Method and device for broadband quadrature calibration of a radio frequency transceiver and radio frequency transceiver

Abstract

The application discloses a wideband quadrature calibration method and device of a radio frequency transceiver and the radio frequency transceiver, relates to the technical field of radio frequency transceivers, and comprises the following steps: controlling injection of a preset in-band single-tone test signal and a preset local oscillator signal, acquiring an I / Q path in-band power value, calculating a difference value, and compensating the frequency-independent IQ gain imbalance in the digital domain through digital gain compensation. Controlling injection of a preset attenuation point test signal and a local oscillator signal, acquiring an I / Q path cutoff frequency point power value, and combining the I / Q path in-band power value to determine the double-frequency point power difference of the I / Q path. According to the deviation of the double-frequency point power difference of the I / Q path and a target attenuation amount, the cutoff frequency of the I / Q path filter is adjusted respectively, and the frequency-dependent IQ imbalance is calibrated in the analog domain. The application replaces the quadrature calibration signal with a single-tone signal for testing, expands the application scenario, simplifies the wideband IQ balance calculation in the digital domain, and reduces the hardware resource consumption and time delay.

Description

Technical Field

[0001] This application relates to the field of radio frequency transceiver technology, and in particular to a broadband orthogonal calibration method and apparatus for radio frequency transceivers and radio frequency transceivers. Background Technology

[0002] Modern wireless communication products pursue low-cost, highly integrated circuit designs, widely employing zero-IF transceiver architectures. While this avoids the use of expensive intermediate frequency (IF) filters, the trade-off is extreme sensitivity to amplitude and phase imbalances between in-phase and quadrature paths. Even minute process deviations during chip manufacturing or slight asymmetries in layout routing can trigger severe image interference. This can lead to a significant deterioration in sensitivity and receive bit error rate when blocking signals appear in the image band during reception; and a deterioration in transmit error vector amplitude during transmission, even causing transmit spurious emissions to fail to meet electromagnetic compatibility (EMC) regulations.

[0003] In existing technologies, image mismatch between the in-phase and quadrature paths is typically calibrated using digital domain calibration. However, digital domain calibration can only compensate for narrowband or fixed-frequency mismatches and cannot effectively compensate for broadband signals. Furthermore, it usually requires the calibration test signal to possess specific orthogonal characteristics, which limits its practical application. To address the broadband compensation problem, a digital algorithm based on Gram-Schmidt orthogonalization has been proposed. This algorithm uses a complex finite impulse response (FIR) filter to model and compensate for the imbalance between the in-phase and quadrature paths. While it can achieve broadband quadrature error calibration, constructing a high-order complex FIR filter requires significant hardware resources and introduces time mismatch errors between the in-phase and quadrature signals, as well as high processing latency, failing to meet the engineering requirements for low power consumption and low latency.

[0004] The above content is only used to help understand the technical solution of this application and does not represent an admission that the above content is prior art. Summary of the Invention

[0005] The main objective of this application is to provide a broadband orthogonal calibration method for radio frequency transceivers, aiming to solve the problems of high hardware resource consumption, high latency, and limited application scenarios of existing broadband orthogonal calibration methods.

[0006] To achieve the above objectives, the broadband orthogonal calibration method for radio frequency transceivers proposed in this application includes: A preset in-band single-tone test signal and a preset local oscillator signal are injected and processed by the receiving path of the RF transceiver to obtain the I-channel in-band power value output by the I-channel filter and the Q-channel in-band power value output by the Q-channel filter of the RF transceiver, respectively. The frequency of the preset local oscillator signal is a first preset frequency, and the frequency of the preset in-band single-tone test signal is the sum of a second preset frequency and the first preset frequency. The second preset frequency is within the common passband of the I-channel filter and the Q-channel filter. Calculate the IQ difference between the in-band power value of the I-channel and the in-band power value of the Q-channel, and perform digital gain compensation on the I-channel and / or Q-channel of the RF transceiver based on the IQ difference, so that the gains of the I-channel and Q-channel of the RF transceiver are the same. The preset attenuation point test signal and the preset local oscillator signal are controlled and injected. After being processed by the receiving path, the power values ​​at the I-channel cutoff frequency point output by the I-channel filter and the power values ​​at the Q-channel cutoff frequency point output by the Q-channel filter are obtained respectively. The I-channel filter and the Q-channel filter have the same preset cutoff frequency attenuation point, and the frequency of the preset attenuation point test signal is the sum of the third preset frequency and the first preset frequency. The third preset frequency is the frequency at the cutoff frequency attenuation point. The power difference between the I-channel dual-frequency points is determined based on the I-channel in-band power value and the I-channel cutoff frequency point power value, and the power difference between the Q-channel dual-frequency points is determined based on the Q-channel in-band power value and the Q-channel cutoff frequency point power value. Obtain the target attenuation amount corresponding to the cutoff frequency attenuation point, and adjust the cutoff frequency of the I-channel filter according to the first deviation between the power difference of the I-channel dual-frequency points and the target attenuation amount; and adjust the cutoff frequency of the Q-channel filter according to the second deviation between the power difference of the Q-channel dual-frequency points and the target attenuation amount.

