Doherty Amplifier

Abstract

To provide a Doherty amplification device capable of shortening an adjustment time.SOLUTION: A Doherty amplification device comprises: a distribution section for distributing a distortion compensated input signal from a distortion compensation section into a first signal and a second signal; a first amplification section for amplifying the first signal; and a second amplification section for amplifying the second signal. The Doherty amplification device comprises: a composition section for composing the first signal from the first amplification section with the second signal from the second amplification section; and an adjustment section which is disposed between the distribution section and the first amplification section or the second amplification section and adjusts a phase delay value of the first signal or the second signal from the distribution section. On the basis of an output signal outputted from the composition section and an input signal inputted to the distortion compensation section, the adjustment section calculates a characteristic indicating a relation of a gain of the output signal with respect to reference power of the input signal. The adjustment section classifies the calculated characteristic into a plurality of regions in accordance with a level of the reference power and calculates a total value of gains on a region basis and on the basis of the total value on the region basis, a phase delay value with best radio characteristics is set to the adjustment section.SELECTED DRAWING: Figure 2

Description

Technical Field

[0001] The present invention relates to a Doherty amplifier device.

Background Art

[0002] Conventionally, power amplifiers for amplifying transmission power have been used in various electronic devices including base stations of mobile communication systems. Particularly in recent years, with the increase in communication speed, it has been expected to amplify transmission power with higher efficiency from the viewpoint of suppressing power consumption. The efficiency of an amplifier is known to be the highest in the output saturation state (non-linear state), and as an amplifier corresponding to this, the Doherty amplifier has been widely spread.

[0003] FIG. 28A is an explanatory diagram showing an example of a conventional Doherty amplifier 100. The Doherty amplifier 100 shown in FIG. 28A includes a carrier amplifier (CA: Carrier Amplifier) 101 and a peak amplifier (PA: Peak Amplifier) 102 connected in parallel with the CA 101. Further, the Doherty amplifier 100 includes a first λ / 4 line 103, a second λ / 4 line 104, and a third λ / 4 line 105. The CA 101 is an AB-class biased amplifier that operates constantly. The PA 102 is a C-class biased amplifier that turns on only when the input power to the Doherty amplifier 100 is a predetermined value or more. The first λ / 4 line 103 is a line connected to the output stage of the CA 101 for converting the output impedance of the CA 101. The second λ / 4 line 104 is a line connected to the input stage of the PA 102 for compensating the phase difference between the CA 101 and the PA 102 caused by the first λ / 4 line 103 to make them in-phase. The third λ / 4 line 105 is a matching circuit between the Doherty amplifier 100 and the load Z0.

[0004] Figure 28B is an explanatory diagram showing an example of the operation of the Doherty amplifier 100 when the input power is below a predetermined value. For example, when the input power is below a predetermined value, CA101 is ON and PA102 is OFF due to Class C bias, so the output impedance viewed from the point of combination with CA101 towards PA102 is ideally open. The load impedance viewed from the output terminal of CA101 is 2Z0 because the load of Z0 / 2 is viewed through the first λ / 4 line 103 with impedance Z0, and the first λ / 4 line 103 acts as an impedance conversion circuit. When the load impedance is 2Z0, CA101 is designed to operate efficiently, although the saturation power decreases.

[0005] Figure 28C is an explanatory diagram showing an example of the operation of the Doherty amplifier 100 when the input power is above a predetermined value. For example, when the input power is above a predetermined value, both CA101 and PA102 are turned ON. With the output load Z0 / 2 connected, the load impedance seen by CA101 and PA102 is Z0. Since the impedance of the first λ / 4 line 103 in the output stage of CA101 is Z0, no impedance transformation is performed, and the load impedance seen from the output terminal of CA101 is also Z0. When the load impedance seen from the output terminal of CA101 is Z0, CA101 and PA102 are designed to have a large saturation power, so the Doherty amplifier 100 will output the desired saturation power.

[0006] However, while conventional Doherty amplifiers 100 are highly efficient, their linearity is poor because the gain and phase of the output power change with respect to the input power. Therefore, digital predistortion (DPD) is known as a technique to compensate for this linearity. Figure 29 is an explanatory diagram showing an example of a conventional Doherty amplifier 150. The Doherty amplifier 150 shown in Figure 29 includes a Doherty amplifier section 151, a DPD 152 positioned before the Doherty amplifier section 151, a DAC (Digital Analog Converter) 153, a driver amplifier 154, and an ADC (Analog Digital Converter) 155.

[0007] The Doherty amplifier section 151 includes CA161, PA162, a first λ / 4 line 163, a second λ / 4 line 164, and a third λ / 4 line 165. CA161 is a Class AB biased amplifier that operates continuously. PA162 is a Class C biased amplifier that turns on only when the input power exceeds a predetermined value. The first λ / 4 line 163 is connected to the output stage of CA161 and is a line for converting the output impedance of CA161. The second λ / 4 line 164 is connected to the input stage of PA162 and is a line for compensating for the phase difference between CA161 and PA162 caused by the first λ / 4 line. The third λ / 4 line 165 is a line that matches the impedance of the combined output of CA161 and PA162 with the impedance of the subsequent stage.

[0008] When the input power is below a predetermined value, the load impedance seen from the output terminal of CA161 is 2Z0 because the output load Z0 / 2 is seen through the first λ / 4 line 163 with impedance Z0, and the first λ / 4 line 163 acts as an impedance conversion circuit. In contrast, for example, when the input power is above a predetermined value, the load impedance seen by CA161 and PA162 with the output load Z0 / 2 connected is Z0. Since the impedance of the first λ / 4 line 163 in the output stage of CA161 is Z0, no impedance conversion is performed, and the load impedance seen from the output terminal of CA161 is also Z0. As a result, the Doherty amplifier 151 combines the signal amplified by CA161 and the signal amplified by PA162 and outputs it.

[0009] The DPD152 comprises a first capture unit 152A, a second capture unit 152B, a calculation unit 152C, and a distortion compensation unit 152D. The first capture unit 152A extracts a portion of the input signal as a REF (Reference) signal. The second capture unit 152B extracts a portion of the output signal from the output stage of the Doherty amplifier 151 as an FB (Feedback) signal. The calculation unit 152C calculates a distortion compensation coefficient such that the difference between the REF signal extracted by the first capture unit 152A and the FB signal extracted by the second capture unit 152B is eliminated, and sets the calculated distortion compensation coefficient in the distortion compensation unit 152D. Based on the set distortion compensation coefficient, the distortion compensation unit 152D compensates for the nonlinear distortion of the input signal such that the difference between the REF signal and the FB signal from the output stage of the Doherty amplifier 151 is eliminated. As a result, the linearity of the output signal at the output stage of the Doherty amplifier 151 can be ensured.

[0010] The DAC153 converts the input signal, which has been compensated by the distortion compensation section 152D within the DPD152, into an analog signal, and outputs the converted analog signal to the driver amplifier 154. The driver amplifier 154 amplifies the input signal to a specified level and outputs the amplified input signal to the Doherty amplifier section 151.

[0011] The ADC155 digitally converts the FB signal, which is part of the output signal from the output stage of the Doherty amplifier 151, and outputs the digitally converted FB signal to the second capture unit 152B.

[0012] The DPD152 compensates the input signal so that the difference between the REF signal and the FB signal of the output stage of the Doherty amplifier 151 is eliminated. By repeating this process, the linearity of the Doherty amplifier 151 can be improved.

[0013] However, in the Doherty amplifier 150 shown in Figure 29, since it has only one input, the phase difference, delay difference, and amplitude difference of the CA161 and PA162 paths are fixed by the hardware design. Therefore, a two-input Doherty amplifier is known that separates the paths of CA161 and PA162 to create two inputs, and places an adjustment unit to adjust the phase difference, delay difference, and amplitude difference for each path. Figure 30 is an explanatory diagram showing an example of a two-input Doherty amplifier 200.

[0014] The two-input Doherty amplifier 200 shown in Figure 30 comprises a DPD 210, a two-input Doherty amplifier section 220, and an ADC 230. The DPD 210 includes a first capture section 211, a second capture section 212, a calculation section 213, and a distortion compensation section 214. The first capture section 211 extracts a portion of the input signal as a REF signal. The second capture section 212 extracts a portion of the output signal from the output stage of the two-input Doherty amplifier section 220 as an FB signal. The calculation section 213 calculates a distortion compensation coefficient such that the difference between the REF signal extracted by the first capture section 211 and the FB signal extracted by the second capture section 212 is eliminated, and sets the calculated distortion compensation coefficient in the distortion compensation section 214. The distortion compensation unit 214 compensates for the nonlinear distortion of the input signal based on a set distortion compensation coefficient, so that the difference between the REF signal and the FB signal of the output stage of the two-input Doherty amplifier unit 220 is eliminated. As a result, the linearity of the output signal of the output stage of the two-input Doherty amplifier unit 220 can be ensured.

[0015] The two-input Doherty amplifier 220 includes a distribution unit 221, a first amplifier 222 located in a first path through which CA222D is placed, a second amplifier 223 located in a second path through which PA223D is placed, a combining unit 224, and an output λ / 4 line 225.

[0016] The first amplification unit 222 includes a first adjustment unit 222A, a first DAC 222B, a first driver amplifier 222C, a CA222D, and a λ / 4 transmission line 222E. The first adjustment unit 222A adjusts the phase difference, delay difference, and amplitude difference between the first path and the second path. The first DAC 222B converts the input signal, which has been compensated by the distortion compensation unit 214 in the DPD 210, into an analog signal and outputs the converted input signal to the first driver amplifier 222C. The first driver amplifier 222C amplifies the input signal to a specified level and outputs the amplified input signal to the CA222D. The CA222D is a continuously operating Class AB bias amplifier. The λ / 4 transmission line 222E is connected to the output stage of the CA222D and is a transmission line for converting the output impedance of the CA222D.

[0017] The second amplification unit 223 includes a second adjustment unit 223A, a second DAC 223B, a second driver amplifier 223C, and a PA 223D. The second adjustment unit 223A adjusts the phase difference, delay difference, and amplitude difference between the second path and the first path. The second DAC 223B converts the input signal, which has been compensated by the distortion compensation unit 214 in the DPD 210, into an analog signal and outputs the converted input signal to the second driver amplifier 223C. The second driver amplifier 223C amplifies the input signal to a specified level and outputs the amplified input signal to the PA 223D. The PA 223D is a Class C bias amplifier that turns on only when the input power is above a predetermined value.

[0018] The combining unit 224 combines the input signals after amplification by CA222D and PA223D. The output λ / 4 line 225 is a line that matches the impedance of the combining unit 224, which combines the outputs of CA222D and PA223D, with the impedance of the subsequent stage.

[0019] The first adjustment unit 222A adjusts the phase difference, delay difference, and amplitude difference between the first path and the second path. As a result, the phase difference, delay difference, and amplitude difference between the first path and the second path can be adjusted. The second adjustment unit 223A adjusts the phase difference, delay difference, and amplitude difference between the second path and the first path. As a result, the phase difference, delay difference, and amplitude difference between the second path and the first path can be adjusted.

[0020] In the 2-input Doherty amplifier 200, the degree of freedom to adjust the phase difference, delay difference, and amplitude difference between paths using adjustment sections for each path is increased, and improvements in amplification efficiency, distortion compensation performance, and bandwidth can be expected. [Prior art documents] [Patent Documents]

[0021] [Patent Document 1] International Publication No. 2021 / 192204 [Patent Document 2] International Publication No. 2021 / 220338 [Overview of the project] [Problems that the invention aims to solve]

[0022] However, in the two-input Doherty amplifier 200, the parameters of the adjustment sections 222A and 223A, which adjust the phase difference and delay difference between paths, must be adjusted while checking the actual distortion by connecting an external measuring device such as a spectrum analyzer. Therefore, it takes time to adjust the phase difference and delay difference between paths to achieve the best distortion characteristics.

