ADAPTIVE IMPEDANCE POWER AMPLIFIER

Abstract

The present invention relates to a method, of providing adaptive impedance in a Power Amplifier (PA), by providing more than one transistors in which one transistor is used to change the load line or to linearize the input signal by adapting the biasing of each transistor, wherein the transistors are connected in parallel.

Claims

1) A method of providing an adaptive impedance in a Power Amplifier (PA), comprising the steps of: a) providing a main amplifier; b) providing at least one auxiliary amplifier, in which during operation each of which switched between active or inactive operation mode; and c) adapting the bias of the main amplifier with or without biasing also the Vg of the main and auxiliary amplifiers, thus the combined output impedance of the PA, as seen by an output load, is maintained constant regardless of whether the auxiliary amplifier is active or inactive.

2) A method according to claim 1, wherein the output impedance maintain constant by choosing Vd and Vg in a way, that one amplifier will compensate the distortion of the second amplifier (Main to Auxiliary and vice versa).

3) A method according to claim 1, further comprising tuning Vg and Vd as to get smoother gain performances.

4) A method according to claim 1, further comprising dynamically changing the Vg and Vd for both amplifiers together in order to have an Adaptive Biasing efficiency performances with an adaptive impedance, and therefore obtaining a flat gain response and/or a higher efficiency response.

5) A power amplifier (PA) system, comprising: a) a main amplifier; b) at least one auxiliary amplifier, wherein both of said amplifier are transistors based; and c) a voltage control means for adapting the bias of the main amplifier in such a way that the output impedance of the main amplifier when the auxiliary amplifier is inactive, is similar to the combined output impedance of the main and auxiliary amplifiers in parallel when the auxiliary amplifier is active.

6) The system according to claim 5, in which the combined output impedance of the amplifiers, as seen by an output load, is substantially the same regardless of whether the auxiliary amplifier is active or inactive.

7) The system according to claim 5, in which a load line as seen by the main amplifier is adaptively changing in such a way that it will get from the main amplifier linear and high efficiency performances.

8) A system according to claim 5, in which during a power interval, one transistor is being used as a main transistor, when the rest of the transistors are being used as an impedance.

9) A system according to claim 5, in which during power interval, one transistor is being used as a main transistor, when the rest of the transistors are being used as a power linearize and/or phase linearize, and/or as efficiency improving.

10) A system according to claim 5, in which for a modulated signal, the transistors can be active during any power level, or just in part of the interval.

11) A system according to claim 5, in which each transistor can change its quiescent current (in FET Vgs—class of operation), or it biasing level (in FET Vds) in function of RF power levels.

12) A system according to claim 6, in which power supply voltage amplitude of the main amplifier is lower in compare to a single transistor envelope tracking, and the main amplifier reach high efficiency even before that envelop tracking is active, as the power supply modulator efficiency is a function of the power drop, with low efficiency at backoff and higher efficiency at the peak levels, efficiency is gained.

13) A system according to claim 5, further comprising providing more than one transistors in which one transistor is used to change the load line or to linearize the input signal by adapting the biasing of each transistor, wherein the transistors are connected in parallel.

14) A system according to claim 5, in which two or more transistors are connected in parallel to an output load so that adjustment of each transistor by the Gate and/or Drain voltage maintains linearity, and/or efficiency, and/or max power delivering, and the load line angle keeps best Power Added Efficiency (PAE).

15) A system according to claim 5, in which changing the biasing of one or more amplifiers, is used as an “automatic” matching correction system, when a change in the load impedance have been performed.

16) A system according to claim 5, further comprising one or more additional amplifiers, which affects each other according to the formula where the impedance equal to the load multiplied by 1+ration between the currents at the loads.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0032] In the drawings:

[0033] FIG. 1A schematically illustrates a block diagram of an ideal Doherty power amplifier (DPA), according to the prior-art;

[0034] FIG. 1B is a graph showing the output currents of the DPA;

[0035] FIG. 1C is a graph showing the output voltages of the DPA;

[0036] FIG. 1D is a graph showing the efficiency characteristics of the DPA;

[0037] FIG. 1E is a graph showing the linearity characteristics of the DPA;

[0038] FIG. 2A schematically illustrates a block diagram of an Adaptive Impedance Amplifier having a main power amplifier (PA) and an auxiliary PA when the main PA is active and the auxiliary PA is inactive, according to an embodiment of the invention.;

[0039] FIG. 2B schematically illustrates a block diagram of the Adaptive Impedance Amplifier when both main and auxiliary PAs are active, according to an embodiment of the invention;

