Band-based amplifier linearity adjustment

Abstract

A system improve amplifier efficiency of operation relative to that of an amplifier with fixed biasing is operating channel dependent. A control circuit determines a bias current for an amplifying transistor of an amplifier circuit based at least in part on an operating channel. The amplifying transistor operates in a multi-channel system, where the bias current for the amplifying transistor operating at channels at an edge of a channel band is different from the bias current for the amplifying transistor operating at channels nearer a center of the channel band.

Claims

1. An amplifier circuit comprising a control circuit configured to determine a bias current for an amplifying transistor based at least in part on an operating channel, the amplifying transistor configured to operate in a multi-channel system, the bias current for the amplifying transistor operating at channels at an edge of a channel band being different from the bias current for the amplifying transistor operating at channels nearer a center of the channel band.

2. The amplifier circuit of claim 1 wherein the control circuit is further configured to provide the bias current to the amplifying transistor through a control signal.

3. The amplifier circuit of claim 2 wherein the control signal includes a digital control signal.

4. The amplifier circuit of claim 2 wherein the control signal is determined to improve an operation efficiency of a mobile device including the amplifier circuit.

5. The amplifier circuit of claim 1 further comprising a matching circuit configured to provide impedance matching between an output port of the amplifying transistor and a load.

6. The amplifier circuit of claim 5 wherein the control circuit is further configured to determine matching characteristics of the matching circuit, the matching characteristics based at least in part on a transmission power level from the amplifying transistor and an indication of amplifier circuit operation.

7. The amplifier circuit of claim 6 wherein the matching characteristics of the matching circuit are different for the channels at the edge of the channel band than for the channels nearer the center of the channel band.

8. The amplifier circuit of claim 6 wherein the indication of amplifier operation includes an error vector magnitude target, an operating point, and the operating channel of the amplifier circuit.

9. The amplifier circuit of claim 1 wherein the control circuit is further configured to determine the bias current for the amplifying transistor based at least in part on an error vector magnitude target and an operating point.

10. A method to determine bias current based at least in part on an operating channel, the method comprising: determining a first bias current for an amplifying transistor when the amplifying transistor is operating at channels at an edge of a channel band, the amplifying transistor configured to operate in a multi-channel system; and determining a second bias current for the amplifying transistor when the amplifying transistor is operating at channels nearer a center of the channel band, the second bias current different from the first bias current.

11. The method of claim 10 further comprising providing impedance matching between an output port of the amplifying transistor and a load.

12. The method of claim 10 further comprising determining matching characteristics of a matching circuit, the matching characteristics based at least in part on a transmission power level from the amplifying transistor and an indication of amplifier circuit operation.

13. The method of claim 12 wherein the matching characteristics of the matching circuit are different for the channels at the edge of the channel band than for the channels nearer the center of the channel band.

14. The method of claim 12 wherein the indication of amplifier operation includes an error vector magnitude target, an operating point, and the operating channel of the amplifier circuit.

15. A wireless communication device comprising: an antenna configured to transmit an amplified radio frequency output signal; and an amplifier circuit configured receive a radio frequency input signal and to amplify the radio frequency input signal to provide the amplified radio frequency output signal, the amplifier circuit including a control circuit configured to determine a bias current for an amplifying transistor based at least in part on an operating channel, the amplifying transistor configured to operate in a multi-channel system, the bias current for the amplifying transistor operating at channels at an edge of a channel band being different from the bias current for the amplifying transistor operating at channels nearer a center of the channel band.

16. The wireless communication device of claim 15 wherein the control circuit is further configured to provide the bias current to the amplifying transistor through a control signal.

17. The wireless communication device of claim 15 wherein the amplifier circuit further includes a matching circuit configured to provide impedance matching between an output port of the amplifying transistor and a load.

18. The wireless communication device of claim 17 wherein the control circuit is further configured to determine matching characteristics of the matching circuit, the matching characteristics based at least in part on a transmission power level from the amplifying transistor and an indication of amplifier circuit operation.

19. The wireless communication device of claim 18 wherein the matching characteristics of the matching circuit are different for the channels at the edge of the channel band than for the channels nearer the center of the channel band.

