9 illustrates an exemplifying user equipment according to

embodiments herein. DETAILED DESCRIPTION

Throughout the following description similar reference numerals have been used to denote similar features, such as elements, units, modules, circuits, nodes, parts, items or the like, when applicable. In the Figures, features that appear in some embodiments are indicated by dashed lines.

Figure 1 depicts an exemplifying power amplifier 100, such as a composite power amplifier, according to embodiments herein. The power amplifier 100 comprises a first and a second sub-amplifier 111 , 112 for amplification of an input signal into an output signal. In this manner, the first and second sub-amplifiers 11 1 , 1 12 may be operated to amplify the input signal into the output signal.

The first and second sub-amplifiers 1 11 , 1 12 are connected to an input network 120 for receiving the input signal at an input port 150 of the input network 120. Furthermore, the first and second sub-amplifiers 11 1 , 112 are connected to an output network 130 for providing the output signal at an output port 140 of the output network 130. The input network 120 comprises a first input branch 101 configured to provide a first signal, derived from the input signal. An average rate of amplitude change of the first signal versus amplitude change of the input signal in a first amplitude range is less than an average rate of amplitude change of the first signal versus amplitude change of the input signal in a second amplitude range. See e.g. the dashed line in Figure 2.

The first signal may have an amplitude envelope that is below 30% of a maximum value of an operational amplitude range of the power amplifier in the first amplitude range of the operational amplitude range of the power amplifier 100

In further examples, the amplitude envelope of the first signal is below 20%, below 10%, about zero, or zero, depending on desired efficiency of the power amplifier. This means that the lower the amplitude envelope is in the first amplitude range, the higher - thus better- becomes the efficiency of the power amplifier.

Accordingly, the highest efficiency is achieved when the amplitude envelope of the first signal is zero in the first amplitude range.

The amplitude envelope of the first signal is further increasing in the second amplitude range of the operational amplitude range, wherein amplitudes of the input signal in the first amplitude range are less than amplitudes of the input signal in the second amplitude range. Accordingly, the second amplitude range follows, directly or indirectly, after the first amplitude range. Expressed differently, the second amplitude range is subsequent to the first amplitude range with respect to magnitude of the amplitudes in the respective ranges.

The amplitude envelope may refer to the power amplifier's response to a test signal, such as the input signal, for which the amplitude is increasingly ramped up.

The input network 120 further comprises a second input branch 102 configured to provide a second signal, derived from the input signal. The second signal has a triangularly shaped amplitude envelope which increases in the first amplitude range and which decreases in the second amplitude range. The power amplifier 100 is configured to provide the first signal to the first and second sub- amplifiers 1 11 , 1 12, and provide the second signal to the first sub-amplifier 1 11.

As is shown in Figure 1 , the second signal may be based on the first signal. This means that the second signal may for example be achieved by subtracting the first signal from the input signal. The second signal may also be generated by subtracting the first signal from a copy of the input signal. In this manner, generation of the first and second signals is independent from each other. An advantage may hence be that amplitudes of the first and second signals also may be independently adjusted. The first signal may sometimes be referred to as a "Class C"-signal due to its characteristic amplitude response.

The second signal may also be generated in many other ways. For example, by means of variable gain amplifiers controlled by respective non-linearly processed peak detector signals.

The following various embodiments are contemplated.

The power amplifier 100 may comprise means configured to enable a concurrent super-positioning of the first and second signals with respect to phase and group delay. The means may be embodied in the form of a phase splitting network. Alternatively or additionally, the means may be realized by means of an offset insertion point as described herein.

As an example, the means may be part of the input network 120, i.e. the input network 120 is configured to enable concurrent super-positioning of the first and second signals with respect to phase and group delay. With the realization of the means in the form of the offset insertion point, it may be that the means may be configured to provide the first signal to the first and second sub-amplifiers 1 11 , 1 12 via a first sub-branch 105 and a second sub- branch 106, respectively. The power amplifier 100 may be configured to insert the second signal into the first sub-branch 105 at an insertion point of the first sub-branch 105. Phase and group delay of the triangularly shaped amplitude envelope due to the insertion point may enable the concurrent super-positioning of the first and second signals with respect to phase and group delay. Thanks to the insertion point a difference in electrical length for the first and second signals is achieved.