[0040] FIG. 3 shows a small signal output reflection coefficient of MRF8S21100;

[0041] FIG. 4 schematically illustrates an optional configuration of the Adaptive Impedance Amplifier with RF detector, Vd switcher block and a Vg switcher block, according to an embodiment of the invention;

[0042] FIG. 5 schematically illustrates an optional configuration for the Vg switcher block of FIG. 4, according to an embodiment of the invention;

[0043] FIG. 6 schematically illustrates an optional configuration for the Vd switcher block of FIG. 4, according to an embodiment of the invention;

[0044] FIG. 7 schematically illustrates an optional configuration of the Adaptive Impedance Amplifier with current sample and switching Vd block, according to an embodiment of the present invention;

[0045] FIG. 8 schematically illustrates an optional configuration of the Adaptive Impedance Amplifier using a modem data, according to an embodiment of the present invention;

[0046] FIG. 9 schematically illustrates an optional configuration of the Adaptive Impedance Amplifier using adaptive biasing technology, according to an embodiment of the invention;

[0047] FIG. 10 schematically illustrates an equivalent amplifier circuitry with two transistors, when one functions as a main transistor (derived in class AB or B), and the other transistor functions as an auxiliary transistor (derived in class C) and works just during peak power intervals, according to an embodiment of the invention;

[0048] FIG. 11 is a graph that shows the currents of the equivalent amplifier circuitry of FIG. 10; and

[0049] FIGS. 12 and 13 schematically illustrates a loadpull data simulation for a MRF8S21100 RF laterally diffused metal-oxidesemiconductor (LDMOS) field-effect transistors (FETs) by Freescale.

DETAILED DESCRIPTION OF THE INVENTION

[0050] Reference will now be made to several embodiments of the present invention, examples of which are illustrated in the accompanying figures. Wherever practicable similar or like reference numbers may be used in the figures and may indicate similar or like functionality. The figures depict embodiments of the present invention for purposes of illustration only. One skilled in the art will readily recognize from the following description that alternative embodiments of the structures and methods illustrated herein may be employed without departing from the principles of the invention described herein.

[0051] In class A the impedance of a transistor that can be seen by the load is varied by the power source in a simple manner:

[00001]$\begin{array}{cc}\mathrm{Rds}=\frac{\mathrm{Vds}}{\mathrm{Ids}}& \left(1\right)\end{array}$

[0052] Assuming at class A that ids (Alternative Current—AC) variation is small, we can take Ids (Direct Current—DC) as a constant. Therefore, Rds is a linear function of Vds. Since at thot class of operation, the goal is to match the load to the conjugate impedance of the amplifier, the impedance will be a function of Vdd.

[0053] FIGS. 2A and 2B show a circuit schematic of a power amplifier (PA) system that can be used in conjunction with the present invention for providing an Adaptive Impedance PA. The system generally indicated by numeral 1 in the figure comprises a main amplifier 2 and at least one auxiliary amplifier 3 for handling the peaks of the envelope signals. In this embodiment, the PA system is designed in such a way that the output impedance of the main amplifier 2 when the auxiliary amplifier 3 is off (i.e., inactive), is similar to the combined output impedance of the main and auxiliary amplifiers in parallel when the auxiliary amplifier 3 is on (i.e., active). Thus, the combined output impedance of the PA 1, as seen by an output load R.sub.L (Rload), is substantially the same regardless of whether the state of the auxiliary amplifier 3 is on or off. According to an embodiment of the invention, this can be achieved by adapting the bias of the main amplifier 2 with or without biasing also the gate voltage (i.e., Vg) applied to a terminal (the gate) of the main and auxiliary amplifiers, as will be described in further details hereinafter.

[0054] The method of such PA system for achieving an Adaptive Impedance, may work in one of the following ways: [0055] Assuming the auxiliary amplifier 3 is at “off” state and then sharply turns “on” when the peak arrives. That can be performed by changing the Vg of the auxiliary amplifier 3. [0056] Keeping the auxiliary amplifier 3 at class C so the amplitude of the Radio Frequency (RF) signal itself will open the amplifier.

[0057] Therefore, the auxiliary amplifier 3 reflects a variable in function of the input RF power that actually affects Vg. The drain voltage (i.e., Vd) in that situation must be an adaptive function of the RF power to keep constant Impedance/Gain/Phase. In both ways, the variation of Vd can work either to “fix” the total impedance to keep it as a constant and as an additional degree of tuning to keep the gain and phase over power flat (AM-PM, AM-AM).