20. The wireless communication device of claim 18 wherein the indication of amplifier operation includes an error vector magnitude target, an operating point, and the operating channel of the amplifier circuit.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) Exemplary embodiments of the invention will now be described in conjunction with the following drawings, in which:

(2) FIG. 1a is a simplified flow diagram of a method according to an embodiment of the invention;

(3) FIG. 1b is a simplified block diagram of an amplifier circuit according to an embodiment of the invention;

(4) FIG. 2a is a graphical representation of EVM and Mask Limit vs. Pout;

(5) FIG. 2b is a graphical representation of EVM vs. power output at two different bias points;

(6) FIG. 3 is a graphical representation of total current vs. power output;

(7) FIG. 4 is a graphical representation of Digital Bias Optimized for 3% EVM;

(8) FIG. 5 is a graphical representation of Current Drain for Optimized 3% EVM;

(9) FIG. 6 is a graphical representation of Efficiency optimized for Spectral Mask;

(10) FIG. 7 is a graphical representation of 802.11b 1 Mbps OOB performance;

(11) FIG. 8 is a graphical representation of 802.11g 54 Mbps OOB performance;

(12) FIG. 9 is a graphical representation of Current Drain increase for Optimum OOB;

(13) FIG. 10 is a graphical representation of burst current vs. output power;

(14) FIG. 11 is a graphical representation of EVM vs. output power;

(15) FIG. 12a is a simplified flow diagram of a method relying on a discrete number of biasing states for the amplifier; and

(16) FIG. 12b is a simplified block diagram of an amplifier for use with the method of FIG. 12a.

DETAILED DESCRIPTION

(17) The following description is presented to enable a person skilled in the art to make and use the invention, and is provided in the context of a particular application and its requirements. Various modifications to the disclosed embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be applied to other embodiments and applications without departing from the scope of the invention. Thus, the present invention is not intended to be limited to the embodiments disclosed, but is to be accorded the widest scope consistent with the principles and features disclosed herein.

(18) In Table 1, System EVM requirements are shown for different modulation rates. As is evident, different modulation schemes and rates have very different EVM requirements and designing to meet these requirements often requires a design that meets the most demanding of the requirements. Often, this limits the power consumption advantages of using other modulation schemes or rates since the system is less efficient when operating at modulation schemes and rates other than that having the most demanding requirements. Hereinafter, modulation schemes or rates for which some standard has defined demanding EVM requirements will be referred to as demanding modulation schemes or rates. Of course, designing the system to be more efficient with less demanding modulation schemes or rates is also ineffective because (a) it only aggravates the drawbacks to the more demanding modulation schemes and (b) often it results in the failure to meet specification requirements for those more demanding modulation schemes or rates. For example, a PA may be designed with a very high linearity performance with a view to the requirement of a data rate equal to 54 Mbps. Operating such a PA at lower data rates would normally be wasteful of the power consumption if the bias point could not be adjusted to a less linear regime of operation for the PA.

(19) TABLE-US-00001 TABLE 1 802.11 g System EVM requirements Modulation Rate (Mbps) EVM (dB) EVM (%) BPSK 6 5 56.2% BPSK 9 8 39.8% QPSK 12 10 31.6% QPSK 18 13 22.4% 16-QAM 24 16 15.8% 16-QAM 36 19 11.2% 64-QAM 48 22 7.9% 64-QAM 54 25 5.6%

(20) A PA with fixed bias points and fixed interstage/output port match has a fixed power consumption, heat dissipation, and linearity performance on all channels, regardless of actual requirements or use. In the context of multi-mode operation, such a PA cannot adjust the operating point or matching circuits to decrease or increase linear performance as required by the mask requirements, EVM requirements, or jurisdiction-dependent operating specifications.

(21) A PA's operating efficiency is beneficially enhanceable by approximately optimizing a trade-off between the PA's efficiency and linearity by a set-up of its operating point and matching circuit based on the de minimis mask and EVM requirements for the modulation scheme and rate of a channel that is to be used in a multi-channel OFDM system. Such an approach maintains the specified EVM performance but reduces power consumption. Controlling the PA's bias provides a straightforward method for approximately optimizing performance as the modulation rate, maximum allowed power level, or in-band channel is changed.