The output network 130 may comprise a first output transmission line 131 , connected to the first sub-amplifier 1 11 , and a second output transmission line 132, connected to the second sub-amplifier 112, wherein the input network 120 may comprise a first input transmission line 121 , connected to the second input branch 102, wherein electrical length of the first input transmission line 121 may be twice as long as electrical length of the first output transmission line 131 , wherein the insertion point may be located at an output 123 of the first input transmission line 121 .

The input network 120 may comprise a second input transmission line 122, connected to the second input branch 102, wherein electrical length of the second input transmission line 122 may be twice as long as electrical length of the second output transmission line 132, wherein a further insertion point at the second sub- branch 106 may be located at an output 124 of the second input transmission line 122, wherein phase and group delay of the triangularly shaped amplitude envelope due to the further insertion point may enable concurrent super-positioning of the first and second signals with respect to phase and group delay. As mentioned above, the means may comprise a phase splitting network configured to enable the concurrent super-positioning of the first and second signals with respect to phase and group delay.

The phase splitting network may be configured to provide the first and second signals at a common insertion point at which electrical length to the first and second sub-amplifiers 111 , 1 12 is equal.

In further embodiments, the means may comprise a constant phase difference network configured to enable the concurrent super-positioning of the first and second signals with respect to phase and group delay.

The power amplifier 100 may be configured to provide the second signal to the first sub-amplifier 11 1 for frequencies in a first frequency range, and to provide the second signal to the second sub-amplifier 1 12 for frequencies in a second frequency range, wherein the first frequency range may be different from the second frequency range.

The power amplifier 100 may have a transition interval whose length between the first and second amplitude ranges is configurable, whereby the first and/or second amplitude range is configurable. In some examples, the transition point is located at the end of the first amplitude range and at the beginning of the second amplitude range.

The input network 120 may comprise a first attenuator 161 , connected to the second input branch 102 and the first sub-amplifier 1 11 , and a second attenuator 162, connected to the second input branch 102 and the second sub-amplifier 112, wherein the power amplifier 100 may be configured to activate the first attenuator 161 when the second signal is in the first frequency range, and wherein the power amplifier 100 may be configured to activate the second attenuator 162 when the second signal is in the second frequency range.

The input network 120 may comprise a third attenuator 163, connected to an output 104 of the first input branch 101 and an input 103 of the second input branch 102, whereby the second input branch 102 is configured to provide the second signal. The third attenuator 163 controls mix of the first signal and the input signal.

High-efficiency multi-stage power amplifiers, aka composite power amplifiers, based on passive output combining networks come in many varieties. They are very different in their output network design and the shapes of drive signals to the sub- amplifiers, or contain this diversity within the wide usage band. Different as they are, they also share two characteristics.

A first characteristic is that the sum of the maximum output power of all sub- amplifiers should be able to combine at the output port 140.

A second characteristic is that there is at least one sub-amplifier that has higher transimpedance to the output port 140 than the other sub-amplifiers. That sub- amplifier is the one that should be active alone in the first amplitude range to increase efficiency. See Figure 2. The in-phase power combining of all sub-amplifiers to full power is achieved by providing the input nodes (gates) of the sub-amplifier transistors with drive signals of the correct phase differences and the correct amplitudes. The maximum drive voltage amplitudes are generally equal if the same types of transistors are used, but the drive currents are proportional to the sizes of the transistors, so the driver amplifiers must be scaled accordingly. The correct phase differences are such that the different phase shifts due to the line lengths, i.e. from amplifier to output, in the output network are cancelled. In this manner, the signal - after going through the different paths - has the same phase at the output.

The efficiency of any sub-amplifier 11 1 , 1 12 operating alone at low output amplitudes, i.e. in the first amplitude range, is proportional to the transimpedance from that transistor to the load. Transimpedance means voltage generated at a second node divided by the current input at the first node. High transimpedance to the load therefore means getting more RF voltage at the load for less RF current from the sub-amplifier's transistor. In case, multiple sub-amplifiers are available, the one with a suitable transimpedance to the output may be chosen as the active one at the lowest output amplitudes. The relative phase of the drive signal to this single sub- amplifier may be chosen so that it has about the same phase at the output as the collective in-phase combined full-amplitude signal.

