WO2014094824A1

WO2014094824A1 - Amplifier circuit and method

Info

Publication number: WO2014094824A1
Application number: PCT/EP2012/075934
Authority: WO
Inventors: Richard Hellberg
Original assignee: Telefonaktiebolaget LM Ericsson AB
Current assignee: Telefonaktiebolaget LM Ericsson AB
Priority date: 2012-12-18
Filing date: 2012-12-18
Publication date: 2014-06-26
Anticipated expiration: 2015-06-18

Abstract

An amplifier circuit (22; 42) comprises afirst amplifier(27) for amplifying an input signal (21), and a second amplifier (29) for amplifying the input signal (21). The output of the first amplifier(27) and an output of the second amplifier (29) are coupled to a common output(23). At leastoneofthe first amplifier (27) or the second amplifier (29) is coupled to receive a supply voltage (31; 33) which, during use, is controlled as a function of the frequency of the input signal (21).

Description

Amplifier circuit and method

Technical field

The present invention relates generally to an amplifier circuit and method, and in particular to improving efficiency in an amplifier circuit and method.

Background

In communications systems, amplifiers need to be able to amplify radio frequencies spread across a fairly wide bandwidth. In addition, they need to be able to do this efficiently in order to reduce power consumption and the need for cooling.

A conventional power amplifier (for example a class B, AB or F amplifier) has a fixed radio frequency (RF) load resistance and a fixed voltage supply. The bias in class B or AB amplifiers 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) current (and hence DC power) is therefore largely proportional to the RF output current amplitude (and voltage). The output power, however, is proportional to the RF output current squared. The efficiency, i.e. output power divided by DC power, is therefore also proportional to the output amplitude. The average efficiency of a power amplifier is consequently low when amplifying signals that on average have a low output amplitude (or power) compared to the maximum required output amplitude (or power). Power amplifiers configured to operate in a Doherty mode or a Chireix mode of operation are much more efficient than conventional amplifiers for amplitude- modulated signals that have a high peak-to-average ratio (PAR), since they have a much lower average sum of output currents from the amplifier

transistors. It will be appreciated that such a reduced average output current leads to high average efficiency. The reduced average output current is obtained by using two amplifier transistors that influence the output voltages and currents of each other through a reactive output network, the reactive output network also being coupled to the load. By driving the constituent amplifier transistors with suitable amplitudes and phases, the sum of RF output currents is reduced at all levels except the maximum. Also, for these amplifiers the RF voltage at one or both transistor outputs is increased. In 2001 the inventor of the present application invented an amplifier system called "Unified High-Efficiency Amplifiers", published as EP1470635. This discloses a 2-stage high-efficiency amplifier with increased robustness against circuit variations, that can avoid tuning of the output network, and with radically increased bandwidth of high efficiency.

The amplifier consists in having a longer and a shorter transmission line from two amplifier transistors to a common output (which is coupled to a load). If the most wideband operation is desired, the lengths of the transmission lines are chosen such that the longer line has an electrical length of half a wavelength at center frequency, while the shorter line is a quarter wavelength long at center frequency. The basic structure of such an amplifier is shown in Figure 1 .

The amplifier circuit 10 of Figure 1 comprises a first amplifier 27 located in a first or "main" branch 1 1 of the amplifier circuit 10 and a second amplifier 29 located in a second or "auxiliary" branch 13 of the amplifier circuit 10. An output of the first amplifier 27 and an output of the second amplifier 29 are coupled to a common output 23 via respective first and second transmission lines 26 and 28. As mentioned above, the first and second transmission lines 26, 28 form a reactive output network which influences the operation of the first and second amplifiers 27, 29. The electrical length of the transmission line 26 can be designed to be shorter than the electrical length of the transmission line 28 (for example a quarter wavelength and a half wavelength, respectively, at center frequency). A fixed supply voltage 25 is applied to the first amplifier 27 and the second amplifier 29. In operation, an input signal 21 is received by the amplifier circuit 10, split by a signal component separator 22 and amplified by the first amplifier 27 and the second amplifier 29.

The amplifier circuit 10 has a wide bandwidth of high efficiency since the shorter/longer transmission lines 26, 28 of the output network form different kinds of amplifiers at different frequencies. Around a center frequency of operation the amplifier circuit 10 operates as a Doherty amplifier, and at 2/3 and 4/3 of that frequency the amplifier circuit 10 operates as a Chireix amplifier. The wide (about 3 to 1 ) high efficiency bandwidth is thus achieved in such an amplifier circuit 10 by devising an output network that has both suitable impedance transformation characteristics and full power output capacity over a wide bandwidth, together with a unified control system that allows high efficiency operation at all "modes" across that bandwidth. The amplifier circuit 10 of Figure 1 therefore allows operation in between and outside the intrinsically narrowband Doherty and Chireix modes.

The amplitude probability density of a mix of sufficiently many independent RF signals, or of a multi-user CDMA (Code Division Multiple Access) signal, tends to be close to a Rayleigh distribution having a high peak-to-average ratio (PAR) for instance. For a typical wideband multicarrier or multiuser signal with about 7 dB PAR, good efficiency, close to the efficiency of the Doherty mode, can be achieved everywhere across a 100% relative bandwidth. However, above a PAR of about 7dB the efficiency deteriorates.