(22) In a broad embodiment, in accordance with system requirements or user demand, a baseband processor determines one or more of EVM, mask, channel to be used, maximum allowable power output level, modulation type, and data rate associated with the RF output signal to be transmitted. The baseband processor then maps this information onto a digital control instruction for the PA. The PA then adjusts, in a predetermined manner, the operating point for one or more amplification stages and/or the matching circuit adjustments for one or more of the interstage/output match circuits. The resulting operating point and matching circuit configuration is one that represents a better trade-off in efficiency whilst meeting the requirements of the system effectively. Some possible requirements include EVM, mask, channel to be used, channel bandwidth, maximum allowable power output level, modulation type, and data rate at a maximum of PA efficiency. Of course, only those requirements specified or deemed essential need be considered. Of note, in some systems fewer or more requirements are specified. This method is illustrated in FIG. 1a for the apparatus of FIG. 1b.

(23) Alternatively, analogue control signals are used, for example, a control signal is provided for each controllable aspect of the amplifier operation.

(24) Alternatively, the information on EVM, mask, channel to be used, maximum allowable power output level, modulation type, and data rate is provided to the PA and the mapping to the appropriate operating point for one or more amplification stages and/or matching circuit adjustments for one or more of the interstage/output match circuits is determined and controlled internally to the integrated PA circuit. In either case, PA efficiency is improvable even in consideration of many operating factors and requirements.

(25) For example, according to an embodiment with a multi-carrier OFDM modulation, the operating point of the amplifying transistors and/or the interstage matching circuit is adjusted for lower or higher power efficiency in a circumstance where: the channel selected to be employed in a transmission system is close to a band edge. For example, in the US, the 2.4 GHz WLAN band ranges from 2400-2483.5 MHz. Typical operating channels include channel 1 (2412 MHz), 6 (2437 MHz), and 11 (2462 MHz). The FCC has defined so-called restricted bands of operation where emissions must be lower than a predefined limit to protect devices operating in these bands. For example, there is a restricted band from 2300-2390 MHz, and another from 2483.5 MHz to 2500 MHz. These restricted bands form a band-edge, and put strict limits on how much power can be radiated into these bands. If the PA is operating on channel 1 or 11, it is obvious that the restricted band is quite close to the operating frequency of the PA (22 MHz below channel 1, and 21.5 MHz above channel 11). However, if the PA is operating in channel 6, the restricted band is now greater than 45 MHz away. Out-of-band emissions from a power amplifier are directly related to PA linearity, so a PA operating in channels 1 or 11 is often required to be more linear than one operating in channel 6. Thus, changing the biasing and/or matching characteristics of the amplifier when operating in channel 1 or 11 is advantageous because efficiency is improvable for a relatively common situation. a data rate to be employed is lower or higher and correspondingly requires less or more linear performance from the PA, the type of modulation to be employed requires a less or more linear performance from the PA (e.g. QPSK, BPSK, QAM 16, etc.) the jurisdiction in which the PA is to be compliant requires more or less linear power output from the PA to meet its standards, an evaluation of mask compliance and EVM performance indicates that an increase or decrease in linear performance or power level is required, and the channel bandwidth is variable and therefore requires more or less linear performance from the PA.

(26) For many WLAN applications, currents can be reduced by 25%. Further enhancements in WLAN applications occur at or near the band edge where bias is approximately optimized to increase transmit power. Further, for lower data rates, less current is required and, as such, when lower data rate transmission is used power savings are supported. It is advantageous to adjust the bias current in a fashion that does not prevent meeting of other requirements of the system such as EVM

(27) Of course, in today's world of international travel and business, often a single mobile communication device is intended to operate in many countries. As such, the ability to meet performance characteristics in different jurisdictions efficiently is highly advantageous. Similarly, the other situations also present advantages to enhancing efficiency of operation when possible.

(28) Preferably, power consumption is approximately minimized in a PA by adjusting of one or more operating points in amplification stages of the PA. Alternatively or in conjunction therewith, the matching circuits of the PA are adjusted. Such a method is useful for balancing different requirements. For example, it supports minimum simultaneous requirements of power output, absolute and relative spectral output outside an allowed frequency bands (mask), and EVM. In particular, in the context of an EVM limited transmission system, if the modulation rate is reduced, the power consumption is reducible using an adjustment in the operating point and optionally an adjustment in the matching circuits to reduce the level of back-off while still meeting mask requirements and the EVM requirement. In other cases, competing requirements are balanced using dynamic alterations to the amplifier circuit as taught hereinabove.