A triangular amplitude drive function, i.e. the second signal, is used for the single sub-amplifier. As mentioned above, the second signal has a peak at a transition point at a medium output amplitude and may preferably be zero at zero and full output amplitude. As mentioned above, zero is understood to have some margin, such as 30% of the maximum operational amplitude. The transition point may separate, or may be located between, the first amplitude range and the second amplitude range. A drive amplitude function, i.e. the first signal, that may start at zero at the transition point and ends at the highest point is used for the collective drive of all amplifiers to the maximum output. These two signals will synthesize to a roughly linear total response (given roughly linear response of the power transistors) at the output if they are weighted correctly in amplitude and phase. Correct amplitude may be defined by a linear drive signal - as a reference - for one sub-amplifier which drive signal goes from zero to an amplitude at full power that for convenience is

normalized to 1. The first signal then also goes to 1 at full power. The second signal will have a linear slope in the low amplitude region, just as the linear reference signal, but with a voltage that is scaled down by the transimpedance, but also divided by the sub-amplifiers part of the output power, i.e. an increase since only one sub-amplifier is responsible for the first part, but all sub-amplifiers contribute to the full power ouptut. The correct phase is the one such that the output phase with single-amplifier drive is the same as the phase of the collectively driven, in-phase combined, output. An idealized example of the amplitudes of these signals as functions of output amplitude is shown in Figure 2 Notably, in practice the graphs may have less straight slopes, and a corner at the junction between the first and second amplitude ranges may be more or less rounded.

An exemplifying application relates to analog single RF input control for wideband amplifiers. In Figure 1 , the power amplifier 100 may be a 3:1 bandwidth wideband amplifier provided with an analog signal processing network, such as the input network 120, for controlling it. The transmission line lengths are given as the electrical length in wavelengths at center frequency of the power amplifier 100, i.e. the center frequency of the operational bandwidth of the power amplifier 100. All parts except the transmission lines are assumed to have zero time delay, or have corresponding matched delay in parallel sub-branches. In more detail, the first input transmission line 121 has an electrical length of one wavelength at center frequency, the second input transmission line 122 has an electrical length of one half wavelength at center frequency, the first output transmission line 131 has an electrical length of one half wavelength at center frequency and the second output transmission line 132 has an electrical length of one quarter wavelength at center frequency.

Still referring to Figure 1 , starting on the input port 150, the first signal may be provided by a class C amplifier, e.g. an amplifier 110, for the collective drive to all sub-amplifiers 1 11 , 112. For wideband operation, the power amplifier 100 may be resistively loaded in the full frequency range of interest. The correct phases for in- phase combining is achieved over the full frequency range by providing delay networks or lengths of transmission line in the sub-branches 105, 106, which thus may include the first and second input transmission lines 121 , 122, to the sub- amplifiers 1 11 , 112. As illustrated in Figure 1 , the total length, i.e. counted as time delay, from a signal splitting point after the class C amplifier, i.e. after the output 104 of the first input branch 101 , to the output port 140, which may be connected to an antenna (not shown) should be the same for the sub-branches 105, 106.

In some examples, the second signal is obtained from the linear input signal minus the class C signal. This second signal, or drive signal, is preferably added only to the best sub-amplifier sub-branch with the correct amplitude adjusted by an attenuator 161 or 162, and the other sub-branch gets no signal at all, or a very small signal. The best sub-amplifier sub-branch is generally the one with the highest transimpedance from the sub-amplifier to the output 140. In the present case of an output network 130 with transmission lines directly coupled to the output 140, this is generally the sub-amplifier whose transmission line is closest to an odd multiple of a quarter wavelengths at the operating frequency in question. The higher

transimpedance occurs due to the impedance transforming properties of the quarterwave line.

Thanks to the properties of the quarterwave transmission line when working as an impedance transformer, the phase shift over frequency, also called group delay, is higher in this type of operation than for the in-phase combination operation of the same sub-amplifier. This phase shift behaves largely as the phase shift due to a length of transmission line twice that of the transmission line in the output network. An effective solution is therefore to compensate by inserting the second signal at a point further into the sub-amplifier sub-branch. The point may be inserted at the transmission line sub-branch corresponding to that delay, giving effectively a negative relative delay.

In Figure 3a, an exemplifying configuration of the power amplifier 100 is shown. In this example, the first output transmission line 131 has an electrical length of one half wavelength at center frequency and the second output transmission line 132 has an electrical length of one quarter wavelength at center frequency.

Accordingly, the first sub-amplifier 11 1 ,1 is plotted with a dotted line and the second sub-amplifier 112, 2 is plotted with a solid line. See also legend in Figure 3a, which shows that reference numerals 1 and 2 are associated to dotted and solid lines, respectively.