With signals of higher PAR values it is only possible to make the prior art solutions have high efficiency for narrow bandwidths. This is because when operating the amplifier at higher PAR, the efficiency is lowered for both the Doherty mode and the Chireix modes. Known solutions to increase the efficiency of the Doherty mode, for example typically rescaling the characteristic impedance of the transmission lines, at the same time has the effect of decreasing the efficiency of the Chireix modes drastically, leading to narrow bandwidth of high efficiency. This is because high-efficiency Chireix operation at high PAR requires almost equal transmission line impedances and transistor sizes. On the other hand, known solutions that increase the efficiency of one Chireix mode, for example typically changing the lengths of the transmission lines, destroys the Doherty mode and the other Chireix mode, and thus takes away most of the wideband high-efficiency response, leaving only a narrow band of high efficiency.

Summary

It is an aim of the embodiments of the invention to obviate or reduce at least some of the above disadvantages and to provide an improved amplifier circuit and method.

According to a first aspect of the invention there is provided an amplifier circuit comprising a first amplifier for amplifying an input signal and a second amplifier for amplifying the input signal. An output of the first amplifier and an output of the second amplifier are coupled to a common output. At least one of the first and second amplifiers is coupled to receive a supply voltage which, during use, is controlled as a function of the frequency of the input signal.

By controlling the supply voltage of at least one of the first and second amplifiers as a function of the frequency of the input signal, this has the advantage of providing an amplifier circuit which is efficient over a wide frequency range and with high PAR signals, both at Doherty and Chireix modes of operation, for example.

According to a another aspect of the present invention there is provided a method for amplifying an input signal. The method comprises the steps of amplifying the input signal at a first amplifier, and amplifying the input signal at a second amplifier, wherein an output of the first amplifier and an output of the second amplifier are coupled to a common output. A supply voltage to at least one of the first amplifier or the second amplifier is controlled as a function of the frequency of the input signal.

Brief description of the drawings

For a better understanding of the present invention, and to show more clearly how it may be carried into effect, reference will now be made, by way of example only, to the following drawings in which:

Figure 1 shows an amplifier circuit according to the prior art; Figure 2 shows an amplifier circuit according to an embodiment of the present invention;

Figure 3 shows a flowchart illustrating the steps performed by an embodiment of the present invention;

Figures 4a to 4f compare RF current amplitude waveforms and RF voltage waveforms describing the effects of trying to improve the efficiency of amplifiers according to embodiments of the present invention; Figure 5 shows an amplifier circuit according to another embodiment of the present invention;

Figure 6a shows how first and/or second supply voltages may be controlled to first and/or second amplifiers according to an embodiment of the present invention;

Figure 6b shows the efficiency in class B mode for an amplifier controlled according to Figure 6a;

Figure 6c shows the maximum RF currents for an amplifier controlled according to Figure 6a;

Figure 7 compares the efficiency curve of an amplifier according to the embodiment of Figures 6a to 6c with the efficiency curves of known amplifiers;

Figure 8a shows how first and/or second supply voltages may be controlled to first and/or second amplifiers according to another embodiment of the present invention;

Figure 8b shows the efficiency in class B mode for an amplifier controlled according to Figure 8a;

Figure 8c shows the maximum RF currents for an amplifier controlled according to Figure 8a;

Figure 9a shows how first and/or second supply voltages may be controlled to first and/or second amplifiers according to another embodiment of the present invention; Figure 9b shows the efficiency in class B mode for an amplifier controlled according to Figure 9a;

Figure 9c shows the maximum RF currents for an amplifier controlled according to Figure 9a;

Figure 10 compares the efficiency curve of an amplifier according to the embodiment of Figures 9a to 9c with the efficiency curves of known amplifiers; Figure 1 1 compares efficiency curves for amplifiers according to the prior art and embodiments of the invention;

Figure 12 shows a set of efficiency curves sampled evenly within one half of the 100 % relative bandwidth for embodiments of the present invention; and

Figure 13 shows efficiency curves for an amplifier known in the prior art and an amplifier according to another embodiment of the invention.

Detailed description

In the following description the same reference designations will be used for the same or similar elements throughout the figures of the drawings.

Figure 2 shows an amplifier circuit 22 according to an embodiment of the present invention. The amplifier 22 comprises a first amplifier 27 for amplifying an input signal 21 , and a second amplifier 29 for amplifying the input signal 21 . An output of the first amplifier 27 and an output 26 of the second amplifier 29 are coupled to a common output 23. According to this embodiment at least one of the first or second amplifiers 27, 29 is coupled to receive a supply voltage (Vvariabie) 31 , 33 which, during operation, is controlled as a function of the frequency of the input signal . By controlling the supply voltage of at least one of the amplifiers, based on the frequency of the input signal, this has the advantage of making the amplifier more efficient for wide bandwidth and high PAR signals, as will be described in greater detail below.