[0058] According to this embodiment, as a replacement to the quarter wave transformer that is matching between the amplifiers. The offered method can be used for as many auxiliary amplifiers as required. Since it doesn't need a transformer between the amplifiers, it will be also size effective. Thus, the combined output impedance of the PA 1, as seen by the output load R.sub.L, is substantially the same regardless of whether the auxiliary amplifier is on or off.

[0059] In FIG. 2A the main amplifier 2 is active and the auxiliary amplifier 3 is inactive. The main amplifier 2 is matched to the output load R.sub.L. In FIG. 2B both the main amplifier 2 and the auxiliary amplifier 3 are active, but with a different biasing of the gate voltage of the main amplifier (Vgm2), the gate voltage of the auxiliary amplifier (Vga2) and the drain voltage of the main amplifier (Vdm2) and the drain voltage of the auxiliary amplifier (Vda2) to achieve similar impedance to the one that was when just the main amplifier 2 was active. In general, several variations are possible. Vg for all PA's can be kept as a constant and the variation may be applied only on Vd. Vd can be equal in both amplifiers or non-equal, when the main amplifier 2 can be biased by Vg in class AB, and the auxiliary amplifier 3 in class C, it is possible to switch between voltage levels in one or more steps, or to adaptively tune them in function of the RF power envelope levels.

[0060] The drain voltages of the auxiliary amplifier 3 are indicated as Vda1 at FIG. 2A and as Vda2 at FIG. 2B. The drain voltages of the main amplifier 2 are indicated as Vdm1 at FIG. 2A and as Vdm2 at FIG. 2B. The gate voltages of the auxiliary amplifier 3 are indicated as Vga1 at FIG. 2A and as Vga2 at FIG. 2B. The gate voltages of the main amplifier 2 are indicated as Vgm1 at FIG. 2A and as Vgm2 at FIG. 2B.

[0061] The Adaptive Impedance method predicts to have flat gain response in compare to the Doherty PA. It is feasible, thanks to the voltage control of the amplifiers 2, 3 that keeps the impedance seen by the load R.sub.L, Xm1 as a constant, and therefore, will have better linearity. In compare to the adaptive biasing method, it is predicted to achieve also better gain behavior thanks to the constant impedance. Efficiency should be similar or even better in compare to both methods.

[0062] In Class A, the adaptive impedance method is that a solution of the equation (3), and flat gain response, can be achieved by varied Vd and Vg of the amplifiers.

[0063] It is clear that:

[00002]$\begin{array}{cc}[\begin{array}{c}\mathrm{Xpa}=X_{M.Math..Math.1}.Math..Math.\mathrm{Auxiliary}=\mathrm{Off}\\ \mathrm{Xpa}=\frac{X_{M.Math..Math.2}X_A}{X_{M.Math..Math.2}+X_A}.Math..Math.\mathrm{Auxiliary}=\mathrm{On}\end{array}& \left(2\right)\end{array}$

Where Xpa represents the impedance

[0064] Since X.sub.M=F(Vd,Vg), we can find Vda1/Vdm1/Vda2/Vdm2 (they can be equal or non-equal) and Vgm1/Vga1/Vga2/Vgm2 such that will keep Xpa as a constant. And therefore we are looking for the solution of:

[00003]$\begin{array}{cc}{X}_{M.Math..Math.1}=\frac{X_{M.Math..Math.2}X_A}{X_{M.Math..Math.2}+X_A}& \left(3\right)\end{array}$

The solution of that equation can be found in RF Amplifiers by looking at the reflection coefficient of an Amplifier high power model.

[0065] In addition, as a standalone method, or as an additional method to the claim of keeping the Impedance as a constant, another level of linearization grantee, can be applied by choosing Vd and Vg in a way, that one PA will compensate the distortion of the second amplifier (Main amplifier to Auxiliary amplifier and vice versa). It can be considered as an active way to the quarter wave transformer of the Doherty Amplifier distortion compensation.

[0066] In this invention, the power combining of the RF signal on the transistor output, is done just by connecting both Drain terminals of the transistors of each amplifier together with a short circuit (with or without a DC coupling capacitor between them). Yet, to get 180 degrees between standing waves, a 90 degree line between the transistors may be considered.

[0067] In Class AB, B, and C the goal is to deliver the max possible power:

[00004]$\begin{array}{cc}P.Math..Math.\max =\mathrm{Vdc}*\frac{I.Math..Math.\max }{4}& \left(4\right)\end{array}$

[0068] The Load Line will be therefore:

[00005]$\begin{array}{cc}\mathrm{Rload}=\frac{V.Math..Math.\max }{I.Math..Math.\max }& \left(5\right)\end{array}$

Where, the Load Line represents the maximum power that a given transistor can deliver, and it is determined by the power supply voltage and the maximum current of the transistor.