In Figure 3b, the average efficiency with assumed ideal class B operation for a 7 dB PAR Rayleigh signal amplitude distribution is shown. As indicated by the solid and dotted lines, the first sub-amplifier 11 1 is operated in intervals at ends of the operational bandwidth and the second sub-amplifier 112 is operated in an

intermediate frequency range of the operational bandwidth.

In Figure 3c, the transition point relative to maximum amplitude is illustrated.

The transition point is in the described implementation adjusted for the frequency of operation by the setting of gate bias voltage to the class C amplifier. It can also be achieved automatically in wide bandwidths by having frequency-dependent gain (e.g. by filters) before and after the class C stage (the second filter equalizing the maximum amplitude across the bandwidth).

In Figure 3d, the phase of the second signal relative to the in-phase combining phase, i.e. the first signal, of the same sub-amplifier is illustrated. The frequency regions are coded with dotted and solid lines according to which sub- amplifier is driven, i.e. operated or active, in the low amplitude range, e.g. the first amplitude range. An ideal relative phase of the second signal, illustrated by the dotted and solid lines, differs somewhat from that achieved by the simple

approximation by insertion point as described above, shown as thin black lines. In this example, the difference is less than 0.25 radians for the worst frequency in the central frequency region, which should affect efficiency only marginally. In examples, where the insertion point is not used, i.e. the second signal is inserted at the same point as the first signal, would give zero radians offset, which in turn would give a maximum deviation of about one radian from the ideal phase. Hence, the insertion point provides an improvement.

To reduce the phase difference between the first and second signals, a 90 degree phase shift and a quarter wavelength offset insertion point for the middle part of the range may be used. It is also possible to have a filter with a response that tracks the phase shift over frequency. Adding phase adjustment means for removing the residual phase difference is yet another possibility.

The combination of the first and second signal may be applied for infinite bandwidths by using resistive combiners. This solution is preferred for low power level. If loss of the resistive combiners is considered to be too high, for example if the combination of the first and second signal is applied at higher power levels, a Wilkinson combiner may be used.

Figure 4 depicts an embodiment of the power amplifier 100. As mentioned above, the same or similar reference numerals have been just for the same or similar features. For simplicity, repetition of already mentioned features is not provided here. In this embodiment, the power amplifier 100 is extended by adding a stage, i.e. by adding a sub-branch (lowest sub-branch in Figure 4) including a third sub-amplifier 113. Here, the modified output network 130 comprises only of transmission lines, including a third output transmission line 133, directly coupled to the output port 140.

The third output transmission line 132 is connected to the third sub- amplifier 1 13. The input network 120 may comprise a third input transmission line 123, connected to the second input branch 102 (not shown in Figure 4). Electrical length of the third input transmission line 123 may be twice as long as electrical length of the third output transmission line 133. Furthermore, the power amplifier 100 of Figure 4 includes a fourth attenuator 164. One attenuator from among the first, second and fourth attenuator 161 , 162, 164 is activated at a time. Thus, the other attenuators are blocking when one is active.

Moreover, Figure 4 illustrates combiners 181 , 182, 183 to be used for insertion of the second signal into respective sub-branches at respective amplitude ranges.

Each of the respective sub-branches may comprise intermediate

transmission lines 171 , 172, 173, each of which are connected to a respective combiner and a respective sub-amplifier. The electrical length of the intermediate transmission lines 171 , 172, 173 may be calculated as a dummy length X (same for all) minus the lengths of the line from the amplifier to the output and the input transmission line lengths. The dummy length X is chosen so that no intermediate transmission line gets a negative length.

Accordingly, Figure 4 shows a 3-stage power amplifier with a 9: 1 bandwidth of high efficiency. The electrical lengths of the first, second and third output transmission lines 131 , 132, 133 in the output network have electrical lengths of a whole, a half and a quarter wavelength at center frequency, respectively.