A control module (not shown) may be provided for determining the frequency of the input signal 21 , and controlling the first supply voltage 31 or the second supply voltage 33, or both, as a function of the frequency of the input signal 21 . The first supply voltage 31 and/or the second supply voltage 33, during operation of the amplifier 22, therefore vary as a function of the frequency of the input signal 21 detected by such a control module. According to one

embodiment the variable supply voltages for the first amplifier 27 and/or the second amplifier 29 can be controlled by determining a frequency band of operation and predicting an expected frequency content, and setting the variable supply voltages 31 , 33 accordingly for the first amplifier 27 and the second amplifier 29 respectively. In such an embodiment the variable supply voltages can be set according to a frequency content which is fairly static.

According to another embodiment the variable supply voltages 31 , 33 for the first amplifier 27 and/or the second amplifier 29 can be controlled on-the-fly during operation of the amplifier circuit 22, whereby the frequency of the input signal is determined on-the-fly, and the variable supply voltage to the first and/or second amplifiers 27, 29 controlled in a dynamic manner. Figure 3 shows a flowchart illustrating the steps performed by an embodiment of the present invention. The steps shown in the flowchart can for instance be performed by the amplifier circuit 22 described in relation to Figure 2 in a method for amplifying the input signal 21 . The method comprises amplifying the input signal 21 at a first amplifier 27, step 301 . The method further comprises amplifying the input signal 21 at a second amplifier 29, step 302. As described in relation to Figure 2, the output of the first amplifier 27 and the output of the second amplifier 29 are coupled to a common output 23. The method further comprises controlling the supply voltage to at least one of the first amplifier 27 or the second amplifier 29 as a function of the frequency of the input signal 21 , step 303.

It is noted that the embodiments of Figures 2 and 3 can be used with

transmission lines (i.e. that couple the outputs of the amplifiers to the common output) having equal impedances, or with transmission lines having unequal electrical lengths or unequal impedances, as will become apparent from the embodiments described later in the application.

It should be noted that during operation the supply voltages which, as explained above, vary and are controlled in relation to the frequency of the input signal 21 , can also be supplied to both the first amplifier 27 and the second amplifier 29.

The variable supply voltages to the first amplifier 27 and the second amplifier 29 may be the same or different. Any variations to the supply voltage may be varied simultaneously, or the period during which the first supply voltage is varied may be before or after the period during which the second supply voltage is varied, or may partially overlap.

Various embodiments will be described below with reference to Figures 4 to 13. In the embodiments described below it will be seen that the present invention can be used with arrangements whereby the transmission lines connecting the respective outputs of the first amplifier 27 and second amplifier 29 to a common output 23 have arrangements in which one transmission line is arranged to have a shorter electrical length than the other, or arrangements in which one or both of the supply voltages are varied, or any combination of the above. According to a first example, it is assumed that the transmission lines 26, 28 coupling the respective outputs of the first amplifier 27 and the second amplifier 29 to the common output 23 of Figure 2 are transmission lines having substantially equal, or comparable line impedances. In this first example the supply voltage to the first amplifier 27 and the supply voltage to the second amplifier 29 are both lowered compared to a nominal supply voltage. Such an arrangement is suitable for an application in which the input signal comprises, for example, a 10 dB PAR Rayleigh distributed amplitude signal. The Chireix operation at around 2/3 and 4/3 of center frequency is adjusted for high efficiency with the higher PAR signal by coupling the lower supply voltage to both the first amplifier 27 and second amplifier 29, i.e. a lowered supply voltage compared to a prior art solution whereby both transistors receive a fixed maximum supply voltage.

Since the line lengths or impedances of the transmission lines 26, 28 are not changed, the symmetry of the frequency response around center frequency is maintained. The effect of this change is that the efficiency at low output amplitudes is increased, and that the maximum RF current amplitudes are increased to compensate.

Figures 4a to 4f show the effect of lowering the supply voltage to the first and second amplifiers 27, 29 for different arrangements. The x-axis represents the output voltage amplitude, and is normalized in the sense that full output amplitude is at the x-axis amplitude 1 . The current amplitudes are normalised so that they are in the unmodified prior art sum to 1 . The RF voltage amplitudes are normalized so that 1 represents the full voltage swing with unmodified supply voltage (the supply voltage sets the limit for RF voltage swing, so in the simplest amplifier model the maximum RF voltage swing is equal to the supply voltage). Figures 4a and 4b show voltage and current amplitudes for an arrangement in which the first and second amplifiers 27, 29 are coupled to a common output via equal or comparable transmission lines, and with conventional supply voltages to the first and second amplifiers (i.e. with fixed supply voltages). Figures 4c and 4d show voltage and current amplitudes for an arrangement in which the first and second amplifiers 27, 29 have conventional supply voltages as above, and are coupled to a common output via transmission lines having unequal impedances, i.e. optimized for Doherty operation. With such an arrangement it can be seen that the Doherty mode at centre frequency is improved, but to the detriment of the Chireix modes. Figures 4e and 4f show voltage and current amplitudes for an arrangement in which the first and second amplifiers 27, 29 are coupled to a common output via transmission lines having equal

impedances, and whereby the supply voltages to the first and/or second amplifiers are varied as a function of the frequency of the input signal.