[0023] One objective is to provide a Doherty amplifier that can significantly reduce the time required to adjust the phase difference and delay difference between the CA and PA paths to achieve optimal distortion characteristics. [Means for solving the problem]

[0024] A Doherty amplifier of one embodiment includes a distribution unit that distributes an input signal after distortion compensation from a distortion compensation unit that compensates for distortion of the input signal into a first signal and a second signal, a first amplification unit that amplifies the first signal, and a second amplification unit that amplifies the second signal. Further, the Doherty amplifier includes a combining unit that combines the first signal from the first amplification unit and the second signal from the second amplification unit, and an adjustment unit that is disposed between the distribution unit and the first amplification unit or the second amplification unit and adjusts the phase delay value of the first signal or the second signal from the distribution unit. The adjustment unit includes a first calculation unit that calculates a characteristic indicating the relationship between the gain of the output signal with respect to the reference power of the input signal based on the output signal output from the combining unit and the input signal input to the distortion compensation unit, and a classification unit that classifies the calculated characteristic into a plurality of regions according to the level of the reference power. The adjustment unit includes a second calculation unit that calculates a statistical value of the gain for each classified region, and a determination unit that sets a phase delay value at which the radio characteristics are optimal in the adjustment unit based on the calculated statistical value for each region.

Advantages of the Invention

[0025] As one aspect, the time for adjusting the phase difference and delay difference between the paths of the first amplification unit (CA) and the second amplification unit (PA) can be significantly shortened so that the distortion characteristics are optimal.

Brief Description of the Drawings

[0026] [Figure 1] FIG. 1 is an explanatory diagram showing an example of the Doherty amplifier of Example 1. [Figure 2] FIG. 2 is an explanatory diagram showing an example of the adjustment unit of Example 1. [Figure 3A] FIG. 3A is an explanatory diagram showing an example of the AM-AM (gain) characteristic before applying distortion compensation by DPD. [Figure 3B] FIG. 3B is an explanatory diagram showing an example of the AM-AM (gain) characteristic after applying distortion compensation by DPD. [Figure 4A] FIG. 4A is an explanatory diagram showing an example of the AM-AM (gain) characteristic before phase delay adjustment by the adjustment unit. [Figure 4B]Figure 4B is an explanatory diagram showing an example of the AM-AM (gain) characteristics after phase delay adjustment in the adjustment section. [Figure 5A] Figure 5A is an explanatory diagram showing an example of a phase table (phase value vs. REF power). [Figure 5B] Figure 5B is an explanatory diagram showing an example of a delay table (delay value vs. REF power). [Figure 6] Figure 6 is a flowchart showing an example of the processing operation of the control unit within the adjustment unit involved in the first optimal value determination process. [Figure 7] Figure 7 is a flowchart showing an example of the processing operation of the control unit involved in the provisional phase value determination process. [Figure 8] Figure 8 is a flowchart showing an example of the processing operation of the control unit involved in the provisional delay value determination process. [Figure 9] Figure 9 is a flowchart showing an example of the processing operation of the control unit involved in the second phase value determination process. [Figure 10] Figure 10 is an explanatory diagram illustrating an example of a method for determining the second phase value of the largest first-order and the second phase value of the largest second-order. [Figure 11] Figure 11 is a flowchart showing an example of the processing operation of the control unit involved in the first phase value determination process. [Figure 12] Figure 12 is a flowchart showing an example of the processing operation of the control unit involved in the first delay value determination process. [Figure 13] Figure 13 is an explanatory diagram illustrating an example of a method for determining the first delay value of the maximum first rank and the first delay value of the maximum second rank. [Figure 14] Figure 14 is a flowchart showing an example of the processing operation of the control unit involved in the second delay value determination process. [Figure 15] Figure 15 is an explanatory diagram showing an example of the Doherty amplifier of Example 2. [Figure 16] Figure 16 is an explanatory diagram showing an example of the adjustment section of Embodiment 2. [Figure 17] Figure 17 is an explanatory diagram showing an example of an amplitude table. [Figure 18]Figure 18 is a flowchart showing an example of the processing operation of the control unit within the adjustment unit involved in the second optimal value determination process. [Figure 19] Figure 19 is a flowchart showing an example of the processing operation of the control unit involved in amplitude pattern determination. [Figure 20] Figure 20 is an explanatory diagram showing an example of the Doherty amplifier of Example 3. [Figure 21] Figure 21 is an explanatory diagram showing an example of the adjustment section of Embodiment 3. [Figure 22] Figure 22 is an explanatory diagram showing an example of an amplitude-frequency correction pattern. [Figure 23] Figure 23 is a flowchart showing an example of the processing operation of the control unit within the adjustment unit involved in the third optimal value determination process. [Figure 24] Figure 24 is a flowchart showing an example of the processing operation of the control unit involved in the amplitude-frequency correction pattern determination process. [Figure 25] Figure 25 is a flowchart showing an example of the processing operation of the control unit involved in the provisional phase value determination process in Example 4. [Figure 26] Figure 26 is a flowchart showing an example of the processing operation of the control unit involved in the provisional delay value determination process in Example 4. [Figure 27] Figure 27 is an explanatory diagram showing an example of a base station hardware configuration, which is an example of an application of a Doherty amplifier. [Figure 28A] Figure 28A is an explanatory diagram showing an example of a conventional Doherty amplifier. [Figure 28B] Figure 28B is an explanatory diagram illustrating an example of the operation of a Doherty amplifier when the input power is below a predetermined value. [Figure 28C] Figure 28C is an explanatory diagram showing an example of the operation of a Doherty amplifier when the input power is above a predetermined value. [Figure 29] Figure 29 is an explanatory diagram showing an example of a conventional Doherty amplifier. [Figure 30] Figure 30 is an explanatory diagram showing an example of a two-input Doherty amplifier. [Modes for carrying out the invention]

[0027] Hereinafter, embodiments of the Doherty amplifier disclosed in this application will be described in detail based on the drawings. However, these embodiments do not limit the disclosed technology. Furthermore, the embodiments described below may be combined as appropriate, provided they do not contradict each other. [Examples]

[0028] Figure 1 is an explanatory diagram showing an example of a Doherty amplifier 1 of Embodiment 1. The Doherty amplifier 1 shown in Figure 1 is a two-input Doherty amplifier. The Doherty amplifier 1 includes a DPD (Digital Predistortion) 10, a two-input Doherty amplifier 20, and an ADC (Analog Digital Converter) 30. The DPD 10 compensates for the nonlinear distortion of the input signal so that the difference between the input signal and the output signal from the two-input Doherty amplifier 20 is eliminated. The two-input Doherty amplifier 20 amplifies the input signal after distortion compensation by the DPD 10. The ADC 30 digitally converts the FB (Feedback) signal, which is a part of the output signal after amplification by the two-input Doherty amplifier 20.

[0029] The DPD10 includes a first capture unit 11, a second capture unit 12, a calculation unit 13, and a distortion compensation unit 14. The first capture unit 11 extracts a portion of the input signal as a REF (Reference) signal. The second capture unit 12 extracts a portion of the output signal from the output stage of the two-input Doherty amplifier 20 as an FB signal. The calculation unit 13 calculates a distortion compensation coefficient so that the difference between the REF signal extracted by the first capture unit 11 and the FB signal extracted by the second capture unit 12 is eliminated, and sets the calculated distortion compensation coefficient in the distortion compensation unit 14. Based on the set distortion compensation coefficient, the distortion compensation unit 14 compensates for the nonlinear distortion of the input signal so that the difference between the RFE signal and the FB signal is eliminated.

[0030] The ADC30 digitally converts the FB signal, which is part of the output signal from the output stage of the 2-input Doherty amplifier 20, and outputs the digitally converted FB signal to the second capture unit 12.

[0031] The two-input Doherty amplifier 20 includes a distribution unit 21, a first amplifier 22 arranged in a first path through which a CA (Carrier Amplifier) ​​22C is located, and a second amplifier 23 arranged in a second path through which a PA (Peak Amplifier) ​​23C is located. Furthermore, the two-input Doherty amplifier 20 includes a combining unit 24, an output λ / 4 line 25, and an adjustment unit 26.

[0032] The first amplification section 22 includes a first DAC 22A, a first driver amplifier 22B, a CA 22C, and a λ / 4 transmission line 22D. The first DAC 22A converts the first signal, which is the input signal after compensation by the distortion compensation section 14 in the DPD 10, into an analog signal and outputs the converted input signal to the first driver amplifier 22B. The first driver amplifier 22B amplifies the input signal to a specified level and outputs the amplified input signal to the CA 22C. The CA 22C is a continuously operating Class AB bias amplifier. The λ / 4 transmission line 22D is connected to the output stage of the CA 22C and is a transmission line for converting the output impedance of the CA 22C.

[0033] The second amplification section 23 includes a second DAC 23A, a second driver amplifier 23B, and a PA 23C. The second DAC 23A converts the second signal, which is the input signal after compensation by the distortion compensation section 14 in the DPD 10, into an analog signal and outputs the converted input signal to the second driver amplifier 23B. The second driver amplifier 23B amplifies the input signal to a specified level and outputs the amplified input signal to the PA 23C. The PA 23C is a Class C bias amplifier that turns on only when the input power is above a predetermined value.

[0034] The combining unit 24 combines the input signals after amplification by CA22C and PA23C. The output λ / 4 line 25 is a line that matches the impedance of the combining unit 24, which combines the outputs of CA22C and PA23C, with the impedance of the subsequent stage. The adjustment unit 26 automatically adjusts the phase difference and delay difference between the second path and the first path.

[0035] Figure 2 is an explanatory diagram showing an example of the adjustment unit 26 of Embodiment 1. The adjustment unit 26 shown in Figure 2 adjusts the phase and delay values ​​of the second signal passing through the second path according to the reference power of the input signal in order to adjust the phase difference and delay difference between the PA23C of the second path and the CA22C of the first path so that the distortion characteristics are best. In other words, the adjustment unit 26 adjusts the delay and phase values ​​of PA23C to maximize the gain near the saturation power after Doherty synthesis of the combining unit 24 and eliminate gain-phase variations. The adjustment unit 26 includes a phase adjustment unit 41, a delay adjustment unit 42 arranged in series with the phase adjustment unit 41, a power address calculation unit 43, a phase table 44, a delay table 45, and a control unit 46.

[0036] The phase adjustment unit 41 adjusts the phase value of the second signal flowing through the second path in which PA23C is located. The phase table 44 is a table that manages the phase value to maximize the gain near the saturation power after Doherty synthesis of the combining unit 24 and to reduce the variation in gain and phase. The phase table 44 pre-manages the optimal phase value of the phase adjustment unit 41 according to the REF power for each amplifier.

[0037] Figure 5A is an explanatory diagram showing an example of a phase table 44 (phase value: REF power). The first phase value a in the phase table 44 shown in Figure 5A is the phase value in the low-level REF power region, for example, the power level region where PA23C is in the OFF state. The second phase value b in the phase table 44 shown in Figure 5A is the phase value in the high-level REF power region, for example, the power level region where PA23C is in the ON state. The first phase value a is -50 degrees, and the second phase value b is -30 degrees. Note that changing the phase value has the effect of improving the distortion characteristics with respect to the output power. Also, the function of the phase table 44 should be changed according to the amplifier being used.

[0038] The delay adjustment unit 42 adjusts the delay value of the second signal flowing through the second path where PA23C is located. The delay table 45 is a table that manages the delay value that minimizes the delay difference with respect to the REF power of the second path where PA23C is located relative to the first path where CA22C is located. The delay table 45 pre-manages the optimal delay value for each amplifier according to the REF power.