Since the first sub-amplifier 11 1 is coupled to the output port 140 by the first output transmission line 133 that is one wavelength at the center frequency, the addition of the second signal in its sub-branch is at a point 2 wavelengths at center frequency from the common split point of the first signal. This means that the first input transmission line 121 has an electrical length of 2 wavelengths at center frequency in this example. The maximum deviation from the ideal phase is then a little more than twice as much as in the two-stage example of Figure 1. In this case, it may be beneficial to add means for phase adjustment or phase tracking filters as outlined above. In Figure 5a, an exemplifying configuration of the power amplifier 100 of

Figure 4 is shown. In this example, the first output transmission line 131 has an electrical length of one wavelength at center frequency, the second output transmission line 132 has an electrical length of one half wavelength at center frequency, and the third output transmission line 133 has an electrical length of one quarter wavelength at center frequency. Accordingly, the first sub-amplifier 11 1 is plotted with a dotted line, the second sub-amplifier 1 12 is plotted with a solid line and the third sub-amplifier 113 is plotted with a dashed line. In Figure 5b, the average efficiency with assumed ideal class B operation for a 7 dB PAR Rayleigh signal amplitude distribution is shown. As indicated by the solid, dashed and dotted lines, the first sub-amplifier 1 11 is operated in four intervals (dotted lines), the second sub-amplifier 112 is operated in two intervals (solid lines), and the third sub-amplifier 1 13 is operated in only one interval (dashed line) at an intermediate frequency of the operational bandwidth.

In Figure 5c, the transition point relative to maximum amplitude is illustrated similarly as in Figure 3c.

In Figure 5d, the phase of the second signal relative to the in-phase combining phase, i.e. the first signal, of the same sub-amplifier is illustrated. The frequency regions are coded as above according to which sub-amplifier is driven in the first amplitude range.

If the sub-amplifiers are not all coupled directly to the output port 140 as in Figure 4 and 5a, the deviations from the phase compensations are generally greater.

Figure 6a shows an exemplifying configuration of the power amplifier 100 in which the second and third sub-amplifier are connected to a common output transmission line of length 1 ,25 wavelengths at center frequency. The second and third sub-amplifiers are also connected to respective transmission lines of length one half wavelength and one quarter wavelength, respectively.

In Figure 6b, the average efficiency with assumed ideal class B operation for a 7 dB PAR Rayleigh signal amplitude distribution is shown. As indicated by the solid, dashed and dotted lines, the first sub-amplifier 1 11 is operated in three intervals (dotted lines), the second sub-amplifier 112 is operated in two intervals (solid lines), and the third sub-amplifier 113 is operated in two intervals (dashed lines).

In Figure 56, the transition point relative to maximum amplitude is illustrated similarly as in Figure 3c and 5c.

As illustrated in Figure 6d, the deviations are greater than for the cases illustrated in Figure 3d and 5d. The solid thin black lines represent the phase shift over frequency, based in this case on the lengths to the nearest common junction.

The first sub-amplifier 11 1 in the three quarter wavelength line sub-branch closely follows the think black line (distinguishable in the Figure), since it is coupled directly to the output. The other two sub-amplifiers 1 12, 1 13 are more off, and possibly needs one of the aforementioned phase compensations. In Figure 7, further embodiments are illustrated. One embodiment variant is to use only some of the frequency intervals illustrated in Figures 3b, 5b and 6b. One example is a single or dual-band implementation. In those cases not all of the attenuators and combiners of the full implementation are needed. The input side, just after the insertion of the first signal but before the insertion point, may also sometimes be made shorter. This is because only the transmission line lengths that precede the insertion point for the actually used intervals are needed. If the intervals associated with the longest transmission line on the output side, i.e. after the insertion point, are not used, i.e. the one wavelength line in this example, the line preceding its insertion point may be disregarded.

The class C function, or signal, may be implemented directly in the sub- amplifiers 1 11 , 1 12 and/or 1 13 for sub-branches that do not need the second signal. In such cases, the amplified input signal is fed directly to the sub-amplifiers 1 11 , 1 12 and/or 1 13 without prior class C processing. Combinations of final stage class C biasing and class C pre-processing are also feasible.

Another variant that is to have a common transmission line that is split into lines 121 ,122, 123 at a point within the length of the shortest of these, i.e. in this case 122. An advantage may be that the splitter may be made electrically long without adding extra length, e.g. Wilkinson variants.

In the exemplifying power amplifier 100 of Figure 7, a fifth attenuator 165 may be connected to the second input transmission line 122 and the second sub- amplifier 1 12. The fifth attenuator 165 may be set to match insertion loss of the combiner 181 in the first sub-branch 105.

