Methods for tuning an adaptive impedance matching network with a look-up table

[0001] The present invention relates generally to methods for tuning an impedance matching network, and more specifically to methods for tuning an impedance matching network with a look-up table.

[0002] Electronic communication products which transmit and receive radio frequency (RF) signals need to match the impedances between the internal circuitry, such as an RF power amplifier, and the antenna for optimum performance of the product. The optimum mode may also depend upon the modulation scheme, the frequency of transmission and other considerations.

[0003] Electronically tunable filters are frequently used to compensate for any impedance mismatch between the circuitry and the antenna. Such tunable filters utilize various tuning elements, such as tunable capacitors, varactors, micro-electronic mechanical systems (MEMS), and doped semiconductor materials.

[0004] Adaptive impedance matching modules (AIMMs) sense continually sense impedance mismatch and retune the impedance for any changed conditions. AIMMs may take up to 20 iterative steps to converge on the best impedance match, such as by using gradient search methods.

[0005] There is a need for an adaptive impedance matching network which can rapidly tune to the best or optimum matching impedance in fewer steps and in less time.

[0006] The present invention is directed to methods for generating and using a look-up table relating a plurality of complex reflection coefficients to a plurality of matched states for a tunable matching network. In one embodiment, typical steps of the methods include measuring a plurality of complex reflection coefficients resulting from a plurality of impedance loads while the tunable matching network is in a predetermined state, determining a plurality of matched states for the plurality of impedance loads, with a matched state determined for each of the plurality of impedance loads, and providing the determined matched states in a look-up table. These steps may be reiterated for different frequency bands, different frequencies or for different use cases.

[0007] In other embodiments, the methods may include additional steps such as interpolating the measured complex reflection coefficients and the determined matching states into a set of complex reflection coefficients with predetermined step sizes, selecting the predetermined state used in measuring the complex reflection coefficients to minimize tolerance variations based upon temperature coefficient, component tolerance or tolerance over time, using the determined matched states in the look-up table to tune the tunable matching network, tuning the tunable matching network to adaptively match the impedance of an antenna, controlling the tunable impedance elements with digital to analog converters.

[0008] The tunable matching network may include a plurality of tunable impedance elements, which may be, for example, ferroelectric capacitors, voltage variable capacitors, MEMS, tunable inductors or networks thereof.

[0009] Yet another embodiment includes methods for generating a look-up table relating a set of measured parameters to a set of matched states for a tunable matching network, including the steps of measuring a set of parameters associated with a plurality of impedance loads while the tunable matching network is in a predetermined state, determining a matched state for each of the plurality of impedance loads, and providing the determined matched states as a look-up table. The set of measured parameters may be selected from the group consisting of complex reflection coefficients, current drain, incident power, reflected power, control setting of the tunable matching network, temperature, input power level, reliability considerations and linearity considerations. Preferably, a sufficient number of matched states are provided such that a transition from one matched state to an adjacent matched state provides a smooth transition.

[0010] A further embodiment may include methods of adaptively tuning a tunable impedance matching network with a lookup table, the lookup table containing a plurality of complex reflection coefficients corresponding to a plurality of impedance mismatches between an input terminal and an output terminal of the tunable impedance matching network, including the steps of determining the impedance mismatch between the input and output terminals of the tunable impedance matching network, determining the closest impedance mismatch in the lookup table, and using the complex reflection coefficients in the lookup table which correspond to the determined closest impedance mismatch to tune the tunable impedance matching network. Further steps may include controlling the ferroelectric capacitors with digital to analog converters, and tuning the tunable impedance matching network to adaptively match the impedance of an antenna.