[0039] Figure 5B is an explanatory diagram showing an example of a delay table 45 (delay value: REF power). The first delay value c in the delay table 45 shown in Figure 5B is the delay value in the low-level REF power range, for example, in the power level range where PA23C is in the OFF state. The second delay value d in the delay table 45 shown in Figure 5B is the delay value in the high-level REF power range, for example, in the power level range where PA23C is in the ON state. The first delay value c is 230, and the second delay value d is 150. For the sake of explanation, the delay values ​​shown in Figure 5B are expressed as digital values ​​(decimal) uniquely associated with the delay time. Changing the delay value has the effect of improving the distortion characteristics with respect to the output power. Also, the function of the delay table 45 should be changed according to the amplifier being used.

[0040] The power address calculation unit 43 calculates the REF power from the input signal from the input stage of the distortion compensation unit 14.

[0041] The control unit 46 includes a first calculation unit 51, a classification unit 52, a second calculation unit 53, and a determination unit 54. The first calculation unit 51 calculates the AM (Amplitude Modulation)-AM (gain) characteristic using the REF signal and FB signal obtained by the calculation unit 13 in the DPD 10, with distortion compensation of the distortion compensation unit 14 not applied. The AM-AM (gain) characteristic is a characteristic that shows the relationship between the gain of the output signal and the reference power (REF power) of the input signal of the two-input Doherty amplifier 20. The REF power and the gain of the output signal of the AM-AM (gain) characteristic shall be expressed normalized by the input power and gain when the two-input Doherty amplifier 20 is transmitting at its rated output. The first calculation unit 51 calculates the gain of the AM-AM (gain) characteristic using (Equation 1).

[0042]

number

[0043] Furthermore, the first calculation unit 51 calculates the REF power of the AM-AM (gain) characteristic using (Equation 2).

[0044]

number

[0045] For example, the AM-AM (gain) characteristics before and after distortion compensation are described when a modulated wave is input to a two-input Doherty amplifier 20 with poor linearity. Figure 3A is an explanatory diagram showing an example of the AM-AM (gain) characteristics before distortion compensation is applied by the DPD 10, and Figure 3B is an explanatory diagram showing an example of the AM-AM (gain) characteristics after distortion compensation is applied by the DPD 10. In the AM-AM (gain) characteristics shown in Figures 3A and 3B, the vertical axis is gain and the horizontal axis is REF power. Each dot represents the gain with respect to REF power.

[0046] The AM-AM (gain) characteristics before distortion compensation is applied, as shown in Figure 3A, show that the gain decreases as the REF power increases, and the gain variation becomes large in the low REF power region. In contrast, the AM-AM (gain) characteristics after distortion compensation is applied, as shown in Figure 3B, become linear, so the gain remains constant with respect to the REF power, and the gain variation can be suppressed in the low REF power region. In other words, by performing distortion compensation on the DPD 10, the linearity of the two-input Doherty amplifier 20 can be improved.

[0047] Next, the AM-AM (gain) characteristics before and after phase delay adjustment in the adjustment unit 26 will be explained. Figure 4A is an explanatory diagram showing an example of the AM-AM (gain) characteristics before phase delay adjustment in the adjustment unit 26, and Figure 4B is an explanatory diagram showing an example of the AM-AM (gain) characteristics after phase delay adjustment in the adjustment unit 26. The AM-AM (gain) characteristics can be classified into Low, High, and Peak regions depending on the level of REF power. Note that 0 dB of REF power shown in Figures 4A and 4B is the REF power (I^2 + Q^2) when transmitting the average transmit power of the device. The Low region is the level range of REF power in which only CA22C in the 2-input Doherty amplifier 20 is ON, i.e., PA23C is OFF, for example, the region of less than 5 dB. The Peak region is the level range of REF power in which both CA22C and PA23C in the 2-input Doherty amplifier 20 are ON, for example, the region of 5 dB or more and less than 7 dB. The High region is the level range of the REF power when transitioning from the Low region to the Peak region, for example, the region of 7 dB or higher.

[0048] The classification unit 52 classifies the AM-AM characteristics (gain) before and after phase delay adjustment into Low, High, and Peak regions according to the level of REF power.

[0049] The second calculation unit 53 calculates statistical values ​​for each region within the AM-AM(gain) characteristic calculated using the input signal without distortion compensation applied to the DPD 10. The distortion compensation effect tends to increase when the statistical values ​​for each region satisfy the first or second condition. The first condition takes precedence over the second condition.

[0050] The first condition is that, as shown in Figure 3B, there is little gain reduction in the region where the REF power is high, for example, in the Peak region. The first condition is used, for example, when determining the phase value, provisional phase value, and delay value, which will be described later. When the first condition is met, the linearity is improved.

[0051] The second condition is that, as shown in Figure 3B, the gain variation is small in the low REF power region, for example, in the High and Low regions. This second condition is used, for example, when determining the provisional delay value, which will be discussed later. When the second condition is met, the memory effect of the amplifier is reduced.

[0052] The second calculation unit 53 calculates statistical values ​​for each region within the calculated AM-AM(gain) characteristic, for example, the average gain Gain ave and the standard deviation Gain std. The second calculation unit 53 calculates the average gain Gain ave using (Equation 3).

[0053]

number

[0054] Furthermore, the second calculation unit 53 calculates the gain standard deviation value Gain std using (Equation 4).

[0055]

number

[0056] Furthermore, (Equation 4) uses (Equation 5).

[0057]

number

[0058] The second calculation unit 53 changes the phase value in the phase adjustment unit 41 as a parameter and sequentially calculates the average gain, which is a statistical value for each region. The determination unit 54 determines a provisional phase value, for example, based on the average gain of the Peak region. The control unit 46 sets the determined provisional phase value in the phase adjustment unit 41.

[0059] Furthermore, the second calculation unit 53, while keeping the provisional phase value set in the phase adjustment unit 41, changes the delay value in the delay adjustment unit 42 as a parameter and sequentially calculates the gain standard deviation value, which is a statistical value for each region. The determination unit 54 determines a provisional delay value that satisfies the second condition, for example, the minimum value from the calculated gain standard deviation value in the High region. The control unit 46 sets the determined provisional delay value in the delay adjustment unit 42.

[0060] The second calculation unit 53, while keeping the provisional delay value set in the delay adjustment unit 42, changes the phase value in the phase adjustment unit 41 as a parameter and sequentially calculates the average gain, which is a statistical value for each region. The determination unit 54 determines the optimal phase value (first phase value a and second phase value b) that satisfies the first condition, for example, the maximum from the calculated average gain of the Peak region, for example, the optimal phase value (first phase value a and second phase value b) that provides the best wireless characteristics based on the correlation between distortion and AM-AM (gain) characteristics. The control unit 46 sets the determined optimal phase value (first phase value and second phase value) in the phase adjustment unit 41.

[0061] The second calculation unit 53, while keeping the optimal phase value set in the phase adjustment unit 41, changes the delay value in the delay adjustment unit 42 as a parameter and sequentially calculates the average gain, which is a statistical value for each region. The determination unit 54 determines, for example, the optimal first delay value c that is the maximum from the calculated average gain of the High region. The control unit 46 sets the determined optimal first delay value c in the delay adjustment unit 42.

[0062] Furthermore, the second calculation unit 53, while keeping the optimal phase value set in the phase adjustment unit 41, changes the delay value in the delay adjustment unit 42 as a parameter and sequentially calculates the average gain, which is a statistical value for each region. The determination unit 54 determines the optimal second delay value d that satisfies the first condition, for example, the maximum value from the calculated average gain of the Peak region. The control unit 46 sets the determined optimal second delay value d in the delay adjustment unit 42.

[0063] The AM-AM (gain) characteristics after phase delay adjustment, shown in Figure 4B, show a reduction in the gain drop in the peak region compared to the AM-AM (gain) characteristics before phase delay adjustment, shown in Figure 4A. As a result, the distortion compensation effect tends to be higher.

[0064] The first calculation unit 51 in the control unit 46 calculates the AM-AM (gain) characteristics after setting predetermined phase and delay values ​​with the DPD 10 distortion compensation not applied. The determination unit 54 in the control unit 46 uses the calculated AM-AM (gain) characteristics to determine the optimal first phase value a, the optimal second phase value b, and the optimal first delay value c, which result in the maximum average gain within the Peak region. The determination unit 54 also uses the calculated AM-AM (gain) characteristics to determine the optimal second delay value d, which results in the maximum average gain within the High region.

[0065] The control unit 46 uses the AM-AM(gain) characteristics calculated without distortion compensation to determine the optimal first phase value a and the optimal second phase value b, which represent the maximum average gain within the Peak region. The control unit 46 then stores and updates the optimal first phase value a and the optimal second phase value b as setting parameters.

[0066] Furthermore, the control unit 46 determines a first delay value c that is the maximum average gain in the High region using the AM-AM(gain) characteristics calculated without distortion compensation. In addition, the control unit 46 determines a second delay value d that is the maximum average gain in the Peak region using the AM-AM(gain) characteristics calculated without distortion compensation. Then, the control unit 46 stores and updates the optimal first delay value c and the optimal second delay value d as setting parameters.

[0067] Next, the operation of the Doherty amplifier 1 of Embodiment 1 will be described. Figure 6 is a flowchart showing an example of the processing operation of the control unit 46 within the adjustment unit 26 involved in the first optimal value determination process. The first optimal value determination process is, for example, the process of determining the optimal phase value and optimal delay value of the two-input Doherty amplifier unit 20 in a state where distortion compensation is not applied. Note that the first optimal value determination process is a process that is executed, for example, at the time of product shipment.

[0068] In Figure 6, the control unit 46 executes the provisional phase value determination process shown in Figure 7 (step S11) to determine a provisional phase value, which is a provisional phase value that maximizes the average gain value in the Peak region of the AM-AM (gain) characteristic. The provisional phase value is a provisional first phase value and a provisional second phase value. After determining the provisional phase value, the control unit 46 executes the provisional delay value determination process shown in Figure 8 (step S12) to determine a provisional delay value, which is a provisional delay value that minimizes the gain standard deviation in the High region of the AM-AM (gain) characteristic. The provisional delay value is a provisional first delay value and a provisional second delay value.

[0069] After determining a provisional delay value, the control unit 46 executes a second phase value determination process as shown in Figure 9 (step S13), which determines an optimal second phase value b, that is the phase value at which the average gain value in the Peak region of the AM-AM (gain) characteristic is maximized. Furthermore, after determining the second phase value b, the control unit 46 executes a first phase value determination process as shown in Figure 11 (step S14), which determines an optimal first phase value a, that is the phase value at which the average gain value in the Peak region of the AM-AM (gain) characteristic is maximized.

[0070] After determining the first phase value a, the control unit 46 executes a first delay value determination process as shown in Figure 12 (step S15) to determine the optimal first delay value c, which is the delay value that maximizes the average gain in the High region of the AM-AM (gain) characteristic. Furthermore, after determining the first delay value c, the control unit 46 executes a second delay value determination process as shown in Figure 14 (step S16) to determine the optimal second delay value d, which is the delay value that maximizes the average gain in the Peak region of the AM-AM (gain) characteristic. Then, it terminates the processing operation shown in Figure 6.

[0071] Figure 7 is a flowchart showing an example of the processing operation of the control unit 46 involved in the provisional phase value determination process. In Figure 7, the control unit 46 sets the number of phase swing trials i to "0" (step S21), and sets the initial phase value as the set phase value (a=b) by referring to the phase table 44 (step S22). For the sake of explanation, the maximum number of phase swing trials is set to N=3, but it is not limited to 3 and can be changed as appropriate. The control unit 46 determines whether the number of phase swing trials being set is i=N (step S23).