According to one embodiment the supply voltage to the first amplifier 27 is controlled such that the supply voltage is lowered to have a minimal value for a centre frequency of the input signal, the centre frequency corresponding to a Doherty mode of operation, and increased from the minimal value as the frequency of the input signal 21 moves above or below the centre frequency.

According to another embodiment, the supply voltage to the second amplifier 29 is controlled such that the supply voltage has a maximum value for a centre frequency of the input signal 21 , the centre frequency corresponding to a Doherty mode of operation, and decreases from the maximum value as the frequency of the input signal moves above or below the centre frequency.

It is noted that the control of the supply voltage to the first amplifier 27 and second amplifier 29, as outlined in the paragraphs above, can be provided alone or in combination. In such embodiments the centre frequency

corresponds to a Doherty mode of operation of the amplifier circuit, and wherein operation in outer regions as the frequency moves above or below the centre frequency corresponds to a Chireix mode of operation.

Figure 5 shows an amplifier circuit 42 according to an another embodiment of the present invention. The first amplifier 27 is coupled to the common output 23 via a first transmission line 26 having a first electrical length λ/4, for example, and the second amplifier 29 is coupled to the common output 23 via a second transmission line having a second electrical length λ/2, for example. As such it can be seen that the electrical length of the first transmission line 26 is less than the electrical length of the second transmission line 28. As an example, the difference between the electrical length of the first transmission line 26 and the electrical length of the second transmission line 28 may be k ^* (λ/4). In this embodiment the wideband high-efficiency response is achieved with

transmission lines that have electrical length at center frequency of a quarter wavelength and a half wavelength, respectively. It is noted that the physical length is substantially constant (measured in mm), but that the electrical length changes (substantially linearly) with frequency. The electrical lengths at the lower Chireix mode is therefore a sixth and a third of a wavelength at that frequency (which is 2/3 of the center frequency).

Figures 6a, 6b and 6c describe how the supply voltages may be varied according to an embodiment of the invention, and the resulting efficiency in class B mode, and the resulting maximum RF currents. The parameters shown in Figures 6a to 6c are those for a Rayleigh distributed signal having a PAR equal to about 10 dB.

The x-axis represents the frequency of the input signal, with "1 " representing a nominal centre frequency of operation, which may correspond to a frequency F_D of a Doherty mode of operation. A frequency F_Ci at 2/3 of the centre frequency F_D and a frequency F_C2 at 4/3 of the centre frequency F_D correspond to the Chireix modes of operation. The y-axis 82 in Figure 6a represents the supply voltage provided to first and second amplifiers of an amplifier circuit (such as the amplifier circuit 42 shown in Figure 5). The curve labelled 97 represents the supply voltage provided to the first amplifier 27. The curve labelled 99 represents the supply voltage provided to the second amplifier 29. According to this embodiment it is assumed that the first amplifier 27 is coupled to the common output via a transmission line 26 having an impedance of λ/4, and the second amplifier 29 coupled to the common output via a transmission line having an impedance λ/2. The effective electrical length of the first transmission line 26 is therefore shorter that the effective electrical length of the second transmission line 28 in this embodiment. As can be seen from Figure 6a, the supply voltage 97 to the first amplifier 27 is controlled such that it has a minimal value around centre frequency

(corresponding to a Doherty mode of operation at frequency F_D). For values of the frequency different from the centre frequency of the input signal, i.e. values lower than the centre frequency or higher than the centre frequency, the supply voltage 97 to the first amplifier 27 has a value higher than this minimal value. In other words, the supply voltage to the first amplifier 27 (coupled to the shorter transmission line 26) is low around centre frequency and high in the uppermost and lowermost frequency ranges. According to the example in Figure 6a the functional relationship between the supply voltage 97 of the first amplifier 27 and the frequency of the input signal, as it varies across the bandwidth, is substantially parabolic. Other functional relationships may also be used, that meet the general criteria of being low near the centre frequency and high at the lowermost and uppermost frequencies. As can also be seen from Figure 6a, the supply voltage 99 to the second amplifier 29 has a substantially maximum value around the centre frequency of operation. For values of the frequency different from the centre frequency of the input signal, i.e. values lower than the centre frequency or higher than the centre frequency, the supply voltage 99 to the second amplifier 29 has a value lower than this maximum value. In other words, the supply voltage to the second amplifier 29 (coupled to the longer transmission line 28) is high around centre frequency and low in the uppermost and lowermost frequency ranges. The functional relationship between the supply voltage 99 of the second amplifier 29 and the frequency of the input signal can also be substantially parabolic in nature. Other functional relationships may also be used, that meet the general criteria of being high near the centre frequency and low at the lowermost and uppermost frequencies.