[0011] Another embodiment may include methods of adaptively tuning a tunable impedance matching network with a lookup table, the lookup table containing sets of parameters corresponding to matched states for the tunable impedance matching network, including the steps of selecting a parameter from the group consisting of complex reflection coefficients, current drain, incident power, reflected power, control setting of the tunable matching network, temperature, input power level, reliability considerations and linearity considerations, determining the closest match to the selected parameter in the lookup table, and using the closest match to the selected parameter in the lookup table to tune the tunable impedance matching network. The parameter may also be selected to minimize tolerance variations based upon temperature coefficient, component tolerance or tolerance over time.

[0019] It will be understood that the present invention may be embodied in other specific forms without departing from the spirit thereof. The present examples and embodiments, therefore, are to be considered in all respects as illustrative and not restrictive, and the invention is not to be limited to the details presented herein.

[0020] FIG. 1 illustrates an impedance matching circuit, generally designated 100, for impedance optimization between a source of radiofrequency (RF) power, such as an RF amplifier 102, and an RF load, such as an antenna 104. An adaptive impedance matching network 106 is coupled between RF amplifier 102 and antenna 104. In the example of FIG. 1, the adaptive impedance matching network 106 is an adaptive impedance matching module (AIMM). AIMM 106 dynamically adjusts its internal impedance matching circuit to minimize reflected power to achieve a near-optimal impedance match. For example, a common application for AIMM 106 is to correct antenna-RF amplifier impedance mismatch that often occurs with handheld radios and body-worn antennas. Similarly, handheld communications devices, such as cellular telephones encounter impedance mismatch when the communication device is held close to the user's head or torso. Yet another application for the AIMM 106 is in sensor networks where the antenna of the sensor can be detuned by proximity effects of the ground, foliage or debris. In general, AIMM 106 may be used in any application where forward power needs to be maximized and reverse power needs to be minimized.

[0021] AIMM 106 may be a multi-chip module comprising a tunable impedance network, which contains one or more tunable ferroelectric capacitors 108-109. Preferably, the tunable ferroelectric capacitors 108-109 are Paratek's ParaTune™ family of passive tunable integrated circuits (PTICs), which are commercially available from Paratek Microwave, Inc. of Columbia, Md. These PTICs 108-109 utilize a Parascan® tunable dielectric material which is further described in U.S. Pat. Nos. 7,107,033 and 6,514,895, which are assigned to the same assignee as the present invention and which are incorporated herein by reference in their entirety. These PTICs 108-109 overcome the power limitations common to other tunable technologies such as varactor diodes and MEMS and can handle in excess of 2 watts of RF power with extremely low inter-modulation distortion.

[0022] The adaptive impedance matching module or AIMM 106 in FIG. 1 has an input terminal 110 for receiving an RF signal from RF amplifier 102. A forward and reverse power detector 112 provides a signal on line 114 to a rectifier 116 which is representative of the forward power from RF amplifier 102. Rectifier 116 provides a rectified value of the forward RF power to an analog to digital converter (ADC) 118, which in turn provides a digital representation of the forward power to a microprocessor 120. In a similar fashion, the power detector 112 provides a signal on line 115 to a rectifier 117 which is representative of the reverse or reflected power from antenna 104. Rectifier 117 provides a rectified value of the reverse RF power to an ADC 119, which in turn provides a digital representation of the reverse power to a microprocessor 120.

[0023] Based upon the values of the determined forward and reverse power levels, and in accordance with one aspect of the present invention, microprocessor 120 uses a lookup table, such as table 400 in FIG. 4 or table 600 in FIG. 6, which may be resident in memory 124 to determine bias adjustments to be made to the PTICs 108-109 for a first step of retuning the adaptive impedance of the AIMM 106. To this end, microprocessor 124 supplies digital signals to digital to analog converters (DACs) 122-123 which control the analog bias signals to PTICs 108-109. Microprocessor 120 may continue with additional steps of retuning the PTICs, as needed, to provide a near-optimum impedance matching between RF amplifier 102 and antenna 104.