[0072] If the number of phase swing trials is not i=N (step S23: No), the control unit 46 determines whether the number of phase swing trials is i=N-1 (step S23A). If the number of phase swing trials is not i=N-1 (step S23A: No), the control unit 46 sets the phase adjustment unit 41 to a set phase value obtained by adding the current set phase value ± adjustment value in order to swing the phase (step S24). For example, if the number of phase swing trials is i=0, the adjustment value is set to 20°, and if the number of phase swing trials is i=1, the adjustment value is set to 10°. In other words, the phase swing is large, 20° when i=0, and small, 10° when i=1, to gradually narrow down to the optimal phase value.

[0073] In step S24, the first calculation unit 51 in the control unit 46 sets a set phase value in the phase adjustment unit 41 and then calculates the AM-AM (gain) characteristics without distortion compensation applied (step S25). The first calculation unit 51 calculates the AM-AM (gain) characteristics based on the FB signal and REF signal without distortion compensation applied by the DPD 10.

[0074] The classification unit 52 within the control unit 46 calculates the AM-AM(gain) characteristics and then classifies the calculated AM-AM(gain) characteristics into Low, High, and Peak regions according to the REF power level (step S26). Furthermore, the second calculation unit 53 within the control unit 46 calculates the average gain of the Peak region from the AM-AM(gain) characteristics (step S27).

[0075] The determination unit 54 within the control unit 46 determines whether the calculated average gain of the Peak region exceeds the average gain of the maximum first-rank phase value (step S28). The maximum first-rank phase value is the phase value at which the average gain of the Peak region is maximized. If the average gain of the Peak region exceeds the average gain of the maximum first-rank phase value (step S28: Yes), the determination unit 54 executes an update process to update the maximum first-rank phase value as the first-rank setting parameter (step S29). The update process updates the maximum first-rank phase value to the maximum second-rank phase value, updates the average gain of the maximum first-rank phase value to the average gain of the maximum second-rank phase value, updates the setting value of the maximum first-rank phase value to the current setting value, and stores the average gain as the average gain of the maximum first-rank. The maximum second-rank phase value is the phase value at which the average gain of the Peak region is second from the top.

[0076] After the control unit 46 performs an update process to update the maximum first-order phase value and the maximum second-order phase value, it increments the number of phase swing trials by +1 (step S30) and proceeds to step S23 to determine whether the number of phase swing trials is i=N.

[0077] If the average gain of the Peak region does not exceed the average gain of the maximum 1st phase value (step S28: No), the determination unit 54 determines whether the average gain of the Peak region exceeds the average gain of the maximum 2nd phase value (step S31). If the average gain of the Peak region exceeds the average gain of the maximum 2nd phase value (step S31: Yes), the determination unit 54 performs an update process to update the maximum 2nd phase value as the 2nd setting parameter (step S32). The update process in step S32 updates the maximum 2nd phase value to the current setting value and stores the average gain as the average gain of the maximum 2nd. After performing the update process in step S32, the control unit 46 proceeds to the process in step S30 to increment the number of phase swing trials by +1. Furthermore, if the average gain of the Peak region does not exceed the average gain of the second-highest phase value (step S31: No), the determination unit 54 proceeds to the process of step S30 incrementing the number of phase swing trials by +1.

[0078] If the number of phase swing trials is i=N-1 (step S23A: Yes), the determination unit 54 sets the phase value calculated by (maximum 1st position phase value + maximum 2nd position phase value) ÷ 2 to the phase adjustment unit 41 (step S33), and proceeds to the process in step S25. For example, if the maximum 1st position phase value is 30°, the average gain of the Peak region is -2dB, the maximum 2nd position phase value is 50°, the average gain of the Peak region is -3dB, the current set phase value is 40°, and the average gain of the Peak region is -1.5dB, then the maximum 1st position phase value will be 40° and the maximum 2nd position phase value will be 30°. The next set phase value will then be (40+30)÷2, which is 35°.

[0079] If the number of phase adjustment trials is i=N (step S23: Yes), the determination unit 54 determines the maximum first-order phase value as the provisional phase value (step S34). Furthermore, the determination unit 54 sets the provisional phase value to the phase adjustment unit 41 (step S35), and terminates the processing operation shown in Figure 7.

[0080] In the provisional phase value determination process shown in Figure 7, the provisional phase value can be determined from the largest first-rank phase value and the largest second-rank phase value with the highest average gain in the Peak region.

[0081] Figure 8 is a flowchart showing an example of the processing operation of the control unit 46 involved in the provisional delay value determination process. In Figure 8, the control unit 46 sets the number of delay adjustment trials i to "0" (step S41), and refers to the delay table 45 to set the initial delay value as the set delay value (c, d=c+α) (step S42). For the sake of explanation, the maximum number of delay adjustment trials is set to N=5, but it is not limited to 5 and can be changed as appropriate. α is set to 10. The control unit 46 determines whether the number of delay adjustment trials being set is i=N (step S43).

[0082] If the number of delay adjustment trials is not i=N (step S43: No), the control unit 46 determines whether the number of delay adjustment trials is i=N-1 (step S43A). If the number of delay adjustment trials is not i=N-1 (step S43A: No), the control unit 46 sets the delay adjustment unit 42 to a set delay value obtained by adding the current set delay value ± adjustment value in order to adjust the delay (step S44). If the number of delay adjustment trials is i=0, for example, the adjustment value is set to 32 and the set delay value is set to ±32. If the number of delay adjustment trials is i=1, for example, if the current set delay value, which is the smallest first-order delay value, is 32, the set delay value c is set to 64, 96, 128, and if the current set delay value is -32, the set delay values ​​are set to -128, -96, -64. Furthermore, when the number of delay adjustment trials is i=2, the current set delay value (the smallest first-order delay value) and the adjustment value are set to ±16. When the number of delay adjustment trials is i=3, the current set delay value (the smallest first-order delay value) and the adjustment value are set to ±8. The delay adjustment unit 42 performs delay adjustment in units of 1 / N·samples (N is an integer). One sample delay is the amount of delay in units of the sampling rate (fs) (1 / fs). For example, if fs=122.88MHz and N=64, then one sample delay is 1 / 122.88MHz = 8.14n seconds, and 1 / N sample delay is 8.14n seconds / 64 = 0.13n seconds.

[0083] In step S44, the first calculation unit 51 in the control unit 46 sets a set delay value in the delay adjustment unit 42, and then calculates the AM-AM (gain) characteristics without distortion compensation applied (step S45). The first calculation unit 51 calculates the AM-AM (gain) characteristics based on the FB signal and REF signal without distortion compensation applied by the DPD 10.

[0084] The classification unit 52 within the control unit 46 calculates the AM-AM(gain) characteristics and then classifies the calculated AM-AM(gain) characteristics into Low, High, and Peak regions according to the REF power level (step S46). Furthermore, the second calculation unit within the control unit 46 calculates the gain standard deviation value for the High region from the AM-AM(gain) characteristics (step S47).

[0085] The determination unit 54 within the control unit 46 determines whether the calculated gain standard deviation of the High region is smaller than the gain standard deviation of the least significant delay value (step S48). The least significant delay value is the delay value at which the gain standard deviation of the High region is minimized. If the calculated gain standard deviation of the High region is smaller than the gain standard deviation of the least significant delay value (step S48: Yes), the determination unit 54 performs an update process to update the least significant delay value as the first setting parameter (step S49). The update process stores and updates the least significant delay value to the least significant delay value (step S49), the gain standard deviation of the least significant delay value (step S48) to the gain standard deviation of the least significant delay value (step S49), the setting value of the least significant delay value (step S49) to the current setting value, and the gain standard deviation value to the least significant gain standard deviation value (step S49). The least significant delay value (step S49) is the delay value at which the gain standard deviation of the High region is second from the bottom.

[0086] After the control unit 46 performs an update process to update the minimum first-order delay value and the minimum second-order delay value, it increments the delay oscillation trial count by +1 (step S50) and proceeds to step S43 to determine whether the delay oscillation trial count is i=N.

[0087] If the decision unit 54 determines that the gain standard deviation of the High region is not smaller than the gain standard deviation of the minimum first-order delay value (step S48: No), it determines whether the gain standard deviation of the High region is smaller than the gain standard deviation of the minimum second-order delay value (step S51). If the decision unit 54 determines that the gain standard deviation of the High region is smaller than the gain standard deviation of the minimum second-order delay value (step S51: Yes), it performs an update process to update the minimum second-order delay value as the second-order setting parameter (step S52). The update process in step S52 updates the minimum second-order delay value to the current setting value and stores the gain standard deviation as the minimum second-order gain standard deviation. After performing the update process in step S52, the control unit 46 proceeds to the process in step S50 incrementing the delay adjustment trial count by +1. Furthermore, if the gain standard deviation in the High region is not smaller than the gain standard deviation of the second smallest delay value (step S51: No), the determination unit 54 proceeds to the process of step S50 incrementing the number of delay adjustment trials by +1.

[0088] If the number of delay adjustment trials is i=N-1 (step S43A: Yes), the determination unit 54 sets the delay value calculated by (smallest 1st digit delay value + smallest 2nd digit delay value) ÷ 2 to the delay adjustment unit 42 (step S53), and proceeds to the process in step S45.

[0089] If the number of delay adjustment trials is i=N (step S43: Yes), the determination unit 54 determines the smallest first-order delay value as the provisional delay value (step S54). The determination unit 54 sets the provisional delay value to the delay adjustment unit 42 (step S55), and terminates the processing operation shown in Figure 8.

[0090] In the provisional delay value determination process shown in Figure 8, the provisional delay value can be determined from the smallest first-order delay value and the smallest second-order delay value with the smallest gain standard deviation in the High region.

[0091] Figure 9 is a flowchart showing an example of the processing operation of the control unit 46 involved in the second phase value determination process. In Figure 9, the control unit 46 sets the number of phase swing trials i to "0" (step S61) and sets the provisional phase value as the second phase value b (step S62). For the sake of explanation, the maximum number of phase swing trials is set to N=3, but it is not limited to 3 and can be changed as appropriate. The control unit 46 determines whether the number of phase swing trials being set is i=N (step S63).

[0092] If the number of phase swing trials is not i=N (step S63: No), the control unit 46 determines whether the number of phase swing trials is i=N-1 (step S63A). If the number of phase swing trials is not i=N-1 (step S63A: No), the control unit 46 sets the phase adjustment unit 41 to a second phase value b obtained by adding the current second phase value b ± adjustment value in order to swing the phase (step S64). For example, if the number of phase swing trials is i=0, the adjustment value is set to 20°, and if the number of phase swing trials is i=1, the adjustment value is set to 10°. In other words, the phase swing is large, 20° when i=0, and small, 10° when i=1, gradually narrowing down to the optimal second phase value b.

[0093] In step S64, the first calculation unit 51 in the control unit 46 sets a second phase value b in the phase adjustment unit 41 and then calculates the AM-AM (gain) characteristics without distortion compensation applied (step S65). The first calculation unit 51 calculates the AM-AM (gain) characteristics based on the FB signal and REF signal without distortion compensation applied by the DPD 10.

[0094] The classification unit 52 within the control unit 46 calculates the AM-AM(gain) characteristics and then classifies the calculated AM-AM(gain) characteristics into Low, High, and Peak regions according to the REF power level (step S66). Furthermore, the second calculation unit 53 within the control unit 46 calculates the average gain of the Peak region from the AM-AM(gain) characteristics (step S67).

[0095] The determination unit 54 within the control unit 46 determines whether the calculated average gain of the Peak region exceeds the average gain of the second phase value with the highest value (step S68). The second phase value with the highest value (step S68) is the second phase value at which the average gain of the Peak region is maximized. If the average gain of the Peak region exceeds the average gain of the second phase value with the highest value (step S68: Yes), the determination unit 54 executes an update process to update the second phase value with the highest value (step S69) as the first setting parameter. The update process updates the second phase value with the highest value (step S69) to the second phase value with the highest value (step S69), the average gain of the second phase value with the highest value (step S69) to the second phase value with the highest value (step S69), the setting value of the second phase value with the highest value (step S69) to the current setting value, and the average gain to the average gain of the first phase value with the highest value (step S69). The second phase value with the highest value (step S69) is the second phase value with the second highest gain from the top.

