High power Doherty amplifier using multi-stage modules

This invention is related to U.S. Pat. Nos. 6,700,444; 6,737,922; 6,791,417; and 6,798,295, assigned to Cree Microwave, Inc.

Filed herewith and incorporated by reference for all purposes are a specification sheet for CREE PFM19030 amplifier, a specification sheet for a Doherty amplifier employing the PFM19030 amplifier module, and a paper titled “60 Watt Doherty Amplifiers Using High Gain 2-Stage Hybrid Amplifier Modules.”

This invention relates generally to RF power amplifiers, and more particularly the invention relates to a high power amplifier having improved efficiency and linearity using multiple stage modules.

Power amplifiers in basestations often operate at output power levels much lower than peak power. Unfortunately, the back-off power level reduces the efficiency of the power amplifier in the transmitter. In a conventional amplifier there is a direct relationship between efficiency and the input drive level. Therefore, high efficiency (DC to RF conversion efficiency) is not obtained until the RF input power level becomes sufficiently high to drive the amplifier into compression or saturation. Since in multicarrier communication systems an amplifier must remain as linear as possible, this region of high efficiency cannot be used.

A power amplifier circuit design which provides improved efficiency in back-off power levels is the Doherty amplifier circuit, which combines power from a main amplifier and from a peak amplifier. See, W. H. Doherty, “A New High-Efficiency Power Amplifier for Modulated Waves,” Proc. IRE Vol. 24, No. 9, pp. 1163–1182, 1936. In the conventional Doherty configuration, the main or carrier amplifier 10 and peak amplifier 12 are designed to deliver maximum power with optimum efficiency to a load R, as shown in FIG. 1A . The main or carrier amplifier is a normal Class B amplifier, while the peak amplifier is designed to only amplify signals which exceed some minimum threshold. For an LDMOS power transistor, this can be accomplished by DC biasing the transistor below its pinch-off voltage for operation similar to Class C. The outputs of the two amplifiers are connected by a quarter-wave transmission line of characteristic impedance R, and a load of one-half of the optimum load R is attached to the output of the peak amplifier. The RF input power is divided equally with a quarter-wave delay at the input to the peak amplifier, thus assuring that the output power of the two amplifiers at the load R/2 will be in phase.

Heretofore, the Doherty amplifier has employed discrete single stage amplifiers in the carrier and peak amplifier modules. The present invention realizes advantages not available when using discrete single stage amplifiers.

In accordance with the invention, multi-stage amplifier modules are employed in a Doherty amplifier for both the main amplifier and the peak amplifier or peak amplifiers.

In one embodiment of a two-way two stage amplifier, the first stage of each amplifier module can include signal pre-distortion whereby the first stage compensates for distortion in both of the first and second stages. The design is simple and the results in a high efficiency amplifier with high gain.

The invention and objects and features thereof will be more readily apparent from the following detailed description and appended claims when taken with the drawings.

FIG. 1 is a functional block diagram of a prior art two stage hybrid amplifier module which can be used in an embodiment of the invention.

FIG. 2 is a functional block diagram of a 2-way two stage RF power amplifier employing hybrid amplifier modules of FIG. 1 .

FIG. 3 is a plot of amplifier gain and input and output return loss versus frequency for one embodiment of the invention.

FIG. 4 is a graph illustrating Doherty amplifier and individual class AB module gain and efficiency versus CW power output.

FIG. 5 is a graph of Doherty amplifier gain and AM/PM versus CW power out.

FIG. 6 is a graph illustrating Doherty 2-Tone CW intermodulation products versus average output power.

FIG. 7 is a graph illustrating Doherty amplifier intermodulation products versus average power out at three frequencies.

FIG. 8 is a plot of Doherty amplifier 2-WCMDMA signal ACLR and intermodulation rejection versus average power output.

FIG. 9 is a graph of Doherty amplifier signal ACLR and IM3 rejection versus average power output for three frequencies.

FIG. 10 is illustrates specifications for a Doherty amplifier with pre-distortion for a 12.5 watt 2WCDMA signal spectrum in accordance with an embodiment of the invention.