[0069] In envelope tracking the ratio between the maximum voltage (Vmax) and the maximum current (Imax) is kept constant and therefore the Rload (R.sub.L) is a constant as well. The drain efficiency is also kept

[00006]$\frac{\pi }{4}$

across the power range.

[0070] In practice, due to the knee affect, to keep maximum efficiency and power delivering, the load resistance that is being reflected to the transistor is increasing. One solution is to work in backoff, but that doesn't allow using 100% of the voltage amplitude, by the cost of losing some efficiency.

[0071] The method presented, is based on the concept that two or more transistors are connected in parallel to the load. A matching network might be needed to match the transistors to the load, as shown with respect to an equivalent circuit in FIG. 10.

[0072] During power interval, one transistor is being used as a main transistor, when the rest of the transistors are being used as an impedance, and/or power linearize, and or phase linearize, and or efficiency improving.

[0073] For modulated signal, the transistors can be active during any power level, or just in part of the interval. Each transistor can change its quiescent current (in FET Vgs—class of operation), or it biasing level (in FET Vds) in function of RF power levels.

[0074] Referring now to FIGS. 10 and 11, one example of such use is two transistors, when one functions as a main transistor 2, and derived in class AB or B. The other transistor functions as an auxiliary transistor 3, and is derived in class C and works just during peak power intervals. The matching network of the main transistor 2 is indicated by numeral 13 and matching network of the auxiliary transistor 3 is indicated by numeral 14. That architecture, when biasing correctly, is able to keep linearity, efficiency, and max power delivering. Actually, adjustment of each transistor by the Gate and Drain voltage (in FET, but in BJT it can work as well), can maintain linearity, efficiency, and max power delivering tough, or better to say, thanks to the variance of the transistor parameters (such as load line angle keeps best Power Added Efficiency (PAE)) with biasing its Gate and Drain voltage. During the peak power intervals the auxiliary transistor 3 is changing the load line that is reflected to the main transistor 2, in a way that the load line will provide to the main transistor matching conditions to have better efficiency, higher power, and or better linearity.

[0075] Furthermore, as shown in the graph of FIG. 11, the behavior of the current I1 is derived from the knee effect of the main transistor, and the behavior the current I2 is derived from the Class C of the auxiliary transistor.

[0076] Still, the system, by manipulating the biasing of the transistors, is also able to overcome the variance of conductance (gm) over Vd as appears in the non-ideal transistors.

[0077] The main amplifier will “see” an impedance of:

[00007]$\begin{array}{cc}Z.Math..Math.1=\mathrm{RL}\left(1+\frac{I.Math..Math.2}{I.Math..Math.1}\right)& \left(6\right)\end{array}$

[0078] If we are changing the biasing parameters of the main amplifier 2, as naturally being performed in Envelop Tracking, which keeps the load that is providing our goal performances of the system (such as Efficiency, linearity, etc.) equal to Z1, that will be a perfect matching conditions. The main claim is that such a combination, that actually reflects dynamic load line, is an advantage since an envelope tracking amplifier will have the highest efficiency when that dynamic load is being presented.

[0079] Another advantage of that method is that the power supply voltage amplitude of the main amplifier 2 can be lower in compare to a single transistor envelope tracking, thanks to the fact that the main amplifier 2 can reach high efficiency even before that envelop tracking is active. Since the power supply modulator efficiency is a function of the power drop as well, with low efficiency at backoff and higher efficiency at the peak levels, we can also gain here in efficiency.

[0080] Changing the biasing of the main and auxiliary amplifiers, is also can be used as an “automatic” matching correction system, when a change in the load impedance have been performed (for example, when holding an antenna of a transmitter with bare hands).

[0081] Referring now to FIGS. 12 and 13, as an example, a loadpull data simulation for a MRF8S21100 RF laterally diffused metal-oxidesemiconductor (LDMOS) field-effect transistors (FETs) by Freescale is shown. For Vgs=2.7V (Class AB) Idq=0.7 A the transistor provides best PAE with the impedance values as indicated in the graph of FIG. 12.

[0082] It can be easily seen from the graph, that in order to keep maximum PAE for any Vds, Z load is not a constant. It can be explained by the transistor non ideal behavior such as the knee affect as shown in FIG. 13, in a low power load line (as indicated by the marker m1) and in a high power load line and a high power load line for maximum power delivery (as indicated by the marker m2).