[0096] Figure 10 is an explanatory diagram illustrating an example of a method for determining the maximum first-order second phase value and the maximum second-order second phase value. The average gain of the Peak region for the provisional phase value of 90° shown in Figure 10 is set to "A0". When i=0, the average gain of the Peak region for the second phase value of 70° is set to "A1", and the average gain of the Peak region for the second phase value of 110° is set to "A2". As a result, when i=0, the maximum first-order second phase value is 90°, corresponding to the average gain of the Peak region "A0", and the maximum second-order second phase value is 70°, corresponding to the average gain of the Peak region "A1". Furthermore, when i=1, the average gain of the Peak region for the second phase value of 80° is set to "A3", and the average gain of the Peak region for the second phase value of 100° is set to "A4". As a result, when i=1, the second phase value of the largest first-order value corresponds to 90°, which is the average gain value "A0" in the Peak region, and the second phase value of the largest second-order value corresponds to 80°, which is the average gain value "A3" in the Peak region.

[0097] Then, the control unit 46 performs an update process to update the second phase value of the maximum first rank and the second phase value of the maximum second rank, increments the number of phase swing trials by +1 (step S70), and proceeds to step S63 to determine whether the number of phase swing trials is i=N.

[0098] If the average gain of the Peak region does not exceed the average gain of the second phase value of the maximum first rank (step S68: No), the determination unit 54 determines whether the average gain of the Peak region exceeds the average gain of the second phase value of the maximum second rank (step S71). If the average gain of the Peak region exceeds the average gain of the second phase value of the maximum second rank (step S71: Yes), the determination unit 54 executes an update process to update the second phase value of the maximum second rank as the setting parameter for the second rank (step S72). The update process in step S72 updates the second phase value of the maximum second rank to the current setting value and stores the average gain as the average gain of the maximum second rank. After executing the update process in step S72, the control unit 46 proceeds to the process in step S70 in order to increment the number of phase swing trials by +1. Furthermore, if the average gain of the Peak region does not exceed the average gain of the second phase value of the second highest value (step S71: No), the determination unit 54 proceeds to the process of step S70 incrementing the number of phase swing trials by +1.

[0099] If the number of phase swing trials is i=N-1 (step S63A: Yes), the control unit 46 sets the second phase value calculated by (the second phase value of the largest first digit + the second phase value of the largest second digit) ÷ 2 to the phase adjustment unit 41 (step S73), and proceeds to the process in step S65. Note that when i=2, as shown in Figure 10, the second phase value b is (the second phase value of the largest first digit + the second phase value of the largest second digit) ÷ 2, that is, (90 + 80) ÷ 2 = 85°.

[0100] If the number of phase swing trials is i=N (step S63: Yes), the determination unit 54 determines the second phase value with the highest value of 1 as the optimal second phase value b (step S74). In other words, as shown in Figure 10, the determination unit 54 determines the phase value of 90° with the highest value of 1 in the Peak region as the optimal second phase value b.

[0101] Then, the determination unit 54 sets the optimal second phase value b to the phase adjustment unit 41 (step S75), and the processing operation shown in Figure 9 is completed.

[0102] In the second phase value determination process shown in Figure 9, the optimal second phase value b can be determined from the second phase value with the largest maximum value in the Peak region (ranked 1st) and the second phase value with the largest maximum value (ranked 2nd).

[0103] Figure 11 is a flowchart showing an example of the processing operation of the control unit 46 involved in the first phase value determination process. In Figure 11, the control unit 46 sets the number of phase swing trials i to "0" (step S81) and sets the provisional phase value as the first phase value a (step S82). For the sake of explanation, the maximum number of phase swing trials is set to N=3, but it is not limited to 3 and can be changed as appropriate. The control unit 46 determines whether the number of phase swing trials being set is i=N (step S83).

[0104] If the number of phase swing trials is not i=N (step S83: No), the control unit 46 determines whether the number of phase swing trials is i=N-1 (step S83A). If the number of phase swing trials is not i=N-1 (step S83A: No), the control unit 46 sets the phase adjustment unit 41 to a first phase value a obtained by adding the current first phase value a ± adjustment value in order to swing the phase (step S84). For example, if the number of phase swing trials is i=0, the adjustment value is set to 20°, and if the number of phase swing trials is i=1, the adjustment value is set to 10°. In other words, the phase swing is large, 20° when i=0, and small, 10° when i=1, gradually narrowing down to the optimal first phase value a.

[0105] In step S84, the first calculation unit 51 in the control unit 46 sets a first phase value a in the phase adjustment unit 41, and then calculates the AM-AM (gain) characteristics in a state where distortion compensation is not applied (step S85). The first calculation unit 51 calculates the AM-AM (gain) characteristics based on the FB signal and REF signal in a state where distortion compensation by the DPD 10 is not applied.

[0106] The classification unit 52 within the control unit 46 calculates the AM-AM(gain) characteristics and then classifies the calculated AM-AM(gain) characteristics into Low, High, and Peak regions according to the REF power level (step S86). Furthermore, the second calculation unit within the control unit 46 calculates the average gain of the Peak region from the AM-AM(gain) characteristics (step S87).

[0107] The determination unit 54 within the control unit 46 determines whether the calculated average gain of the Peak region exceeds the average gain of the first phase value of the maximum first rank (step S88). The first phase value of the maximum first rank is the first phase value at which the average gain of the Peak region is maximized. If the average gain of the Peak region exceeds the average gain of the first phase value of the maximum first rank (step S88: Yes), the determination unit 54 executes an update process to update the first phase value of the maximum first rank as the setting parameter of the first rank (step S89). The update process stores and updates the first phase value of the maximum first rank to the first phase value of the maximum second rank, the average gain of the first phase value of the maximum first rank to the average gain of the first phase value of the maximum second rank, the setting value of the first phase value of the maximum first rank to the current setting value, and the average gain to the average gain of the maximum first rank. The first phase value of the maximum second rank is the first phase value at which the average gain of the Peak region is second from the top.

[0108] After the control unit 46 performs an update process to update the first phase value of the maximum first rank and the first phase value of the maximum second rank, it increments the number of phase swing trials by +1 (step S90) and proceeds to step S83 to determine whether the number of phase swing trials is i=N.

[0109] If the average gain of the Peak region does not exceed the average gain of the first phase value of the maximum first rank (step S88: No), the determination unit 54 determines whether the average gain of the Peak region exceeds the average gain of the first phase value of the maximum second rank (step S91). If the average gain of the Peak region exceeds the average gain of the first phase value of the maximum second rank (step S91: Yes), the determination unit 54 executes an update process to update the first phase value of the maximum second rank as the setting parameter of the second rank (step S92). The update process in step S92 updates the first phase value of the maximum second rank to the current setting value and stores the average gain as the average gain of the maximum second rank. After executing the update process in step S92, the control unit 46 proceeds to the process in step S90 in order to increment the number of phase swing trials by +1. Furthermore, if the average gain of the Peak region does not exceed the average gain of the first phase value of the second highest value (step S91: No), the determination unit 54 proceeds to the process of step S90 incrementing the number of phase swing trials by +1.

[0110] If the number of phase swing trials is i=N-1 (step S83A: Yes), the control unit 46 sets the first phase value calculated by (maximum first-order first phase value + maximum second-order first phase value) ÷ 2 to the phase adjustment unit 41 (step S93) and proceeds to the process in step S85. Then, if the number of phase swing trials is i=N (step S83: Yes), the determination unit 54 determines the maximum first-order first phase value as the optimal first phase value a (step S94). Then, the determination unit 54 sets the optimal first phase value a to the phase adjustment unit 41 (step S95) and terminates the processing operation shown in Figure 11.

[0111] In the first phase value determination process shown in Figure 11, the optimal first phase value a can be determined from the first and second largest first phase values ​​with the highest average gain in the Peak region.

[0112] Figure 12 is a flowchart showing an example of the processing operation of the control unit 46 involved in the first delay value determination process. In Figure 12, the control unit 46 sets the number of delay adjustment trials i to "0" (step S101) and sets the provisional delay value as the first delay value c (step S102). For the sake of explanation, the maximum number of delay adjustment trials is set to N=5, but it is not limited to 5 and can be changed as appropriate. Let α be 10. The control unit 46 determines whether the number of delay adjustment trials being set is i=N (step S103).

[0113] If the number of delay-swing trials is not i=N (step S103: No), the control unit 46 determines whether the number of delay-swing trials is i=N-1 (step S103A). If the number of delay-swing trials is not i=N-1 (step S103A: No), the control unit 46 sets a first delay value c, obtained by adding the current first delay value ± adjustment value, to the delay adjustment unit 42 in order to swing the delay (step S104). If the number of delay-swing trials is i=0, for example, if the adjustment value is 32, the first delay value c to be set will be ±32. If the number of delay-swing trials is i=1, for example, if the current first delay value, the smallest first-order delay value c1, is 32, the first delay value c to be set will be 64, 96, 128, and if the current first delay value c is -32, the first delay value to be set will be -128, -96, -64. Furthermore, when the number of delay adjustment trials is i=2, the first delay value c1, which is the minimum first-order delay value in the current setting, and the adjustment value are set to ±16. When the number of delay adjustment trials is i=3, the first delay value c1, which is the minimum first-order delay value in the current setting, and the adjustment value are set to ±8. The delay adjustment unit 42 performs delay adjustment in units of 1 / N·samples (N is an integer). One sample delay is the amount of delay in units of the sampling rate (fs) (1 / fs). For example, if fs=122.88MHz and N=64, then one sample delay is 1 / 122.88MHz = 8.14n seconds, and 1 / N sample delay is 8.14n seconds / 64 = 0.13n seconds.

[0114] In step S104, the first calculation unit 51 in the control unit 46 sets a first delay value c in the delay adjustment unit 42, and then calculates the AM-AM (gain) characteristics without distortion compensation applied (step S105). The first calculation unit 51 calculates the AM-AM (gain) characteristics based on the FB signal and REF signal without distortion compensation applied by the DPD 10.

[0115] The classification unit 52 within the control unit 46 calculates the AM-AM(gain) characteristics and then classifies the calculated AM-AM(gain) characteristics into Low, High, and Peak regions according to the REF power level (step S106). Furthermore, the second calculation unit 53 within the control unit 46 calculates the average gain value of the High region from the AM-AM(gain) characteristics (step S107).

[0116] Figure 13 is an explanatory diagram illustrating an example of a method for determining the first maximum delay value and the second maximum first delay value. The average gain in the High region for the provisional delay value "11" shown in Figure 13 is assumed to be "A0". When i=0, the average gain in the High region for the first delay value "-32" is assumed to be "A1", and the average gain in the High region for the first delay value "32" is assumed to be "A2". As a result, when i=0, the first maximum first delay value is "-32", which corresponds to the average gain in the High region "A1", and the second maximum first delay value is "11", which corresponds to the average gain in the High region "A0".

[0117] Furthermore, when i=1, the average gain in the High region for the first delay value "-128" is "A3", the average gain in the High region for the first delay value "-96" is "A4", and the average gain in the High region for the first delay value "-64" is "A5". As a result, when i=1, the largest first-rank first delay value corresponds to "-32", which is the average gain in the High region "A1", and the largest second-rank first delay value corresponds to "11", which is the average gain in the High region "A0".

[0118] Furthermore, when i=2, the average gain in the High region for the first delay value "-48" is "A6", and the average gain in the High region for the first delay value "-16" is "A7". As a result, when i=2, the largest first-rank first delay value is "-16", which corresponds to the average gain in the High region "A7", and the largest second-rank first delay value is "-32", which corresponds to the average gain in the High region "A1".