In the embodiment shown in Figure 6a both supply voltages to the first amplifier 27 and second amplifier 29 are shown as being controlled as a function of frequency. It is noted, however, that the invention is intended to embrace just one of the supply voltages being varied or controlled. Furthermore, the supply voltages to the first amplifier and second amplifier can be controlled in the manner shown in Figure 6a for transmission lines that have comparable impedance or electrical length. With regard to such an embodiment, however, it is noted that in the case of having electrical lengths which are substantially equal, there is lower efficiency at low amplitudes. In the case of electrical lengths that are substantially equal around a quarter wavelength (λ/4 - Δ) or (λ/4 + Δ), the embodiments function by improving the operation outside of the central Chireix mode (where the operation tends to be Doherty-like) by making the transistor sizes appear unequal (which favours backed-off Doherty operation). The voltages may be controlled in such an arrangement to be generally equal at the center frequency, with one voltage being higher and the other lower to the sides of the center frequency (high-low on one side and low- high on the other). It is noted that the manner in which the voltages are changed or controlled with frequency can be different for different amplifier types and frequency ranges.

Returning to the embodiment of Figure 6a (i.e. in which the electrical lengths of the transmission lines are different), in addition to controlling the supply voltage 97 of the first amplifier 27 to be low near the centre frequency, and/or the supply voltage 99 of the second amplifier 29 to be high near the centre frequency, according to one embodiment the supply voltages are further controlled such that the supply voltage provided to the first amplifier 27 is substantially equal to the supply voltage provided to the second amplifier 29, but lower than a nominal supply voltage, at first and second side lobes of the centre frequency of the input signal. The frequencies of the first and second side lobes may correspond to the respective frequencies of a Chireix mode of operation, that is the frequencies F_Ci and F_C2- The frequencies may also correspond, respectively, to about 2/3 or 4/3 of the centre frequency of the input signal.

It can be seen from Figure 6a that the supply voltages are of comparable size or substantially equal at these points. This has the advantage that the first amplifier 27 and the second amplifier 29 can be of comparable size, and can have the same voltage rating.

The optimal supply voltages 97, 99 for the first and/or second amplifiers 27, 29 can be determined using experiments or simulations to determine what supply voltage should be provided to each of the first and/or second amplifiers 27, 29 during different frequency modes of operation. Alternatively, the optimal supply voltage for the first amplifier and/or second amplifier can be adapted

dynamically, or on-the-fly, during use of the amplifier circuit. Figure 6b shows the efficiency in class B mode for an amplifier controlled according to Figure 6a, the efficiency being represented on the y-axis, and frequency on the x-axis. As can be seen, the amplifier circuit is efficient at the centre frequency (or Doherty frequency F_D), and at the Chireix frequencies F_Ci , Fc2 corresponding to 2/3 and 4/3 of the centre frequency. In other words, the efficiencies at these points are higher than what they would be otherwise, for example as shown further in Figure 7 below. As will be appreciated by a person skilled in the art, the question of efficiency depends on whether one is interested in long term power consumption for signals that can randomly fall anywhere within a large bandwidth (for which average efficiency across the bandwidth is most important), or worst case single frequency operation at one frequency within the bandwidth.

Figure 6c shows the maximum RF currents for an amplifier controlled according to Figure 6a. The curve labelled 91 1 represents the RF current of the first amplifier 27, and the curve labelled 913 represents the RF current of the second amplifier 29. Compared to prior art solutions, it is noted that in some

embodiments of the invention the maximum RF currents of the amplifier transistors, and hence DC currents, are generally increased (for example at the center and outermost frequencies for the embodiment of Figure 6). The maximum RF currents are the output currents which can be measured at the outputs of amplifier 27 and 29, as represented by the y-axis in Figure 6c. For example, in some embodiments the output currents of the transistors are increased by about 30%. This may call for larger final and driver transistors within an amplifier (such as having higher power rating, or higher output current at the same rated supply voltage, which may be achieved by having larger channel width (for power transistors measured in mm or even cm)). In some cases it is possible to achieve more current from a transistor by sacrificing some gain. The amplifiers 27 and 29 are considered to contain both final transistors and drivers when assessing such current ratings. Figure 7 compares the efficiency curve for an amplifier according to the embodiment of Figures 6a to 6c with the efficiency curves for amplifiers known in the prior art. The x-axis 91 represents the frequency of the input signal, while the y-axis 103 represents the efficiency η. The curve labelled with reference 105 represents the efficiency for an amplifier circuit having fixed supply voltages provided to both the first and second amplifiers, and with "equal" transmission lines coupling the first and second amplifiers to a common output (and load). The curve labelled with reference 107 represents the efficiency for an amplifier circuit having fixed supply voltages provided to both the first and second amplifiers, and with "unequal" transmission lines coupling the first and second amplifiers to a common output (and load). The curve labelled with reference 109 represents the efficiency for an amplifier according to the invention and as described in relation to Figures 6a to 6c. It can be seen in Figure 7 that the amplifier according to an embodiment of the invention has a higher efficiency for all values in the regions corresponding to the centre frequency (or Doherty frequency F_D) and the Chireix frequencies F_Ci , F_C2- Thus the amplifier according to the invention performs better in these regions than the amplifiers known from the prior art. From Figure 7 it can be seen that the efficiency is always better than for the "equal-size" prior art, shown in curve 105. Modifying the prior art with prior art methods gives the "unequal-size" example shown in curve 107, which can be as efficient as the invention only in a narrow frequency band at center frequency. The efficiency of an embodiment of the invention as shown in curve 109 is therefore better than the wideband prior art shown in curve 105, and has wider bandwidth than the modified prior art shown in curve 107.