(29) In a more specific embodiment, an extremely critical performance metric of a WLAN Power Amplifier is EVM performance. To meet System EVM performance requirements, power amplifiers are typically specified to 3% EVM performance from the PA at the rated transmit power. This specification leaves some margin for the total system EVM. Referring to FIG. 2a, a diagram of EVM (%) vs. P.sub.out(dBm) is shown. As is evident from the graph, the EVM of 3% as discussed herein is a horizontal dashed line but the mask is only an approximation thereof several dB below the threshold to provide additional margin. At low P.sub.out, the margin is approximately 20 dB whereas as P.sub.out increases, the margin shrinks to about 6 dB. Because of the high peak to average power of OFDM modulation, power amplifier power is reduced significantly from the 1 dB compression point to meet the 3% EVM requirement. As noted hereinabove, reducing the PA output power results in poor PA efficiency when the PA is actually operated at the lower power level. In fact, Mask compliance at the 3% EVM level has several dB of margin, which makes the system EVM limited as seen in FIG. 2a where the threshold of 3% EVM is crossed by the 54 Mbps EVM curve much before the Mask (spectral energy) threshold with increasing power output level (P.sub.out). Though lower modulation rates reduce the EVM requirement and therefore support a lower bias setting which improves efficiency and extends product battery life, these are not always available or used.

(30) When optimizing EVM performance, particular care is given to biasing a circuit for efficiency of operation at the EVM performance. Consideration of EVM at lower power levels or backed off EVM performance is taken into consideration to not exceed a given specification. Typically, lower bias results in increased backed-off EVM, commonly referred to as EVM humping. This is illustrated in FIG. 2b where EVM at a lower bias point crosses the 3% threshold briefly at approximately 16 dBm power output and then the EVM crosses the threshold again at 20 dBm. This lower power operating point correspondingly results, in a lower ICC consumption (power consumption) over the full range of output powers (see FIG. 3). As such, it is desirable, but problematic.

(31) According to an embodiment of the invention, bias at each power level is adjusted to approximately optimize EVM and efficiency for better performance. This results in a dramatic improvement in overall PA performance. FIGS. 4 and 5 show the approximately optimized performance of a PA with digital bias control set for EVM and efficiency. There is a substantial improvement in current drain by dynamically controlling bias at each power level to meet the EVM requirements with a minimum current consumption performance. Advantageously, such a system reduces the overall power consumption, thereby increasing battery life, while maintaining other performance characteristics of the PA.

(32) Some applications allow for higher transmit power levels at lower modulation rates, where the performance is mask limited instead of EVM limited. Dynamic bias control allows for further increasing transmit power or improving efficiency by approximately re-optimizing power amplifier bias for approximately optimum spectral mask performance. Referring to FIG. 6, shown is an EVM vs. bias current graph wherein efficiency of operation is determinable. As is evident from the graph, the bias current is reduceable from 170 to 150 mA, improving the efficiency from 17% to 20%. By dynamically adjusting operating points in response to operation of the PA, efficiency is improved while maintaining performance within requirements.

(33) Band edge requirements are taken from the FCC restricted band limits, outlined in the FCC Part 15.205 specification, along with the radiated emission limits defined in FCC Part 15.209. Meeting the FCC Band Edge requirements typically requires setting lower transmit power levels at channels 1 and 11 (2412 MHz and 2462 MHz). Typically, power at the band edge channels1 and 11is 3 to 4 dB lower than that of channels that are not band edge limited. As noted above, this results in reduced efficiency of operation.

(34) As can be seen in FIGS. 7, 8, and 9, increasing PA bias current greatly improves Band Edge performance (or Out Of BandOOB) for 802.11b and 802.11g Band Edge performance. In FIG. 7, the performance improvement with the lower bias current is evident. Here, not only is the specified limit crossed at a higher power level, the OOB is lower and more consistent across the entire range where the higher bias current operation is within the specified limit. This allows for increased amplification when desired. From these plots it is evident that the transmit power can be increased up to 3 dB for 802.11b and 1.5 dB for 802.11g signals at the band edge channels by appropriately adjusting the power amplifier bias.