[0119] Furthermore, when i=3, the average gain in the High region for the first delay value "-24" is "A8", and the average gain in the High region for the first delay value "-8" is "A9". As a result, when i=3, the first largest delay value corresponds to "-16", which is the average gain in the High region "A7", and the second largest first delay value corresponds to "-32", which is the average gain in the High region "A1".

[0120] The determination unit 54 within the control unit 46 determines whether the calculated average gain of the High region exceeds the average gain of the first delay value with the highest value (step S108). The first delay value with the highest value is the first delay value at which the average gain of the High region is maximized. If the average gain of the High region exceeds the average gain of the first delay value with the highest value (step S108: Yes), the determination unit 54 executes an update process to update the first delay value with the highest value as the setting parameter with the highest value (step S109). The update process updates the first delay value with the highest value (maximum 1st) to the first delay value with the highest value (maximum 2nd), the average gain of the first delay value with the highest value (maximum 1st) to the average gain of the first delay value with the highest value (maximum 2nd), the setting value of the first delay value with the highest value (maximum 1st) to the current setting value, and the average gain to the average gain of the first delay value with the highest value (maximum 1st). The first delay value with the highest value (maximum 2nd) is the first delay value with the second highest average gain in the High region.

[0121] The control unit 46 performs an update process to update the first delay value of the maximum first rank and the first delay value of the maximum second rank, then increments the delay swing trial count by +1 (step S110), and proceeds to step S103 to determine whether the delay swing trial count is i=N.

[0122] If the average gain in the High region does not exceed the average gain of the first delay value of the maximum first rank (step S108: No), the determination unit 54 determines whether the average gain in the High region exceeds the average gain of the first delay value of the maximum second rank (step S111). If the average gain in the High region exceeds the average gain of the first delay value of the maximum second rank (step S111: Yes), the determination unit 54 executes an update process to update the first delay value of the maximum second rank as the setting parameter of the second rank (step S112). The update process in step S112 updates the first delay value of the maximum second rank to the current setting value and stores the average gain as the average gain of the maximum second rank. After executing the update process in step S112, the control unit 46 proceeds to the process in step S110 in order to increment the delay adjustment trial count by +1. Furthermore, if the average gain in the High region does not exceed the average gain of the first delay value which is the second highest (step S111: No), the determination unit 54 proceeds to the process of step S110 incrementing the number of delay adjustment trials by +1.

[0123] If the number of delay adjustment trials is i=N-1 (step S103A: Yes), the control unit 46 sets the first delay value calculated by (maximum first-order first delay value + maximum second-order first delay value) ÷ 2 to the delay adjustment unit 42 (step S113), and proceeds to the process in step S105. For example, if i=4, the provisional first delay value is (maximum first-order first delay value + maximum second-order first delay value) ÷ 2, as shown in Figure 13, that is, (-16-32) ÷ 2 = -24.

[0124] If the number of delay adjustment trials is i=N (step S103: Yes), the determination unit 54 determines the first maximum delay value c as the optimal first delay value (step S114). As a result, when i=4, the gain average value "A7" is determined to be "-16", which is the first maximum delay value, and A7 is determined to be the optimal first delay value c. The determination unit 54 sets the optimal first delay value c to the delay adjustment unit 42 (step S115), and terminates the processing operation shown in Figure 12.

[0125] In the first delay value determination process shown in Figure 12, the optimal first delay value c can be determined from the first largest and second largest first delay values ​​with the highest average gain in the High region.

[0126] Figure 14 is a flowchart showing an example of the processing operation of the control unit 46 involved in the second delay value determination process. In Figure 14, the control unit 46 sets the number of delay adjustment trials i to "0" (step S121) and sets the provisional delay value as the second delay value (d=c+α) (step S122). For the sake of explanation, the maximum number of delay adjustment trials is set to N=5, but it is not limited to 5 and can be changed as appropriate. α is set to 10. The control unit 46 determines whether the number of delay adjustment trials being set is i=N (step S123).

[0127] If the number of delay-swing trials is not i=N (step S123: No), the control unit 46 determines whether the number of delay-swing trials is i=N-1 (step S123A). If the number of delay-swing trials is not i=N-1 (step S123A: No), the control unit 46 sets a second delay value d obtained by adding the current second delay value ± adjustment value to the delay adjustment unit 42 in order to swing the delay (step S124). If the number of delay-swing trials is i=0, for example, if the adjustment value is 32, the set second delay value d will be ±32. If the number of delay-swing trials is i=1, for example, if the current second delay value, the smallest first-order second delay value d1, is 32, the set second delay value d will be 64, 96, 128, and if the current second delay value d is -32, the set second delay value will be -128, -96, -64. Furthermore, if the number of delay adjustment trials is i=2, the second delay value d1, which is the minimum first-order second delay value in the current settings, and the adjustment value are set to ±16. If the number of delay adjustment trials is i=3, the second delay value d1, which is the minimum first-order second delay value in the current settings, and the adjustment value are set to ±8.

[0128] In step S124, the first calculation unit 51 in the control unit 46 sets a second delay value d in the delay adjustment unit 42, and then calculates the AM-AM (gain) characteristics without distortion compensation applied (step S125). The first calculation unit 51 calculates the AM-AM (gain) characteristics based on the FB signal and REF signal without distortion compensation applied by the DPD 10.

[0129] The classification unit 52 within the control unit 46 calculates the AM-AM(gain) characteristics and then classifies the calculated AM-AM(gain) characteristics into Low, High, and Peak regions according to the REF power level (step S126). Furthermore, the second calculation unit 53 within the control unit 46 calculates the average gain of the Peak region from the AM-AM(gain) characteristics (step S127).

[0130] The determination unit 54 within the control unit 46 determines whether the calculated average gain of the Peak region exceeds the average gain of the second delay value with the highest value (step S128). The second delay value with the highest value (step S128) is the second delay value at which the average gain of the Peak region is maximized. If the average gain of the Peak region exceeds the average gain of the second delay value with the highest value (step S128: Yes), the determination unit 54 executes an update process to update the second delay value with the highest value (step S129) as the first setting parameter. The update process updates the second delay value with the highest value (step S128) to the second delay value with the highest value (step S129), the average gain of the second delay value with the highest value (step S128) to the average gain of the second delay value with the highest value (step S128), the setting value of the second delay value with the highest value (step S128) to the current setting value, and the average gain to the average gain of the first value with the highest value (step S128). Furthermore, the second largest delay value is the second largest delay value among the peak region's gain average values.

[0131] After the control unit 46 performs an update process to update the second delay value of the first largest and the second largest second largest, it increments the delay swing trial count by +1 (step S130) and proceeds to step S123 to determine whether the delay swing trial count is i=N.

[0132] If the average gain of the Peak region does not exceed the average gain of the second delay value of the maximum first rank (step S128: No), the determination unit 54 determines whether the average gain of the Peak region exceeds the average gain of the second delay value of the maximum second rank (step S131). If the average gain of the Peak region exceeds the average gain of the second delay value of the maximum second rank (step S131: Yes), the determination unit 54 executes an update process to update the second delay value of the maximum second rank as the second-rank setting parameter (step S132). The update process in step S132 updates the second delay value of the maximum second rank to the current setting value and stores the average gain as the average gain of the maximum second rank. After executing the update process in step S132, the control unit 46 proceeds to the process in step S130 in order to increment the delay adjustment trial count by +1. Furthermore, if the average gain of the Peak region does not exceed the average gain of the second delay value of the second highest value (step S131: No), the determination unit 54 proceeds to the process of step S130 incrementing the number of delay adjustment trials by +1.

[0133] If the number of delay adjustment trials is i=N-1 (step S123A:Yes), the control unit 46 sets the second delay value calculated by (maximum first-order second delay value + maximum second-order second delay value) ÷ 2 to the delay adjustment unit 42 (step S133), and proceeds to the process in step S125. Then, if the number of delay adjustment trials is i=N (step S123:Yes), the determination unit 54 determines the maximum first-order second delay value as the optimal second delay value d (step S134). The determination unit 54 sets the optimal second delay value d to the delay adjustment unit 42 (step S135), and the processing operation shown in Figure 14 is completed.

[0134] In the second delay value determination process shown in Figure 14, the optimal second delay value d can be determined from the second delay value with the largest maximum value (1st place) and the second delay value with the largest maximum value (2nd place) in the Peak region.

[0135] The Doherty amplifier 1 of Embodiment 1 has an adjustment unit 26 positioned between the distribution unit 21 and PA23C, which adjusts the phase value and delay value of the second signal from the distribution unit 21. The adjustment unit 26 calculates the AM-AM (gain) characteristics based on the FB signal and the REF signal. The adjustment unit 26 classifies the calculated AM-AM (gain) characteristics into multiple regions according to the level of REF power. The adjustment unit 26 calculates statistical gain values ​​for each classified region and sets the phase value and delay value that provides the best wireless characteristics based on the calculated statistical values ​​for each region. As a result, the time required to adjust the phase difference and delay difference between the paths of CA22C and PA23C in the two-input Doherty amplifier 20 to achieve the best distortion characteristics can be significantly reduced. For example, while conventional manual adjustment took about 1 hour, in this embodiment, automatic adjustment can be performed in about 2 minutes.

[0136] Furthermore, the Doherty amplifier 1 uses the FB and REF signals within the device for adjustment, eliminating the need for conventional external measuring devices. For example, while external measuring devices monitor the average power output of a power amplifier, the Doherty amplifier 1 monitors the FB and REF signals on a sample-by-sample basis, enabling highly accurate correction. Moreover, the Doherty amplifier 1 uses phase and delay values ​​for adjustment, further enhancing its ability to perform highly accurate correction.

[0137] The first calculation unit 51 sets predetermined phase and delay values ​​in the adjustment unit 26 without applying distortion compensation, and then calculates the AM-AM (gain) characteristics based on the FB signal and REF signal. As a result, the AM-AM (gain) characteristics of the DPD 10 without distortion compensation can be obtained.

[0138] The adjustment unit 26 includes a phase adjustment unit 41 that adjusts the phase value of the second signal, and a delay adjustment unit 42 connected in series with the phase value that adjusts the delay value of the second signal. The adjustment unit 26 sets a set phase value in the phase adjustment unit 41 that maximizes the average gain in the Peak region, and sets a set delay value in the delay adjustment unit 42 that maximizes the average gain in the High region. By reducing the decrease in gain in the Peak region, linearity can be improved. Furthermore, by reducing the variation in gain in the High and Low regions, the memory effect of the amplifier is reduced. As a result, the wireless characteristics are improved. In other words, the adjustment unit 26 adjusts the phase value and delay value of PA23C to maximize the gain near the saturation power after Doherty synthesis and eliminate gain-phase variation. As a result, the wireless characteristics can be optimized.

[0139] Furthermore, the example given shows how to determine the provisional phase value, the optimal first phase value, the optimal second phase value, and the optimal second delay value using the average gain in the Peak region, the optimal first delay value using the average gain in the High region, and the provisional delay value using the standard deviation of the gain in the High region. However, the target region and statistical values ​​should be changed according to the amplifier characteristics used, and are not limited to this.

[0140] In the adjustment section 26 within the Doherty amplifier 1 of Example 1, the case in which the phase adjustment section 41 and the delay adjustment section 42 are arranged in series in the second path where the PA23C is located was illustrated. However, in addition to the phase adjustment section 41 and the delay adjustment section 42, an amplitude adjustment section 47 may also be arranged in series, and such an embodiment will be described below as Example 2. [Examples]

[0141] Figure 15 is an explanatory diagram showing an example of the Doherty amplifier 1A of Embodiment 2. Note that components identical to those in the Doherty amplifier 1 of Embodiment 1 are denoted by the same reference numerals, and explanations of their overlapping components and operations are omitted. The difference between the Doherty amplifier 1A shown in Figure 15 and the Doherty amplifier 1 shown in Figure 1 lies in the configuration of the adjustment unit 26A.