The graphs shown in Figures 8a to 8c are similar to those shown in Figures 6a to 6c and represent a further embodiment of the invention. The parameters shown in Figures 8a to 8c are those for a Rayleigh distributed signal having a PAR equal to about 10 dB, and having unequal line impedances. In this embodiment the first amplifier 27 is coupled to the common output 23 via a first transmission line 26 which has an impedance of substantially 1 .56 times the load resistance. The second amplifier 29 is coupled to the common output 23 via a second transmission line 28 which has an impedance of substantially 2.78 times the load resistance. Thus, according to this embodiment the first amplifier 27 and the second amplifier 29 are coupled to the common output 23 via respective first and second transmission lines 26, 28, and wherein an electrical length of the first transmission line 26 is less than an electrical length of the second transmission line 28.

The curve labelled 1 17 represents the supply voltage provided to the first amplifier 27. The curve labelled 1 19 represents the supply voltage provided to the second amplifier 29. As can be seen, the curves representing the supply voltages 1 17, 1 19 comprise a discontinuity at first and second side lobes of the centre frequency of the input signal, or about 2/3 and 4/3 of the centre frequency of the input signal; or at Chireix modes of operation. The first and second supply voltages 1 17, 1 19 are substantially equal at these points.

For the embodiment of Figure 8a, it is noted that some of the over-dimensioning needed to achieve a high efficiency within the whole 100% relative bandwidth can be provided from higher RF voltage capability of transistors in the amplifier circuits.

Figure 8b shows the efficiency in class B mode for an amplifier controlled according to Figure 8a, the efficiency being represented on the y-axis, and frequency on the x-axis. As can be seen, this embodiment provides almost the same minimum local efficiency within the bandwidth, for example as compared with Figure 6b. Figure 8c shows the maximum RF currents for an amplifier controlled according to Figure 8a. The curve labelled 1 1 1 1 represents the RF current of the first amplifier 27, and the curve labelled 1 1 13 represents the RF current of the second amplifier 29.

It is noted that optimal efficiency can be achieved by lowering the supply voltage of the amplifier connected to the "shorter" transmission line near the centre frequency, with the amplifier connected to the "longer" transmission line being at its maximum value near the centre frequency.

The graphs shown in Figures 9a to 9c are similar to Figures 8a to 8c, however instead of a PAR of 10 dB the input signal 21 has a PAR of 13 dB. The curve labelled 127 represents the supply voltage provided to the first amplifier 27. The curve labelled 129 represents the supply voltage provided to the second amplifier 29. In this embodiment the first amplifier 27 is coupled to the common output 23 via a first transmission line 26 which has an impedance of two times the load resistance. The second amplifier 29 is coupled to the common output 23 via a second transmission line 28 which has an impedance of two times the load resistance.

As can be seen, the curves representing the supply voltages 127, 129 comprise a discontinuity at first and second side lobes of the centre frequency of the input signal, or about 2/3 and 4/3 of the centre frequency of the input signal; or at Chireix modes of operation. The first and second supply voltages 127, 129 are substantially equal at these points. It can be seen, however, that in this embodiment (having a PAR of 13dB compared with 10dB for Figures 8a to 8c), the supply voltages need to be controlled such that they differ more between different frequency regions. Figure 9b shows the efficiency in class B mode for an amplifier controlled according to Figure 9a, the efficiency being represented on the y-axis, and frequency on the x-axis. Figure 9c shows the maximum RF currents for an amplifier controlled according to Figure 9a. The curve labelled 121 1 represents the RF current of the first amplifier 27, and the curve labelled 1213 represents the RF current of the second amplifier 29. In this example, where the transmission line impedances are equal, the "different-sizing" required to obtain an improved efficiency in back-off is carried out by the supply voltage variations. Around center frequency, the amplifier 27 is the "main" Doherty amplifier and therefore has a rather low supply voltage. Outside the Chireix frequencies, the roles are reversed, which is evident by the sharp cross over at the Chireix frequencies. Figure 10 is similar to Figure 7, but comparing the signal having PAR of 13dB of Figures 9a to 9c with prior art solutions rather than a signal having PAR of 10 dB. As in Figure 7, the x-axis 91 represents the frequency of the input signal, with the y-axis 103 representing the efficiency η. The curve labelled with reference 139 represents the efficiency for an amplifier circuit having fixed supply voltages provided to both the first and second amplifiers, and with transmission lines having equal characteristic impedance coupling the first and second amplifiers to a common output (and load). The curve labelled with reference 131 1 represents the efficiency for an amplifier circuit having fixed supply voltages provided to both the first and second amplifiers, and with transmission lines having unequal characteristic impedance coupling the first and second amplifiers to a common output (and load). The curve labelled with reference 137 represents the efficiency for an amplifier circuit according to the embodiment described in Figures 9a to 9c. It can be seen in Figure 10 that the efficiency and bandwidth improvements are also considerable for this

embodiment of the invention when compared with the prior art. For example, the efficiency is improved for all values in the regions corresponding to the centre frequency (or Doherty frequency F_D) and the Chireix frequencies F_Ci , Fc2- Thus the amplifier according to the invention performs better in these regions that the amplifiers known from the prior art. In addition, by comparing the curve labelled 109 shown in Figure 7 with the curve labelled 137 shown in Figure 10, it can be seen that the efficiency is more constant for a signal with a PAR of 13 dB. Usually, the minimum efficiency within the bandwidth of interest is an important measure, although the average efficiency over a bandwidth can also be an important measure. Thus, although this embodiment is less efficient at a PAR of 13dB compared to 10dB, the amplifier according to this

embodiment of the invention still performs better in such regions than the amplifiers known from the prior art.