[0007] In one embodiment, adjusting the cutoff frequency of the I-channel filter based on the first deviation between the power difference of the I-channel dual-frequency points and the target attenuation includes: If the power difference between the two frequency points of the I-channel is less than the target attenuation, then the cutoff frequency of the I-channel filter is reduced. If the power difference between the two frequency points of the I-channel is greater than the target attenuation, then the cutoff frequency of the I-channel filter is increased.

[0008] In one embodiment, adjusting the cutoff frequency of the Q-channel filter based on the second deviation between the power difference at the Q-channel dual-frequency points and the target attenuation includes: If the power difference between the two frequency points of the Q-channel is less than the target attenuation, then the cutoff frequency of the Q-channel filter is increased. If the power difference between the two frequency points of the Q-channel is greater than the target attenuation, then the cutoff frequency of the Q-channel filter is reduced.

[0009] In one embodiment, the I-channel filter includes a first adjustable impedance array for adjusting the cutoff frequency of the I-channel filter, and the Q-channel filter includes a second adjustable impedance array for adjusting the cutoff frequency of the Q-channel filter. Adjusting the cutoff frequency of the I-channel filter includes: Adjust the equivalent capacitance and / or equivalent resistance of the first adjustable impedance array to change the RC time constant of the first adjustable impedance array; Adjusting the cutoff frequency of the Q-channel filter includes: Adjust the equivalent capacitance and / or equivalent resistance of the second adjustable impedance array to change the RC time constant of the second adjustable impedance array.

[0010] In one embodiment, the first adjustable impedance array / second adjustable impedance array is an N-bit binary weighted capacitor array, the N-bit binary weighted capacitor array comprising: A basic branch consisting of a fixed resistor and a fixed capacitor connected in parallel; And a switchable capacitor array consisting of N-bit binary weighted capacitor units, where the value of the k-th capacitor is... Each capacitor unit is connected in parallel to both ends of the basic branch through an independent switch; The equivalent capacitance value of the adjustable impedance array satisfies:

[0011] in, This represents the capacitance value of the capacitor unit. For a fixed capacitance value, 1 ≤ k ≤ N, 0 ≤ n ≤ And k and n are both integers; Adjusting the equivalent capacitance value of the first adjustable impedance array / the second adjustable impedance array includes: A successive approximation algorithm is used to control the on / off state of each independent switch in the N-bit binary weighted capacitor array in order from the most significant bit to the least significant bit.