In the exemplifying power amplifier 100 of Figure 7, a sixth attenuator 166 may be connected to the third input transmission line 123 and the third sub-amplifier 1 13. The sixth attenuator 166 may also be set to match insertion loss of the combiner 181 in the first sub-branch 105.

In some cases, the transition point amplitude is fairly consistent over the design frequency band or bands. The gate bias setting and attenuator setting, etc. can then be fixed in production, by a priori design or by trimming.

Figure 8 shows an exemplifying radio network node 200.

As used herein, the term "radio network node" may refer to is a piece of equipment that facilitates wireless communication between user equipment (UE) and a network. Accordingly, the term "radio network node" may refer to a Base Station (BS), a Base Transceiver Station (BTS), a Radio Base Station (RBS), a NodeB in so called Third Generation (3G) networks, evolved Node B, eNodeB or eNB in Long Term Evolution (LTE) networks, or the like. In UMTS Terrestrial Radio Access Network (UTRAN) networks, where UTMS is short for Universal Mobile

Telecommunications System, the term "radio network node" may also refer to a Radio Network Controller. Furthermore, in Global System for Mobile Communications (GSM) EDGE Radio Access Network (GERAN), where EDGE is short for Enhanced Data rates for GSM Evolution, the term "radio network node" may also refer to a Base Station Controller (BSC).

The radio network node 200 comprises a power amplifier 210, e.g. the power amplifier 100, according to the embodiments described above.

Furthermore, the radio network node 200 may comprise a processing circuit 220 and/or a memory 230.

The radio network node 200 may further comprise additional transceiver circuitry (not shown) for facilitating transmission and reception of data, e.g. in the form of radio signals.

Figure 9 shows an exemplifying user equipment 300.

As used herein, the term "user equipment" may refer to a mobile phone, a cellular phone, a Personal Digital Assistant (PDA) equipped with radio

communication capabilities, a smartphone, a laptop or personal computer (PC) equipped with an internal or external mobile broadband modem, a tablet PC with radio communication capabilities, a portable electronic radio communication device, a sensor device equipped with radio communication capabilities or the like. The sensor may be any kind of weather sensor, such as wind, temperature, air pressure, humidity etc. As further examples, the sensor may be a light sensor, an electronic switch, a microphone, a loudspeaker, a camera sensor etc.

The user equipment 300 comprises a power amplifier 310, e.g. the power amplifier 100, according to the embodiments described above.

Furthermore, the user equipment 300 may comprise a processing circuit 320 and/or a memory 330. The means of the terms "processing circuit" and

"memory" as explained above applies also for the user equipment 300.

The user equipment 300 may further comprise additional transceiver circuitry (not shown) for facilitating transmission and reception of data, e.g. in the form of radio signals.

As used herein, the term "processing circuit" may be a processing unit, a processor, an application specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or the like. As an example, a processor, an ASIC, an FPGA or the like may comprise one or more processor kernels. In some examples, the processing circuit may be embodied by a software or hardware module. Any such module may be a determining means, estimating means, capturing means, associating means, comparing means, identification means, selecting means, receiving means, transmitting means or the like as disclosed herein. As an example, the expression "means" may be a unit, such as a determining unit, selecting unit, etc.

As used herein, the term "memory" may refer to a hard disk, a magnetic storage medium, a portable computer diskette or disc, flash memory, random access memory (RAM) or the like. Furthermore, the term "memory" may refer to an internal register memory of a processor or the like.

As used herein, the terms "number", "value" may be any kind of digit, such as binary, real, imaginary or rational number or the like. Moreover, "number", "value" may be one or more characters, such as a letter or a string of letters, "number", "value" may also be represented by a bit string.

As used herein, the expression "in some embodiments" has been used to indicate that the features of the embodiment described may be combined with any other embodiment disclosed herein.

Even though embodiments of the various aspects have been described, many different alterations, modifications and the like thereof will become apparent for those skilled in the art. The described embodiments are therefore not intended to limit the scope of the present disclosure.