[0083] All the above will be better understood through the following illustrative and non-limitative examples.

[0084] As an example for the implementation of system 1 (as described hereinabove with respect to FIGS. 2A and 2B), the Freescale MRF8S21100 RF laterally diffused metal-oxidesemiconductor (LDMOS) field-effect transistors (FETs) can be used, which its load impedance curves can be followed at FIG. 3.

[0085] In FIG. 3, the two markers, m1 and m2 reflect a change at the output impedance of the transistor. In this case the impedance is close to be multiplied just by varying Vd when Vg being kept constant. In impedance Biasing we want to drive both the main amplifier 2 and the auxiliary amplifier 3 with Vd=24V and the gate voltage of the main amplifier 2 (Vgm=2.8V) when the gate voltage of the auxiliary amplifier 3 (Vga) is below the peaking level. When reaching the peaking level and up to the peak high Vd for both main and auxiliary amplifiers will be set to approximately 31V, and the Vg of the auxiliary amplifier 3 will be changed to 2.8V. So in that example both amplifiers will function at Class A or Class AB when they both active, while the parallel impedance of both amplifiers will be kept as a constant. Naturally, it is possible to slightly tune Vg and Vd to get smoother gain performances. It is also possible to dynamically change the Vg and Vd for both amplifiers together in order to have an Adaptive Biasing efficiency performances with a constant impedance, and therefore a flat gain response. Another option may be to choose two different transistors for matching and efficiency considerations.

[0086] As will be appreciated by a person skilled in the art, there are several solutions in the market to adaptively bias an amplifier, similar techniques with some “levels” modifications could be performed here as well. For example, at FIG. 4 we can find an Envelope Detector (ED) 6 that being active above reference level (set the high of the peak that first active the biasing circuitry). A VGS switcher block 5 sets the bias of the Vg, and a VDS switcher block 4 sets the bias of the Vd of the main amplifier 2 and the auxiliary amplifier 3.

[0087] To build those blocks, couple of circuits may be offered. Some from the portfolio of companies like Qualcom or Nujira. For example, as shown in FIG. 5, the VGS switcher block 5 may involve the following:

A comparator A1 gets the envelope level, Vrefg2 sets the level it become at high position, an Operational Amplifier (OPA) A2 sets the required gain, and an amplifier A3 sets the final Vg as a constant that being sets by Vrefg1+the value of OPA A2. That circuit with equal or different resistors and reference values can be used for both PA's gates.

[0088] VDS switcher block 4 that serves as a Vd switcher, can be performed in several ways as well, one way is described below with respect to FIG. 6.

[0089] The suggested Vd switcher in FIG. 6 is delivering to the PA V1 before reaching the peaking level, and then Q1, Q2, Q3, or Qn are being chosen for higher voltage levels (up to Vn) depending on the required voltage level. This circuit may be used for both amplifiers 2 and 3 together (shares the same Vd) or separately.

[0090] Referring now to FIG. 7, an alternative to the envelope detector can be offered. It is using a current sampler on the main amplifier 2 or on the auxiliary amplifier 3 to control the biasing levels of the amplifiers.

[0091] If there is access to the base band data, it is optional to use adaptive biasing “off the shelf” parts as the element QFE1100 by Qualcom (as indicated by numeral 11 in FIG. 8). A transfer function 10 might be required to adjust the device to the adaptive biasing requirements for the main amplifier 2 and the auxiliary amplifier 3. Such a configuration is shown in FIG. 8.

[0092] Another topology that can be used is based on Nujira Adaptive Biasing technology (as indicated by numeral 12 and presented in FIG. 9).

[0093] Farther degree of linearization may be provided by an envelope detector added to the load with a feedback to the Gate and Darin voltage levels. It can be added to any of the diagrams suggested hereinabove with respect to FIGS. 4 and 7-9.

[0094] There should be no limitation to the transistors technologies that might be used. Any technology that can work for a regular RF amplifier will probably fit here as well. LDMOS, GaAs, and GaN devices will fit. Following, the concept of the Adaptive Impedance with an internally matched device MRF8S21100 from Freescale is shown, but non matched devices can be used as well.

[0095] All the above description and examples have been given for the purpose of illustration and are not intended to limit the invention in any way. Those skilled in the art will perceive improvements, changes, and modifications. Such improvements, changes, and modifications within the skill of the art are intended to be covered by the appended claims. Many different scheme and electronic elements can be employed, all without exceeding the scope of the invention.