[0012] This application also proposes a broadband orthogonal calibration device for a radio frequency transceiver, comprising: The injection control and power acquisition module is used to control the injection of a preset in-band single-tone test signal and a preset local oscillator signal. After processing by the receiving path of the RF transceiver, the module acquires the I-channel in-band power value output by the I-channel filter and the Q-channel in-band power value output by the Q-channel filter of the RF transceiver, respectively. The frequency of the preset local oscillator signal is a first preset frequency, and the frequency of the preset in-band single-tone test signal is the sum of a second preset frequency and the first preset frequency. The second preset frequency is within the common passband of the I-channel filter and the Q-channel filter. A digital gain compensation module is used to calculate the IQ difference between the in-band power value of the I-channel and the in-band power value of the Q-channel, and to perform digital gain compensation on the I-channel and / or Q-channel of the RF transceiver based on the IQ difference, so that the gains of the I-channel and Q-channel of the RF transceiver are the same. The injection control and power acquisition module is also used to control the injection of a preset attenuation point test signal and the preset local oscillator signal. After processing by the receiving path, the power values ​​at the I-channel cutoff frequency point output by the I-channel filter and the power values ​​at the Q-channel cutoff frequency point output by the Q-channel filter are obtained respectively. The I-channel filter and the Q-channel filter have the same preset cutoff frequency attenuation point, and the frequency of the preset attenuation point test signal is the sum of a third preset frequency and a first preset frequency. The third preset frequency is the frequency at the cutoff frequency attenuation point. The cutoff frequency adjustment module is used to determine the power difference of the I-channel dual-frequency point based on the I-channel in-band power value and the I-channel cutoff frequency point power value, and to determine the power difference of the Q-channel dual-frequency point based on the Q-channel in-band power value and the Q-channel cutoff frequency point power value. In addition, it is used to obtain the target attenuation amount corresponding to the cutoff frequency attenuation point, and adjust the cutoff frequency of the I-channel filter according to the first deviation between the power difference of the I-channel dual-frequency points and the target attenuation amount; and adjust the cutoff frequency of the Q-channel filter according to the second deviation between the power difference of the Q-channel dual-frequency points and the target attenuation amount.

[0013] In one embodiment, when the cutoff frequency adjustment module adjusts the cutoff frequency of the I-channel filter based on the first deviation between the power difference of the I-channel dual-frequency points and the target attenuation, it is specifically used for: If the power difference between the two frequency points of the I-channel is less than the target attenuation, then the cutoff frequency of the I-channel filter is reduced. If the power difference between the two frequency points of the I-channel is greater than the target attenuation, then the cutoff frequency of the I-channel filter is increased. When the cutoff frequency adjustment module adjusts the cutoff frequency of the Q-channel filter based on the second deviation between the power difference of the Q-channel dual-frequency points and the target attenuation, it is specifically used for: If the power difference between the two frequency points of the Q-channel is less than the target attenuation, then the cutoff frequency of the Q-channel filter is increased. If the power difference between the two frequency points of the Q-channel is greater than the target attenuation, then the cutoff frequency of the Q-channel filter is reduced.

[0014] In one embodiment, the I-channel filter includes a first adjustable impedance array for adjusting the cutoff frequency of the I-channel filter, and the Q-channel filter includes a second adjustable impedance array for adjusting the cutoff frequency of the Q-channel filter. When adjusting the cutoff frequency of the I-channel filter, the cutoff frequency adjustment module is specifically used to: adjust the equivalent capacitance value and / or equivalent resistance value of the first adjustable impedance array to change the RC time constant of the first adjustable impedance array. When adjusting the cutoff frequency of the Q-channel filter, the cutoff frequency adjustment module is specifically used to: adjust the equivalent capacitance and / or equivalent resistance of the second adjustable impedance array to change the RC time constant of the second adjustable impedance array.

[0015] In one embodiment, the first adjustable impedance array / second adjustable impedance array is an N-bit binary weighted capacitor array, the N-bit binary weighted capacitor array comprising: A basic branch consisting of a fixed resistor and a fixed capacitor connected in parallel; And a switchable capacitor array consisting of N-bit binary weighted capacitor units, where the value of the k-th capacitor is... Each capacitor unit is connected in parallel to both ends of the basic branch through an independent switch; The equivalent capacitance value of the adjustable impedance array satisfies:

[0016] in, This represents the capacitance value of the capacitor unit. For a fixed capacitance value, 1 ≤ k ≤ N, 0 ≤ n ≤ And k and n are both integers; Adjusting the equivalent capacitance value of the first adjustable impedance array / the second adjustable impedance array includes: A successive approximation algorithm is used to control the on / off state of each independent switch in the N-bit binary weighted capacitor array in order from the most significant bit to the least significant bit.