[0024] FIG. 2 illustrates a table, generally designated 200, which is an example of how independent variables 202 may be defined and organized. For example, the independent variables 202 may include frequency information 204, reflection coefficient magnitude 205 and reflection coefficient phase 206. Furthermore, each of these independent variables may have a symbol 208, a normalized independent variable 210, a normalized symbol 212 and a number of points 214.

[0025] As shown in the example of FIG. 2, the frequency information 204 may be associated with a symbol Fq and a normalized symbol Fq, and may consist of 12 values including one value for each transmit frequency band and one value for each receive frequency band. The frequency variable for the frequency information 204 may also be accorded 12 points. Similarly, the reflection coefficient magnitude 205 has a symbol mag_S11, a normalized symbol N_mag_S11, and a normalized independent variable determined as mag_S11 times 8 and then rounded to the nearest integer. The reflection coefficient magnitude 205 is accorded 6 points. Lastly, the reflection coefficient phase 206 has a symbol ph_S11, a normalized symbol N_ph_S11, and a normalized independent variable determined as ph_S11 divided by 45 and then rounded to the nearest integer. The reflection coefficient phase 106 is accorded 8 points.

[0026] The purpose of a look-up table, such as look-up table 400 in FIG. 4 or look-up table 600 in FIG. 6 is to make it possible for an adaptive impedance matching network, such as adaptive impedance matching module 106 in FIG. 1, to take a large first step in tuning. A large first tuning step reduces the time to convergence of final settings, reduces power consumption and decreases traffic on the communication bus.

[0027] In accordance with one embodiment of the present invention, the look-up table 400 may contain pairs (or sets) of digital-to-analog (DAC) settings 402 that are to be put into the high voltage application specific integrated circuit (HV-ASIC) for controlling tunable impedance elements in the adaptive impedance matching module 106. The pairs (or sets) of DAC settings 402 can be identified by an index which is correlated to independent variables 202, for example, frequency 204, magnitude S11 205 and phase S11 206. The index would be related to the position in the table 400 of the desired information. Hence, the index information would not need to be stored in the table. The index could be the sum of three independent variables, such as independent variables 202.

[0028] In this example, the address pointer, such as address pointer 602 in FIG. 6, may be calculated as pointer=6*3*Fq+8*N_mag_S11+N_ph_s11. The table then has 432 rows (=12*6*8). Each has 2 or 3 bytes of information, one for each tunable element setting. The total memory usage for a two tunable element adaptive impedance network is 1152 bytes. The total memory usage for an adaptive impedance matching network with 3 tunable elements is 1728 bytes in this example.

[0029] A worst case analysis may typically be required to determine how many different phases and magnitudes are sufficient or insufficient. A key consideration is how accurate the first step needs to be and how much the tolerances may degrade the accuracy. Also, the magnitude and phase of S11 do not have to be scaled linearly. They could be scaled non-linearly to give better accuracy to areas of the Smith chart that are common.

[0030] The operation of the adaptive impedance matching module 106 with the look-up table 400 will now be considered. When the adaptive impedance matching module 106 is initially turned-on, it may be programmed to an initial or default state. From the initial state, adaptive impedance matching module 106 will measure the magnitude and phase of the reflection coefficient 205 and 206, look-up the DAC setting 402 in table 400 that corresponds to that reflection coefficient, and take a large first tuning step. The first tuning step can be improved by interpolating between table entries. The first step may be broken into several steps if the modulation accuracy or time mask specifications do not allow for a full step. The default setting may have the following characteristics: a) all tunable elements may be set to the same voltages, and b) the voltage that the tunable elements are set to would correspond to the voltage at which the tolerances have minimal impact, for example: a temperature coefficient that is O ppm/C. Many tolerances may degrade the accuracy of the first step. Using the O ppm/C temperature coefficient will reduce the contribution of temperature to the tolerance stack-up. The default setting could also be chosen to correspond to the expected load impedance, such as the load-pull system 306, if known.