Figure 1 1 compares efficiency curves for amplifiers according to the prior art and embodiments of the invention. The x-axis 61 of Figure 1 1 represents output amplitude. The y-axis 81 in Figure 1 1 represent the efficiency η. The curves represent efficiency versus amplitude at the Chireix frequency of the examples of Figures 4a to 4f, and represent equal impedance prior art, unequal impedance prior art, and embodiments of the invention. In particular, the curve labelled with reference 87 represents the efficiency for an amplifier having fixed supply voltages provided to both the first and second amplifiers, and with "equal" transmission lines coupling the first and second amplifiers to a common output (and load). as known in the prior art. The curve labelled with reference 85 also represents the efficiency for an amplifier circuit having fixed supply voltages provided to both the first and second amplifiers, and with "unequal" transmission lines coupling the first and second amplifiers to a common output (and load) as known in the prior art. The curve labelled with reference 89 represents the efficiency for an amplifier according to embodiments of the invention, for example as described in Figure 4. As seen from Figure 1 1 , the efficiency at low output amplitudes for the curve labelled 85, relating to a prior art solution, is lower than the prior art version shown by the curve labelled 87, contrary to what one wants to achieve. The embodiments of the invention, however, show an increased efficiency in this region, as illustrated by curve 89, while (largely inconsequentially) sacrificing a little efficiency at the highest amplitudes.

Figure 12 shows a set of efficiency curves sampled evenly within one half of the 100 % relative bandwidth, comparing the prior art and embodiments of the present invention. The x-axis 74 of Figure 12 represents output amplitude. The y-axis 72 in Figure 12 represent the efficiency η.

Figure 13 shows efficiency curves for a traditional Doherty amplifier known in the prior art and an amplifier according to an embodiment of the invention. The input signal 21 in this case has a PAR of 7 dB and again a Rayleigh distribution. The y axis 143 represents the efficiency, and the x-axis 145 the frequency of the input signal. In this embodiment, however, the first supply voltage 31 provided to either the first amplifier 27 or the supply voltage 33 provided to the second amplifier 29 is varied as a function of the frequency of the input signal 21 . This means that the other supply voltage is kept constant. So in case the first supply voltage 31 is varied the supply voltage 33 is kept constant, and vice versa. In Figure 13, the curve labelled 142 represents such a amplifier circuit. The curve labelled 141 represents an amplifier circuit as known in the prior art. As can be seen from Figure 13, the efficiency of the amplifier circuit according to the invention is at least equal and most of the time higher than the efficiency of the amplifiers circuit known in the prior art. The amplifier circuit according to the invention performs thus substantially better than those known from the prior art.

The embodiments of the present invention provide a way of changing the Chireix modes of operation to provide increased efficiency that does not destroy the efficiency in the Doherty mode, and a way of changing the Doherty mode that does not impair the efficiency of the Chireix modes.

The embodiments of the invention are applicable to other amplifier

configurations, in addition to the specific examples provided in the description, including amplifier arrangements having even more wideband characteristics.

It is noted that the amplifier circuits and methods described in the embodiments of the invention may be used in any terminal of a telecommunications network including, but not limited to, radio base stations or eNodeBs (or other similar nodes in other telecommunication platforms), mobile or portable terminals, or any other device which requires a wideband amplifier with good efficiency across the bandwidth. It should also be noted that, although the embodiments of the invention have been described in relation to a telecommunications environment, the

embodiments of the invention may also be used with any application whereby a wideband amplifier is required with good efficiency across the bandwidth, including non-telecommunication applications.

It should be noted that the above-mentioned embodiments illustrate rather than limit the invention, and that those skilled in the art will be able to design many alternative embodiments without departing from the scope of the appended claims. The word "comprising" does not exclude the presence of elements or steps other than those listed in a claim, "a" or "an" does not exclude a plurality, and a single processor or other unit may fulfil the functions of several units recited in the claims. Any reference signs in the claims shall not be construed so as to limit their scope.

Claims

1 . An amplifier circuit (22; 42) comprising:

a first amplifier (27) for amplifying an input signal (21 );

a second amplifier (29) for amplifying the input signal (21 ), wherein an output of the first amplifier (27) and an output of the second amplifier (29) are coupled to a common output (23); and

wherein at least one of the first and second amplifiers (27, 29) is coupled to receive a supply voltage (31 ; 33) which, during use, is controlled as a function of the frequency of the input signal (21 ).