[0017] This application also proposes a radio frequency transceiver, including a first phase-locked loop (PLL), a second PLL, a quadrature mixer, an I-channel filter, a Q-channel filter, a first analog-to-digital converter (ADC), and a second ADC, as well as a broadband quadrature calibration device including the radio frequency transceiver described above; the quadrature mixer is connected to the output terminals of the first PLL, the second PLL, the I-channel filter, and the Q-channel filter, respectively; the output terminal of the I-channel filter is connected to the input terminal of the first ADC, and the output terminal of the Q-channel filter is connected to the input terminal of the second ADC; the output terminal of the first ADC, the controlled terminal of the first PLL, and the controlled terminal of the second PLL are respectively connected to the injection control and power acquisition module; the controlled terminals of the I-channel filter and the Q-channel filter are connected to the cutoff frequency adjustment module.

[0018] This application employs a broadband orthogonal calibration method for RF transceivers, including: Step S10, controlling the injection of a preset in-band single-tone test signal and a preset local oscillator signal. After processing by the receiving path of the RF transceiver, the in-band power values ​​of the I-channel filter and the Q-channel filter are obtained respectively. The frequency of the preset local oscillator signal is a first preset frequency, and the frequency of the preset in-band single-tone test signal is the sum of a second preset frequency and the first preset frequency. The second preset frequency is within the common passband of the I-channel filter and the Q-channel filter. Step S20, calculating the IQ difference between the I-channel in-band power value and the Q-channel in-band power value, and performing digital gain compensation on the I-channel and / or Q-channel of the RF transceiver based on the IQ difference, so that the gains of the I-channel and Q-channel of the RF transceiver are the same. Thus, this application calibrates the frequency-independent IQ gain imbalance in the digital domain. Step S30: Control the injection of a preset attenuation point test signal and a preset local oscillator signal. After processing by the receiving path, obtain the I-channel cutoff frequency power value output by the I-channel filter and the Q-channel cutoff frequency power value output by the Q-channel filter, respectively. The I-channel filter and the Q-channel filter have the same preset cutoff frequency attenuation point, and the frequency of the preset attenuation point test signal is the sum of the third preset frequency and the first preset frequency, where the third preset frequency is the frequency at the cutoff frequency attenuation point. In this way, the single-tone test signal is used as the calibration test signal, eliminating the dependence on orthogonal calibration signals and expanding the application scenarios. Step S40: Determine the I-channel dual-frequency power difference based on the I-channel in-band power value and the I-channel cutoff frequency power value, and determine the Q-channel dual-frequency power difference based on the Q-channel in-band power value and the Q-channel cutoff frequency power value. Step S50: Obtain the target attenuation amount corresponding to the cutoff frequency attenuation point, and adjust the cutoff frequency of the I-channel filter according to the first deviation between the power difference of the I-channel dual-frequency points and the target attenuation amount; and adjust the cutoff frequency of the Q-channel filter according to the second deviation between the power difference of the Q-channel dual-frequency points and the target attenuation amount. In this way, frequency-related IQ imbalance is calibrated in the analog domain. This application eliminates the dependence on orthogonal calibration signals, expands application scenarios, simplifies the complex broadband IQ balance calculation in the digital domain, and reduces the hardware resource consumption and latency for implementing broadband IQ imbalance calibration of RF transceivers. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on the structures shown in these drawings without creative effort.

[0020] Figure 1 A schematic diagram of a structure of an embodiment of the broadband orthogonal calibration method for a radio frequency transceiver provided in this application; Figure 2 A schematic diagram of another embodiment of the broadband orthogonal calibration method for the radio frequency transceiver provided in this application; Figure 3 A schematic diagram of a structure of an embodiment of the broadband orthogonal calibration device for a radio frequency transceiver provided in this application; Figure 4 Provided for this application Frequency response diagram; Figure 5 Provided for this application Frequency response diagram; Figure 6 An electronic circuit diagram of an adjustable impedance array for an embodiment of the broadband orthogonal calibration method for a radio frequency transceiver provided in this application.

[0021] The realization of the purpose, functional features and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0022] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of the embodiments. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of this application.

[0023] It should be noted that all directional indicators (such as up, down, left, right, front, back, etc.) in the embodiments of this application are only used to explain the relative positional relationship and movement of each component in a certain specific posture (as shown in the figure). If the specific posture changes, the directional indicator will also change accordingly.