The present invention provides a high power Doherty amplifier with improved gain and linearity compared to conventional single stage Doherty amplifiers employing discrete transistors. In the present invention, a Doherty amplifier having a main or carrier amplifier and one or more peak amplifiers employs multi-stage amplifier units which can be manufactured in hybrid assemblies.

FIG. 1 is a functional block diagram of a commercially available two stage thick film hybrid microelectronic amplifier (CREE PFM19030SM) which can advantageously be employed in a Doherty amplifier in accordance with an embodiment of the present invention. The first stage of the module includes a field effect transistor Q 1 connected to a RF input through input matching circuitry and pre-distortion circuitry. The output of transistor Q 1 is applied through an output matching network through distributed surface mount technology matching circuitry to the second stage including a second field effect transistor Q 2 . Transistor Q 2 is connected to receive the output of Q 1 through input matching circuitry and applies an amplified output through output matching circuitry to the RF output.

The partitioning of this circuit places the devices with high power dissipation on metal interposers that constitute RF ground and also function as heat spreaders; while matching circuitry and bias injection and bypassing is realized using thick film microstrip circuits including SMT passive elements. The block diagram in FIG. 1 demonstrates the amplifier partitioning.

Q 1 and Q 2 are LDMOS transistors, and each LDMOS transistor die exhibits input or output impedances on the order of 1–5 ohms (before adding chip and-wire matching). The combination of chip-and-wire matching and distributed circuit matching on the output of the 30 W transistor (Q 2 ) transforms the optimum power-match impedance (for Class AB operation) to a nominal 20 Ohm level. This simplifies the off-module matching circuitry required for a Doherty amplifier subsystem.

Each amplifier module assembly includes the two (5 W and 30 W) die carriers soldered to a bottom plate of copper (1.0 mm thick). The silicon LDMOS transistor die (manufactured by Cree Microwave) are eutectically attached to metal interposers. The bottom plate also supports a 0.5 mm thick single-layer alumina thick-film substrate which has cutouts where the die carriers are located. The alumina substrate is attached to the copper base with a conductive epoxy. The heat from the die is spread through the die carrier (interposer) and then through the thick copper base, before it encounters the external interface.

When mounted on a PCB layout, the two-stage surface-mount module occupies only slightly more board space than would a single discrete-package 30 W transistor (with conventional bolt-down metal flange). In addition to the RF input and RF output leads, there are leads accessing the first-stage drain and internal temperature compensation circuitry. Bypass circuits are included internal to the module to support wide video bandwidths and thereby minimize memory effects.

FIG. 2 is a functional block diagram of a 2-way two stage Doherty amplifier in accordance with an embodiment of the present invention which employs two PFM19030SM modules of FIG. 1 . As shown in FIG. 2 , the amplifier includes a main module 20 and a peaking module 22 which receive an RF input signal through splitter 24 . The main module 20 is biased for class AB operation, and the peak amplifier module is biased nominally for class C operation.

Module inputs are connected directly to a 3 dB quadrature hybrid. The outputs are matched using short low impedance transmission line elements and shunt capacitors Cp and Cm. Because the output impedances of the modules are much higher than unmatched discrete LDMOS transistors, the additional matching circuitry is minimized. Cp and Cm are of different values (Cp<<Cm) as is appropriate for the different operating modes (Class C vs. Class AB) of the 30 W devices.

Further impedance transformation is accomplished using transmission lines TL 1 (each side). The main module side TL 2 section is nominally 90°, as is typical of classic 2-Way Doherty designs. The output section TL 3 and associated capacitors constitute an impedance transformer. All element (transmission line and shunt capacitor) values are adjusted in the circuit analysis and optimization process. The validity and applications power of the CMC device models was confirmed by the experience that only capacitor values were adjusted in prototype circuitry to obtain reported results (transmission line lengths and widths were left at turn-on values).