2. An amplifier circuit (22; 42) according to claim 1 , wherein the first amplifier (27) and the second amplifier (29) are adapted to receive a supply voltage (31 ; 33) which, during use, is controlled in relation to the frequency of the input signal (21 ).

3. An amplifier circuit (22; 42) according to claim 1 or 2, wherein the first amplifier (27) and the second amplifier (29) are coupled to the common output (23) via respective first and second transmission lines (26, 28), and wherein an electrical length of the first transmission line (26) is less than an electrical length of the second transmission line (28).

4. An amplifier circuit (22; 42) according to any one of claims 1 to 3, wherein the supply voltage (31 ; 97; 1 17; 127) to the first amplifier (27) is controlled such that the supply voltage (31 ; 97; 1 17; 127) has a minimal value for a centre frequency of the input signal, the centre frequency corresponding to a Doherty mode of operation, and increases from the minimal value as the frequency of the input signal (21 ) moves above or below the centre frequency.

5. An amplifier circuit (22; 42) according to any one of claims 1 to 4, wherein the supply voltage (33; 99; 1 19; 129) to the second amplifier (29) is controlled such that the supply voltage (33; 99; 1 19; 129) has a substantially maximum value for a centre frequency of the input signal (21 ), the centre frequency corresponding to a Doherty mode of operation, and decreases from the maximum value as the frequency of the input signal moves above or below the centre frequency.

6. An amplifier circuit (22; 42) according to claim 4 or 5, wherein the centre frequency corresponds to a Doherty mode of operation of the amplifier circuit, and wherein operation in outer regions as the frequency moves above or below the centre frequency corresponds to a Chireix mode of operation.

7. An amplifier circuit (22; 42) according to claim 4 or 5, wherein the functional relationship between the supply voltage (31 , 97, 1 17, 127; 33, 99, 1 19, 129) and frequency is substantially parabolic.

8. An amplifier circuit (22; 42) according to any one of claims 1 to 7, wherein the supply voltage (31 ; 97; 1 17; 127) provided to the first amplifier (27) is substantially equal to the supply voltage (33; 99; 1 19; 129) provided to the second amplifier (29), but lower than a nominal supply voltage, at:

first and second side lobes of the centre frequency of the input signal; or two thirds or four thirds of the centre frequency of the input signal; or a Chireix mode of operation.

9. An amplifier circuit (22; 42) according to claim 4 or 5, wherein the supply voltage (31 , 97, 1 17. 127; 33, 99, 1 19, 129) comprises a discontinuity at:

first and second side lobes of the centre frequency of the input signal (21 ) or two thirds or four thirds of the centre frequency of the input signal (21 ); or

a Chireix mode of operation.

10. An amplifier circuit (22; 42) according to any one of claims 3 to 9 wherein the first transmission line (26) comprises an electrical length of a quarter wavelength at the centre frequency, and wherein the second transmission line (28) comprises an electrical length of a half wavelength at the centre frequency.

A method for amplifying an input signal (21 ), the method comprising:

amplifying the input signal at a first amplifier (301 );

amplifying the input signal at a second amplifier (302), wherein an output of the first amplifier and an output of the second amplifier are coupled to a common output; and

controlling a supply voltage to at least one of the first amplifier or the second amplifier as a function of the frequency of the input signal (303).

12. A method according to claim 1 1 , wherein the step of controlling includes the step of controlling a supply voltage (31 , 97, 1 17, 127; 33, 99, 1 19, 129) to at least the first amplifier (27) and the second amplifier (29) as a function of the frequency of the input signal (21 ).

13. A method as claimed in claim 1 1 or 12, wherein the first amplifier (27) and the second amplifier (29) are coupled to the common output (23) via respective first and second transmission lines (26, 28), and wherein an electrical length of the first transmission line (26) is less than an electrical length of the second transmission line (28).

14. A method as claimed in any one of claims 1 1 to 13, wherein the step of controlling includes the steps of controlling the supply voltage (31 ; 97; 1 17; 127) to the first amplifier (27) such that the supply voltage (31 ; 97; 1 17; 127) has a minimal value for a centre frequency of the input signal, the centre frequency corresponding to a Doherty mode of operation, and increased from the minimal value as the frequency of the input signal (21 ) moves above or below the centre frequency.

15. A method as claimed in any one of claims 1 1 to 14, wherein the step of controlling includes the steps of controlling the supply voltage (33; 99; 1 19; 129) to the second amplifier (29) such that the supply voltage (33; 99; 1 19; 129) has a substantially maximum value for a centre frequency of the input signal (21 ), the centre frequency corresponding to a Doherty mode of operation, and decreases from the maximum value as the frequency of the input signal moves above or below the centre frequency.

16. A method as claimed in any one of claims 1 1 to 15, further comprising the steps of controlling the supply voltages provided to the first amplifier (27) and the second amplifier (29) such that the supply voltages are substantially equal, but lower than a nominal supply voltage, at:

17. A method as claimed in any one of claims 1 1 to 16, wherein the steps of controlling the supply voltages provided to the first amplifier (27) and/or the second amplifier (29) are performed dynamically during operation of the amplifier circuit, based on a determined frequency of the input signal.