[0031] The magnitude, phase and band information will then be used to calculate the index of the look-up table that represents the best first guess for the first tuning step. Thereafter, adaptive impedance matching network 304 may switch to its standard adaptive search algorithm. Once optimal tuning is achieved, adaptive impedance matching network 304 can switch to a tracking mode, in which step sizes are smaller and/or less often.

[0032] Consideration will now be given to generating the look-up table, such as tables 400 and 600. These tables may be generated during the design phase after the topology and the tuner values are fixed. A block diagram of the characterization test set-up 300 is shown in FIG. 3. it consists of a signal generator 302 coupled via a conductor 303 to the adaptive impedance matching network 304, which in turn, is coupled via a conductor 305 to a load-pull system 306. This test set-up 300 simulates an end use application of the adaptive impedance matching network 304, in which signal generator 302 may be an RF power amplifier, for example, in a portable communication device such as a cellular telephone, personal digital assistant, or the like. In such an end use application, the load-pull system 306 may be the RF antenna for the portable communication device.

[0033] The procedure to generate the table 400 begins with the adaptive impedance matching network 304 in a pause mode, and the tunable element in the adaptive impedance matching network 304 set to the default setting. This may be the bias voltage at which the capacitive temperature coefficient is O ppm/c. Note that different settings may be preferable if it results in improved tolerance performance. For example, if there is knowledge of the expected load, the default setting could also be set based on the expected load, such as presented by the load-pull system 306.

[0034] The signal generator is preferably set to the middle of the frequency range being characterized. As the design is better understood, the designer may choose to use a different frequency that better represents the center of performance. If frequency information is available, a look-up table could be generated for each frequency, as the availability of memory allows or permits.

[0035] The load-pull system 306 is run through a fine mesh of the magnitudes and phases. The magnitude and phase of S11 is recorded as measured by adaptive impedance matching network 304 for each load. Alternatively, other parameters could be recorded for use in the table. These parameters may include (but are not limited to) complex reflection coefficient, current drain, incident power, reflected power, reliability metrics, linearity metrics, and the like.

[0036] The adaptive impedance matching module 106 is set in the run mode. That is the adaptive impedance matching module 106 is permitted to optimize the match as it would in an end use application. The optimization may be part of the module 106 or apart from it. If the optimization is apart from the module 106, it may be referred to as a tunable impedance matching network. If the optimization is part of the network, it may be referred to as an adaptive impedance matching network. An adaptive impedance matching module 106 always includes a tunable impedance matching network.

[0037] As the load-pull system 306 runs through a fine mesh of the magnitudes 404 and phases 405, the DAC settings 402 that adaptive impedance matching network 304 settles on for each load 306 are recorded, as by storage in available memory. At this point, all of the information is available for a look-up table, such as look-up table 400 in FIG. 4.

[0038] Using the two data sets 304 and 305, generate contours correlating the S11 measurements to the preferred DAC settings 302. The contours are then used to interpolate to find the preferred or interpolated DAC settings 404 in table 400 (FIG. 4) for each S11 that is to be represented in the look-up table 400.

[0039] At this point, the look-up table would be like that of table 600 in FIG. 6. This is an extremely memory efficient table implementation because only the output interpolated DAC settings 604 need to be stored. An address pointer 602 will retrieve the applicable interpolated DAC settings 604. The interpolated DAC settings 604 are preferably organized such that no searching is required, thereby saving processing time.

[0040] It will be appreciated by those skilled in the art that the above steps of generating the look-up tables will be performed by a microprocessor or the like. For example, the above steps may be performed by a microprocessor in the product for which the antenna matching is occurring, such as in a cellular telephone, PDA, or the like. Alternately, the microprocessor may be provided in the adaptive impedance matching module 106.

[0041] While particular embodiments of the invention have been shown and described, it will be obvious to those skilled in the art that changes and modifications may be made therein without departing from the invention in its broader aspects.