The capacitor values are adjusted primarily to achieve optimum peak power levels. Bias conditions are the most sensitive determinate of amplifier linearity and efficiency. By choosing Class AB & Class C, circuit linearity is optimized at some tradeoff in efficiency. Linearity is critical to the intended applications, in which further correction by pre-distortion can be anticipated. A major objective is to achieve system-level linearity standards with pre-distortion only (avoiding feed-forward losses). This strategy can potentially maximize system efficiency and reduce complexity.

Gain and return loss for the two way, two stage amplifier of FIG. 2 are presented in FIG. 3 . Gain is 26±0.2 dB over 1930–1960 MHz. Measured gain and efficiency versus CW output power are compared for the Doherty amplifier and an individual module operated with normal Class AB bias in FIG. 4 . The Class AB module has a peak output capability on nominally 30 Watts (+44.8 dam), whereas the Doherty output is nominally 60 Watts (47.8 dam). Note that both circuits have device quiescent bias levels adjusted for optimum linearity, not for optimum peak power or for efficiency. The comparison of the shape of the efficiency versus output power characteristics is of particular interest. Even though the Doherty amplifier has twice the output power capability, its efficiency is similar to that of the individual module at low power levels.

FIG. 5 presents measured gain and relative phase versus output power (AM/PM) for the Doherty amplifier. The very low AM/PM contributes to 2-tone CW and WCDMA linearity.

A characteristic of this 2-Way Doherty amplifier is the degradation of linearity as one deviates from the band center frequency. This is thought to be typical of 2-Way Doherty amplifiers in general. Center-band and band-edge measurements of CW 2-Tone 3rd order IMD products are included in FIG. 7 . The 2-tone CW Imps show more dependence with frequency than do the 2-WCDMA signal tests.

Standard WCDMA testing involved two signals separated by 10 MHz (3GPP with 8.5 dB peak-to-average), centered at 1960 MHz. Measurements ( FIG. 8 ) show that the IM3 products at 10 MHz offsets tend to be the dominant distortion, and a degree of IM3 asymmetry is evident at lower power levels (this also occurs for the individual modules operated in standard Class AB conditions). ACLR and IM3 rejection are plotted at three RF frequencies in FIG. 9 . Measurement system dynamic range is limited to about −55 dab at low power(˜−60 dab at higher power levels).

A key application objective is to further improve linearity by use of pre-distortion. FIG. 10 shows 2-WCDMA signal data using a PMC-Sierra Paladin 15 digital pre-distortion to enhance the Doherty amplifier linearity.

The signal in this case is two WCDMA signals with crestfactor reduction to 7.5 dB. Average output power is 12.5 Watts, and efficiency is 26.8%, with ACLR & IM3 at −51 & −54 dBc. Efficiency across the RF band varies from 28% to 26% (1930–1990), and ACLR is ≦−49 dBc and IM3 is ≦−50 dBc. IM3 asymmetry is very small after application of pre-D. This demonstrates excellent linearity and efficiency using this Doherty design in conjunction with pre-distortion.

The two-way two stage Doherty amplifier in FIG. 2 uses small surface-mount hybrid modules as the active elements in a 2-way 60 W Doherty amplifier. The design demonstrates good efficiency (26%) for two 3GPP WCDMA signals at 10 W average output, with ACLR of −40 dBc and IM3 of −38 dBc (uncorrected). When augmented by pre-distortion, 12.5 W of WCDMA is produced at 26% efficiency & ACLR/IM3 rejections of −49 dBc/−50 dBc across the full 1930–1990 MHz band. The associated 26 dB gain and low input return loss simplifies system design.

While the invention has been described with reference to a specific embodiment, the description is illustrative of the invention and is not to be construed as limiting the invention. For example, the transistors can be lateral DMOS silicon field effect transistors, MESFETs, HEMTs, HBTs, and bipolar transistors. Further, the invention has applicability to amplifiers having more than one peak amplifier and using modules with two or more stages. For example, a three way two stage amplifier using three CREE PFM19030 modules has been simulated for producing over 90 watts of single tone power with overall gain of 24 dB. Thus, various modifications and applications may occur to those skilled in the art without departing from the true spirit and scope of the invention as defined by the appended claims.