(35) In an alternative embodiment, a bias of the PA is adjusted in a discrete stepwise fashion. Referring to FIG. 10, an impact of using only 2 bias points with a single transition is shown. Icq is reduced from 250 mA to 128 mA, and power consumption is reduced. This provides some of the benefits outlined hereinabove but with a much simpler implementation. For example, 2-step control could be implemented using a separate control pin. Alternatively, a serial interface is used allowing reduced complexity as the number of bias points is increased. As shown in FIG. 11, EVM remains below 3% over the entire power range and the single bias change improves efficiency within one of the two ranges. In the example shown, there is a step discontinuity in EVM at 17 dBm. Of course, the switch point could be lowered if EVM<2% is required at backed-off powers.

(36) Alternatively, more than two bias points are used.

(37) In another alternative embodiment, a fixed bias point is used over one or more ranges while a variable bias point is used over one or more other ranges. Thus, the benefits of the simplicity of the embodiment of FIGS. 10 and 11 are achieved with the improved EVM behavior of the first embodiment over the ranges specified.

(38) Referring to FIG. 12a, a method is shown wherein a control signal is received for selectively switching the amplifier between a first biasing mode of operation with reduced current and a second other biasing mode of operation with higher bias current. For example, the control signal is provided at a pin wherein a 0 indicates low bias current and a 1 indicates higher bias current. Alternatively, the signal is provided within a digital control signal comprising other control values. Referring to FIG. 12b, an amplifier circuit is shown for use with the method of FIG. 12a.

(39) Advantageously, using CMOS technology within a SiGe BiCMOS technology allows for digital control of an amplifier according to embodiments of the present invention. The use of a digital control and interface for a control signal presents several advantages. Firstly, the control signal optionally comprises a digital selection of two or more amplifier configurations. For example, those shown in FIG. 12a. Using a digital control signal is advantageous as it allows N different amplifier configurations to be selectable without requiring numerous input ports or pins. Secondly, it allows for the amplifier to select an amplifier configuration based on data provided in a simple form which can, with little further processing, be used in the configuration of the amplifier. For example, the control signal comprises an indication of a standard, a channel, a modulation type and rate, etc. and the amplifier integrated circuit comprises logic for mapping those values onto a lookup table to determine a configuration. Alternatively, the values are processed to determine a configuration of the amplifier supporting continuous variation of biasing and/or matching circuits. Of course, it is also supported that a lookup table establishes the processing to be performed in order to support continuous variation of biasing and/or matching circuits.

(40) Within the amplifier, bias is controllable. Also within the amplifier, filters can be designed in a fashion to provide controllability. These tunable filters are useful in that better filteringhigher rejectionusually has an associated higher insertion loss. As such, an ability to tune a filter to reduce rejection would also advantageously reduce insertion loss. Filter tuning is useful in improving efficiency in some applications.

(41) In some embodiments, digital control, which supports separate control over each tunable aspect of an amplifier, is provided. This optionally includes each stage of the amplifier as well. As such, bias is controllable for each amplifier stage and filters are tunable independently for each amplification stage. By providing independent control, a control that results in an approximately most efficient operation of the amplifier is selectable. This also adds significant flexibility in the event of new standards and new modulation techniques.

(42) Optionally only some of the stages are independently controllable. Further optionally, each stage receives a same control signal.

(43) In an alternative embodiment, the matching circuits are controlled instead of the bias current in order to achieve similar benefits. In yet another embodiment, the bias current and matching circuits are controlled. For example, the bias point of the PA is adjusted and then the matching circuit is adjusted to account for any changes in the PA load line. Of course, as will be evident to those of skill in the art, other methods of achieving control over biasing of the PA and/or control over the matching circuits are also applicable to embodiments of the invention.

(44) Though the invention has been described with reference to a set of potential variables for use in tuning of the amplifier, anyone or more of those variables is independently useful for amplifier control or useful in conjunction with one or more other variables. For example, channel bandwidth is useful in isolation or with modulation type, for example. Similarly, each is useful with more than one other variable or in isolation.

(45) Numerous other embodiments may be envisaged without departing from the scope of the invention.