Claims

1. A power amplifier (100) comprising a first and a second sub-amplifier (1 11 , 1 12) for amplification of an input signal into an output signal, wherein the first and second sub-amplifiers (1 11 , 112) are connected to an input network (120) for receiving the input signal at an input port (150) of the input network (120), and the first and second sub-amplifiers (11 1 , 112) are connected to an output network (130) for providing the output signal at an output port (140) of the output network (130), wherein the power amplifier (100) is characterized in that the input network (120) comprises:

a first input branch (101) configured to provide a first signal, derived from the input signal, wherein an average rate of amplitude change of the first signal versus amplitude change of the input signal in a first amplitude range is less than an average rate of amplitude change of the first signal versus amplitude change of the input signal in a second amplitude range,

a second input branch (102) configured to provide a second signal, derived from the input signal, wherein the second signal has a triangularly shaped amplitude envelope which increases in the first amplitude range and which decreases in the second amplitude range, wherein the power amplifier (100) is configured to:

provide the first signal to the first and second sub-amplifiers (11 1 , 1 12), and

provide the second signal to the first sub-amplifier (1 11).

2. The power amplifier (100) according to claim 1 , wherein the power amplifier (100) comprises means configured to enable a concurrent super-positioning of the first and second signals with respect to phase and group delay.

3. The power amplifier (100) according to claim 2, wherein the means is configured to provide the first signal to the first and second sub-amplifiers (1 11 , 1 12) via a first sub-branch (105) and a second sub-branch (106), respectively, wherein the power amplifier (100) is configured to insert the second signal into the first sub- branch (105) at an insertion point of the first sub-branch (105), wherein phase and group delay of the triangularly shaped amplitude envelope due to the insertion point enable the concurrent super-positioning of the first and second signals with respect to phase and group delay.

4. The power amplifier (100) according to claim 3, wherein the output network (130) comprises a first output transmission line (131), connected to the first sub- amplifier (1 11), and a second output transmission line (132), connected to the second sub-amplifier (1 12), wherein the input network (120) comprises a first input transmission line (121), connected to the second input branch (102), wherein electrical length of the first input transmission line (121) is twice as long as electrical length of the first output transmission line (131), wherein the insertion point is located at an output (123) of the first input transmission line (121) .

5. The power amplifier (100) according to the preceding claim, wherein the input network (120) comprises a second input transmission line (122), connected to the second input branch (102), wherein electrical length of the second input transmission line (122) is twice as long as electrical length of the second output transmission line (132), wherein a further insertion point at the second sub- branch (106) is located at an output (124) of the second input transmission line (122), wherein phase and group delay of the triangularly shaped amplitude envelope due to the further insertion point enable concurrent super-positioning of the first and second signals with respect to phase and group delay.

6. The power amplifier (100) according to any one of claims 1-5, wherein the means comprises a phase splitting network () configured to enable the concurrent super- positioning of the first and second signals with respect to phase and group delay.

7. The power amplifier (100) according to claim 6, wherein the phase splitting

network is configured to provide the first and second signals at a common insertion point at which electrical length to the first and second sub-amplifiers (1 1 1 , 1 12) is equal.

8. The power amplifier (100) according to any one of claim 1-7, wherein the means comprises a constant phase difference network configured to enable the concurrent super-positioning of the first and second signals with respect to phase and group delay.

9. The power amplifier (100) according to any one of the preceding claims, wherein the power amplifier (100) is configured to provide the second signal to the first sub-amplifier (1 11) for frequencies in a first frequency range, and to provide the second signal to the second sub-amplifier (1 12) for frequencies in a second frequency range, wherein the first frequency range is different from the second frequency range.

10. The power amplifier (100) according to any one of the preceding claims, wherein the power amplifier (100) has a transition interval whose length between the first and second amplitude ranges is configurable, whereby the first and/or second amplitude range is configurable.

1 1. The power amplifier (100) according to any one of the preceding claims, wherein the input network (120) comprises a first attenuator (161), connected to the second input branch (102) and the first sub-amplifier (1 11), and a second attenuator (162), connected to the second input branch (102) and the second sub-amplifier (1 12), wherein the power amplifier (100) is configured to activate the first attenuator (161) when the second signal is in the first frequency range, and wherein the power amplifier (100) is configured to activate the second attenuator (162) when the second signal is in the second frequency range.

12. The power amplifier (100) according to any one of claims 1-11 , wherein the input network (120) comprises a third attenuator (163), connected to an output (104) of the first input branch (101) and an input (103) of the second input branch (102), whereby the second input branch (102) is configured to provide the second signal.

13. The power amplifier (100) according to any one of claims 1-12, wherein the

power amplifier (100) is a composite power amplifier.

14. A radio network node (200) comprising the power amplifier (100) according to any one of claims 1-13.

15. A user equipment (300) comprising the power amplifier (100) according to any one of claims 1-13.