[0142] Figure 16 is an explanatory diagram showing an example of the adjustment unit 26A of Embodiment 2. The adjustment unit 26A shown in Figure 16 includes a phase adjustment unit 41, a delay adjustment unit 42, a phase table 44, a delay table 45, a power address calculation unit 43, and a control unit 46. Furthermore, the adjustment unit 26A includes an amplitude adjustment unit 47 located downstream of the delay adjustment unit 42 in the second path where the PA23C is placed, and an amplitude table 48 that stores multiple amplitude patterns.

[0143] Figure 17 is an explanatory diagram showing an example of an amplitude table 48. The amplitude table 48 shown in Figure 17 pre-manages the amplitude pattern of the amplitude adjustment unit 47 according to the REF power that provides the best wireless characteristics. The amplitude adjustment unit 47 sequentially sets multiple amplitude patterns in the amplitude table 48. The amplitude pattern is a characteristic amplitude pattern according to the REF power that provides good wireless characteristics. For the sake of explanation, it is assumed that the amplitude table 48 shown in Figure 17 manages three types of amplitude patterns. The determination unit 54 sets the amplitude adjustment unit 47 to the setting amplitude pattern that minimizes the gain standard deviation value, which is a statistical value for each setting amplitude pattern in the High region.

[0144] Figure 18 is a flowchart showing an example of the processing operation of the control unit 46 within the adjustment unit 26A involved in the second optimal value determination process. In Figure 18, the control unit 46 performs a provisional delay determination process in step S12 to determine a provisional delay value, and then performs the amplitude pattern determination process shown in Figure 19 to determine the optimal amplitude pattern (step S140). Furthermore, after determining the optimal amplitude pattern in step S140, the control unit 46 performs a second phase value determination process in step S13.

[0145] In the second optimal value determination process, after determining the optimal amplitude pattern, the optimal second phase value b, the optimal first phase value a, the optimal first delay value c, and the optimal second delay value d are determined sequentially.

[0146] Figure 19 is a flowchart showing an example of the processing operation of the control unit 46 involved in the amplitude pattern determination process. In Figure 19, the control unit 46 sets the number of amplitude pattern selection trials i to "0" (step S141) and sets the provisional phase value to the phase adjustment unit 41 (step S142). Furthermore, the control unit 46 sets the provisional delay value to the delay adjustment unit 42 (step S143). The control unit 46 determines whether the number of amplitude pattern selection trials i = N or not (step S144). For the sake of explanation, the maximum number of amplitude pattern selection trials that can be selected in the amplitude table 48 is set to N = 3, but it is not limited to 3 and can be changed as appropriate.

[0147] If the number of amplitude pattern selection trials is not i=N (step S144: No), the control unit 46 sets an unselected amplitude pattern from the amplitude table 48 to the amplitude adjustment unit 47 (step S145). After setting an unselected amplitude pattern to the amplitude adjustment unit 47 in step S145, the first calculation unit 51 in the control unit 46 calculates the AM-AM (gain) characteristics without distortion compensation applied (step S146). The first calculation unit 51 calculates the AM-AM (gain) characteristics based on the FB signal and REF signal without distortion compensation applied by the DPD 10.

[0148] The classification unit 52 within the control unit 46 calculates the AM-AM(gain) characteristics and then classifies the calculated AM-AM(gain) characteristics into Low, High, and Peak regions according to the REF power level (step S147). Furthermore, the second calculation unit 53 within the control unit 46 calculates the gain standard deviation value for the High region from the AM-AM(gain) characteristics (step S148).

[0149] The determination unit 54 in the control unit 46 determines whether the calculated gain standard deviation value in the High region is smaller than the gain standard deviation value of the smallest first-order amplitude pattern (step S149). The smallest first-order amplitude pattern is the amplitude pattern that has the smallest gain standard deviation value in the High region. If the calculated gain standard deviation value in the High region is smaller than the gain standard deviation value of the smallest first-order amplitude pattern (step S149: Yes), the determination unit 54 performs an update process to update the smallest first-order amplitude pattern as the first-order setting pattern (step S150). The update process in step S150 updates the setting value of the smallest first-order amplitude pattern to the current setting value and stores the gain standard deviation value as the smallest first-order gain standard deviation value.

[0150] After executing the update process in step S150, the control unit 46 increments the number of amplitude pattern selection trials by +1 (step S151) and proceeds to step S144 to determine whether the number of amplitude pattern selection trials is i=N. If the determination unit 54 determines that the calculated gain standard deviation value in the High region is not smaller than the gain standard deviation value of the smallest first amplitude pattern (step S149: No), it proceeds to the process in step S151 to increment the number of amplitude pattern selection trials by +1.

[0151] If the number of amplitude pattern selection trials is i=N (step S144: Yes), the control unit 46 determines the smallest first-order amplitude pattern as the optimal amplitude pattern (step S152). Furthermore, the determination unit 54 sets the determined optimal amplitude pattern to the amplitude adjustment unit 47 (step S153), and the processing operation shown in Figure 19 is completed.

[0152] The amplitude pattern determination process shown in Figure 19 can determine the optimal amplitude pattern that minimizes the gain standard deviation in the High region.

[0153] The Doherty amplifier 1A of Embodiment 2 further includes an amplitude adjustment unit 47 connected in series with the phase adjustment unit 41 or the delay adjustment unit 42 to adjust the amplitude pattern of the second signal. The Doherty amplifier 1A sets the amplitude adjustment unit 47 to the setting amplitude pattern that minimizes the gain standard deviation in the high region. As a result, when making adjustments, the amplitude pattern is used in addition to the phase value and delay value, enabling high-precision adjustment that takes the amplitude pattern into account. Note that the adjustment of the amplitude pattern is not limited to changing the target region and statistical values ​​according to the amplifier characteristics.

[0154] In the adjustment section 26 within the Doherty amplifier 1 of Embodiment 1, the case in which the phase adjustment section 41 and the delay adjustment section 42 are arranged in series in the second path where the PA23C is located was illustrated. However, in addition to the phase adjustment section 41 and the delay adjustment section 42, an amplitude frequency adjustment section 47A may also be arranged in series, and such an embodiment will be described below as Embodiment 3. [Examples]

[0155] Figure 20 is an explanatory diagram showing an example of the Doherty amplifier 1B of Embodiment 3. Note that components identical to those in the Doherty amplifier 1 of Embodiment 1 are denoted by the same reference numerals, and the explanation of their overlapping components and operations is omitted. The difference between the Doherty amplifier 1B shown in Figure 20 and the Doherty amplifier 1 shown in Figure 1 lies in the configuration of the adjustment unit 26B.

[0156] Figure 21 is an explanatory diagram showing an example of the adjustment unit 26B of Embodiment 3. The adjustment unit 26B shown in Figure 21 includes a phase adjustment unit 41, a delay adjustment unit 42, a phase table 44, a delay table 45, a power address calculation unit 43, and a control unit 46. Furthermore, the adjustment unit 26B includes an amplitude frequency adjustment unit 47A located downstream of the delay adjustment unit 42 in the second path in which the PA23C is located.

[0157] Figure 22 is an explanatory diagram showing an example of an amplitude-frequency correction pattern. The amplitude-frequency correction pattern shown in Figure 22 is the amplitude-frequency correction pattern of the amplitude-frequency adjustment unit 47A according to the REF power that provides the best wireless characteristics. The vertical axis of the amplitude-frequency correction pattern is amplitude, and the horizontal axis is normalized frequency (normalized by the difference between frequency and center frequency (fc) by fc). For the sake of explanation, the amplitude-frequency correction pattern shown in Figure 22 is three types of amplitude-frequency correction patterns. The amplitude-frequency adjustment unit 47A sequentially sets multiple amplitude-frequency correction patterns. The determination unit 54 sets the amplitude-frequency adjustment unit 47A to the setting amplitude-frequency correction pattern that minimizes the gain standard deviation value in the High region.

[0158] Figure 23 is a flowchart showing an example of the processing operation of the control unit 46 in the adjustment unit involved in the third optimal value determination process. In Figure 23, the control unit 46 performs a provisional delay determination process in step S12 to determine a provisional delay value, and then performs the amplitude frequency correction pattern determination process shown in Figure 24 to determine the optimal amplitude frequency correction pattern (step S160). Furthermore, after determining the optimal amplitude frequency correction pattern in step S160, the control unit 46 performs a second phase value determination process in step S13.

[0159] In the third optimal value determination process, after determining the optimal amplitude-frequency correction pattern, the optimal second phase value b, the optimal first phase value a, the optimal first delay value c, and the optimal second delay value d are determined sequentially.

[0160] Figure 24 is a flowchart showing an example of the processing operation of the control unit 46 involved in the amplitude-frequency correction pattern determination process. In Figure 24, the control unit 46 sets the number of amplitude-frequency correction pattern selection trials i to "0" (step S161) and sets the provisional phase value to the phase adjustment unit 41 (step S162). Furthermore, the control unit 46 sets the provisional delay value to the delay adjustment unit 42 (step S163). The control unit 46 determines whether the number of amplitude-frequency correction pattern selection trials i = N (step S164). For the sake of explanation, the maximum number of selectable amplitude-frequency correction pattern selection trials is set to 3, but it is not limited to 3 and can be changed as appropriate.

[0161] If the number of attempts to select the amplitude frequency correction pattern is not i=2 (step S164: No), the control unit 46 sets the unselected amplitude frequency correction pattern to the amplitude frequency adjustment unit 47A (step S165). After setting the unselected amplitude frequency correction pattern to the amplitude frequency adjustment unit 47A in step S165, the first calculation unit 51 in the control unit 46 calculates the AM-AM (gain) characteristics in the state without distortion compensation applied (step S166). The first calculation unit 51 calculates the AM-AM (gain) characteristics based on the FB signal and REF signal in the state without distortion compensation applied by the DPD 10.

[0162] The classification unit 52 within the control unit 46 calculates the AM-AM(gain) characteristics and then classifies the calculated AM-AM(gain) characteristics into Low, High, and Peak regions according to the REF power (step S167). Furthermore, the second calculation unit within the control unit 46 calculates the gain standard deviation value of the High region from the AM-AM(gain) characteristics (step S168).

[0163] The determination unit 54 in the control unit 46 determines whether the calculated gain standard deviation value in the High region is smaller than the gain standard deviation value of the minimum first-order amplitude frequency correction pattern (step S169). The minimum first-order amplitude frequency correction pattern is the amplitude frequency correction pattern that minimizes the gain standard deviation value in the High region. If the calculated gain standard deviation value in the High region is smaller than the gain standard deviation value of the minimum first-order amplitude frequency correction pattern (step S169: Yes), the determination unit 54 performs an update process to update the minimum first-order amplitude frequency correction pattern as the first-order setting pattern (step S170). The update process in step S170 updates the setting value of the minimum first-order amplitude frequency correction pattern to the current setting value and stores the gain standard deviation value as the minimum first-order gain standard deviation value.

[0164] After executing the update process in step S170, the control unit 46 increments the number of amplitude frequency correction pattern selection trials by +1 (step S171) and proceeds to step S164 to determine whether the number of amplitude frequency correction pattern selection trials is i=N. If the calculated gain standard deviation value in the High region is not smaller than the gain standard deviation value of the smallest first amplitude frequency correction pattern (step S169: No), the determination unit 54 proceeds to the process in step S171 to increment the number of amplitude frequency correction pattern selection trials by +1.

[0165] If the number of attempts to select an amplitude-frequency correction pattern is i=N (step S164: Yes), the control unit 46 determines the smallest first-order amplitude-frequency correction pattern as the optimal amplitude-frequency correction pattern (step S172). The determination unit 54 sets the determined amplitude-frequency correction pattern to the amplitude-frequency adjustment unit 47A (step S173), and terminates the processing operation shown in Figure 24.

[0166] The amplitude-frequency correction pattern determination process shown in Figure 24 can determine the optimal amplitude-frequency correction pattern that minimizes the gain standard deviation in the High region.

[0167] The Doherty amplifier 1B of Embodiment 3 further includes an amplitude frequency adjustment unit 47A connected in series with the phase adjustment unit 41 or the delay adjustment unit 42 to adjust the amplitude frequency correction pattern of the second signal. The Doherty amplifier 1B sets the amplitude frequency adjustment unit 47A to a setting amplitude frequency correction pattern that minimizes the gain standard deviation in the high region. As a result, when making adjustments, the amplitude frequency correction pattern is used in addition to the phase value and delay value, enabling highly accurate adjustment that takes the amplitude frequency correction pattern into consideration. Note that the adjustment of the amplitude frequency correction pattern is not limited to changing the target region and statistical values ​​according to the amplifier characteristics.

[0168] In the Doherty amplifier 1 of Example 1, when determining the provisional phase value, the first phase value, and the second phase value, the case where the phase value with the maximum first-order average gain and the phase value with the maximum second-order average gain are used is illustrated. Furthermore, in the Doherty amplifier 1, when determining the first delay value and the second delay value, the case where the delay value with the maximum first-order average gain and the delay value with the maximum second-order average gain are used is illustrated. Furthermore, in the Doherty amplifier 1, when determining the provisional delay value, the case where the delay value with the minimum first-order standard deviation of the gain and the delay value with the minimum second-order standard deviation are used is illustrated. However, it is not necessary to use the second-order phase value or the second-order delay value, and such an embodiment will be described below as Example 4. Note that components identical to those in the Doherty amplifier 1 of Example 1 are denoted by the same reference numerals, and the description of the redundant components and operations will be omitted. [Examples]

[0169] The determination unit 54 performs the provisional phase value determination process shown in Figure 25 instead of Figure 7, and the provisional delay value determination process shown in Figure 26 instead of Figure 8.

[0170] Figure 25 is a flowchart showing an example of the processing operation of the control unit 46 involved in the provisional phase value determination process in Embodiment 4. In Figure 25, the control unit 46 determines in step S23 whether or not the number of phase swing trials being set is i=N.

[0171] If the number of phase adjustment trials is not i=N (step S23: No), the control unit 46 executes the process of step S24, which sets a first phase value a obtained by adding the current first phase value a ± adjustment value to the phase adjustment unit 41 in order to adjust the phase.

[0172] Furthermore, in step S27, the second calculation unit in the control unit 46 calculates the average gain of the Peak region from the AM-AM (gain) characteristics and then determines whether the average gain exceeds the average gain of the maximum first-order phase value (step S28A). If the average gain exceeds the average gain of the maximum first-order phase value (step S28A: Yes), the determination unit 54 performs an update process to update the maximum first-order phase value as a setting parameter (step S29A). The update process in step S29A updates the setting value of the maximum first-order phase value to the current setting value and stores the average gain as the average gain of the maximum first-order.

[0173] After the control unit 46 performs an update process to update the maximum first-order phase value, it proceeds to the process of step S30 in order to increment the number of phase swing trials by +1. Also, if the average gain value does not exceed the average gain value of the maximum first-order phase value (step S28A: No), the determination unit 54 proceeds to the process of step S30 in order to increment the number of phase swing trials by +1.

[0174] If the number of phase swing trials is i=N (step S23: Yes), the control unit 46 proceeds to the process in step S34 to determine the first phase value of the largest first magnitude as the provisional phase value.

[0175] In the provisional phase value determination process shown in Figure 25, the provisional phase value can be determined using the maximum first-order phase value that maximizes the average gain in the Peak region. Furthermore, in the first and second phase value determination processes, the optimal first and second phase values ​​may also be determined using only the maximum first-order phase value.

[0176] Figure 26 is a flowchart showing an example of the processing operation of the control unit 46 involved in the provisional delay value determination process in Embodiment 4. In Figure 26, if the number of delay adjustment trials is not i=N in step S43, the control unit 46 sets the set delay value obtained by adding the current set delay value ± adjustment value to the delay adjustment unit 42 in step S44.

[0177] The second calculation unit 53 calculates the gain standard deviation value in the High region from the AM-AM (gain) characteristics in step S47, and then determines whether the gain standard deviation value is smaller than the gain standard deviation value of the minimum first-order delay value (step S48A). If the calculated gain standard deviation value in the High region is smaller than the gain standard deviation value of the minimum first-order delay value (step S48A: Yes), the determination unit 54 performs an update process to update the minimum first-order delay value as a setting parameter (step S49A). The update process in step S49A updates the setting value of the minimum first-order delay value to the current setting value and stores the gain standard deviation value as the minimum first-order gain standard deviation value.

[0178] After the control unit 46 performs an update process to update the minimum first-order delay value, it proceeds to the process of step S50 to increment the delay swing trial count by +1. Also, if the determination unit 54 determines that the gain standard deviation is not smaller than the gain mean value of the minimum first-order delay value (step S48A: No), it proceeds to the process of step S50 to increment the delay swing trial count by +1.

[0179] If the number of delay adjustment trials in step S43 is i=N, the control unit 46 proceeds to the process in step S54 to determine the smallest first-order delay value as the provisional delay value.

[0180] In the provisional delay value determination process shown in Figure 26, the provisional delay value can be determined using the smallest first-order delay value that minimizes the gain standard deviation in the High region. Furthermore, in the first and second delay value determination processes, the optimal first and second delay values ​​may also be determined using only the first-order delay value.

[0181] In the Doherty amplifier 1 of Example 1, a two-input Doherty amplifier 20 is employed, and therefore, the case in which the adjustment unit 26 is placed in the second path where PA23C is located is illustrated. However, if a two-input inverse Doherty amplifier is used instead of the two-input Doherty amplifier 20, the adjustment unit 26 may be placed in the first path where CA22C is located. Also, in the Doherty amplifier 1 of Example 1, the case in which the adjustment unit 26 is placed in the first path where PA23C is located is illustrated, but the adjustment unit 26 may be placed in both paths where CA22C and PA23C are located, and this can be changed as appropriate.

[0182] Furthermore, while the adjustment unit 26A in the Doherty amplifier 1A of Example 2 is exemplified as having a phase adjustment unit 41, a delay adjustment unit 42, and an amplitude adjustment unit 47, an amplitude-frequency adjustment unit 47A may also be connected in series with the phase adjustment unit 41, the delay adjustment unit 42, and the amplitude adjustment unit 47. As a result, when making adjustments, an amplitude-frequency correction pattern is used in addition to the phase value, delay value, and amplitude pattern, enabling high-precision adjustment that takes into account the amplitude pattern and amplitude-frequency pattern.

[0183] Figure 27 is an explanatory diagram showing an example of the hardware configuration of a base station 60, which is an example of the application of the Doherty amplifier 1. The base station 60 shown in Figure 27 has a communication interface 61, a wireless communication circuit 62, a storage device 63, and a processor 64. The communication interface 61 is an interface that connects to a higher-level device and acquires information such as transmission signals. The wireless communication circuit 62 incorporates the Doherty amplifier 1, which amplifies RF signals. The storage device 63 stores various tables such as the phase table 44 and delay table 45 within the Doherty amplifier 1. The processor 64 is the control unit 46 within the adjustment unit 26 of the Doherty amplifier 1.

[0184] Furthermore, the adjustment unit 26 (26A, 26B) and DPD 10 are implemented as hardware, for example, by a processor. Examples of processors 64 include CPU (Central Processing Unit), DSP (Digital Signal Processor), FPGA (Field Programmable Gate Array), etc. The storage device 63 is implemented as hardware, for example, by RAM (Random Access Memory) such as SDRAM (Synchronous Dynamic Random Access Memory), ROM (Read Only Memory), or flash memory. The distribution unit 21, the first DAC 22A, the second DAC 23A, the first driver amplifier 22B, the second driver amplifier 23B, CA 22C and PA 23C, the combining unit 24, and ADC 30 are implemented as analog circuits, for example. [Explanation of Symbols]

[0185] 1, 1A, 1B Doherty Amplifiers 10 DPD 14 Distortion compensation section 21 Distribution section 22C CA 23C PA 24 Synthesis section 26, 26A, 26B adjustment section 41 Phase adjustment section 42 Delay adjustment section 44 Phase Table 45 Delay Table 46 Control Unit 47 Amplitude adjustment section 47A Amplitude-frequency adjustment section 48 Amplitude Table 51 First calculation unit 52 Classification Department 53 Second calculation unit 54 Decision Section

Claims

1. A distortion compensation unit that compensates for distortion in the input signal, and a distribution unit that distributes the distortion-compensated input signal into a first signal and a second signal, A first amplification unit that amplifies the first signal described above, A second amplification unit that amplifies the second signal, A combining unit that combines the first signal from the first amplification unit and the second signal from the second amplification unit, An adjustment unit is positioned between the distribution unit and the first amplification unit or the second amplification unit, and adjusts the phase delay value of the first signal or the second signal from the distribution unit. It has, The adjustment unit is, A first calculation unit calculates a characteristic showing the relationship between the gain of the output signal and the reference power of the input signal, based on the output signal output from the synthesis unit and the input signal input to the distortion compensation unit. A classification unit that classifies the calculated characteristics into multiple regions according to the level of the reference power, A second calculation unit calculates the statistical value of the gain for each of the classified regions, A determination unit sets the adjustment unit a phase delay value that provides the best wireless characteristics based on the calculated statistical values ​​for each region, A Doherty amplifier characterized by having the following features.

2. The first calculation unit is, The Doherty amplifier according to claim 1, characterized in that, after setting a predetermined phase delay value in the adjustment unit with distortion compensation not applied in the distortion compensation unit, the characteristics are calculated based on the output signal output from the combining unit and the input signal input to the distortion compensation unit.

3. The first calculation unit is, The Doherty amplifier according to claim 2, characterized in that the AM-AM (gain) characteristic is calculated as the aforementioned characteristic.

4. The adjustment unit is, A phase adjustment unit that adjusts the phase value of the first signal or the second signal, It includes a delay adjustment unit connected in series with the phase adjustment unit to adjust the delay value of the first signal or the second signal, The aforementioned classification unit is The characteristics are classified into a Low region, a High region (higher than the Low region), and a Peak region (higher than the High region) according to the level of the reference power within the aforementioned characteristics. The aforementioned determination unit, The phase adjustment unit is set to a set phase value that maximizes the average gain, which is a statistical value of the Peak region. The Doherty amplifier according to claim 1, characterized in that the delay adjustment unit is set to a set delay value that maximizes the average gain, which is a statistical value of the High region or the Peak region.

5. The aforementioned determination unit, The Doherty amplifier according to claim 4, characterized in that the phase delay value is set in the adjustment unit to minimize the decrease in gain in the Peak region and reduce the variation in gain in the High region and the Low region so that the wireless characteristics are best.

6. The first calculation unit is, The Doherty amplifier according to claim 1, characterized in that, at a predetermined timing before product shipment of the Doherty amplifier, a predetermined phase delay value is set in the adjustment unit in the distortion compensation unit with distortion compensation not applied, and then the characteristics are calculated based on the output signal output from the combining unit and the input signal input to the distortion compensation unit.

7. The adjustment unit is, The Doherty amplifier according to claim 4, further comprising an amplitude adjustment unit connected in series with the phase adjustment unit or the delay adjustment unit to adjust the amplitude pattern of the first signal or the second signal.

8. The adjustment unit is, The Doherty amplifier according to claim 4, further comprising an amplitude frequency adjustment unit connected in series with the phase adjustment unit or the delay adjustment unit for adjusting the amplitude frequency correction pattern of the first signal or the second signal.

Images