[0024] Furthermore, the use of terms such as "first" and "second" in this application is for descriptive purposes only and should not be construed as indicating or implying their relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. Additionally, the technical solutions of the various embodiments can be combined with each other, but only on the basis of being achievable by those skilled in the art. When the combination of technical solutions is contradictory or impossible to implement, such a combination of technical solutions should be considered non-existent and not within the scope of protection claimed in this application.

[0025] In existing technologies, image mismatch (i.e., IQ imbalance) between in-phase and quadrature signals can be calibrated using both analog and digital domain methods. Image mismatch refers to the phenomenon in zero-IF transceiver architectures where the amplitude of the I / Q local oscillator signals is unequal or their phases are not strictly 90° orthogonal, resulting in the inability to completely suppress image frequency interference. However, digital domain calibration methods can only compensate for narrowband or fixed-frequency mismatches and cannot effectively compensate for broadband signals. Furthermore, they typically require the calibration test signal to possess specific orthogonal characteristics, limiting their practical application. To address the broadband compensation problem, a digital algorithm based on Gram-Schmidt orthogonalization has been proposed. This algorithm uses a complex finite impulse response (FIR) filter to model and compensate for the imbalance between in-phase and quadrature signals. While it can achieve broadband quadrature error calibration, constructing a high-order complex FIR filter requires significant hardware resources and introduces time mismatch errors between the in-phase and quadrature signals, as well as high processing latency, failing to meet the engineering requirements of low power consumption and low latency.

[0026] This application proposes a broadband orthogonal calibration method for radio frequency transceivers.

[0027] Please see Figure 1 In the first embodiment of this application, the broadband orthogonal calibration method for the radio frequency transceiver includes steps S10 to S50: Step S10: Control the injection of a preset in-band single-tone test signal and a preset local oscillator signal. After processing by the receiving path of the RF transceiver, obtain the I-channel in-band power value output by the I-channel filter and the Q-channel in-band power value output by the Q-channel filter of the RF transceiver respectively. The frequency of the preset local oscillator signal is the first preset frequency, and the frequency of the preset in-band single-tone test signal is the sum of the second preset frequency and the first preset frequency. The second preset frequency is within the common passband of the I-channel filter and the Q-channel filter.

[0028] It should be noted that existing technologies typically require injecting calibration test signals with orthogonal characteristics for testing, in order to facilitate the separation of broadband IQ imbalance parameters through calculation, thus limiting their application scenarios. In this embodiment, a preset in-band single-tone test signal is controlled as the first calibration test signal. This preset in-band single-tone test signal is represented as a discrete frequency point in the frequency domain, and its frequency is the sum of the second preset frequency and the first preset frequency, with the second preset frequency falling within the common passband of the I-path filter and the Q-path filter. The first preset frequency is set to... The second preset frequency is The frequency of the preset in-band single-tone test signal is then... You can set The frequency is set to a low value, so that the obtained I-band power value and Q-band power value are approximately DC power, which facilitates subsequent calculations.

[0029] In one implementation, please refer to Figure 3 The quadrature mixer inside the RF transceiver is connected to the output of the low-noise amplifier via a switching circuit. The first analog-to-digital converter and the second analog-to-digital converter inside the RF transceiver are connected to the I-channel filter and the Q-channel filter, respectively, to convert the analog signals output from the I-channel filter and the Q-channel filter into digital signals for subsequent analysis and calculation. Before step S10, the broadband quadrature calibration method for this RF transceiver also includes: Step S01: Control the disconnect switch circuit.

[0030] Step S02: Control the first analog-to-digital converter and the second analog-to-digital converter to perform corrections respectively to remove their respective DC offset errors.

[0031] It should be noted that the receiving path of the RF transceiver includes an orthogonal mixer, an I-channel filter, a Q-channel filter, a first analog-to-digital converter, and a second analog-to-digital converter.

[0032] In this embodiment, the connection of the low-noise amplifier can be disconnected first, and the DC bias error between the first and second analog-to-digital converters can be removed. Then, the generation frequency of the second phase-locked loop inside the RF transceiver can be controlled to be... The preset local oscillator signal, after buffering, is output to one input of the quadrature mixer; the frequency generated by the first phase-locked loop inside the RF transceiver is controlled to be... The preset in-band single-tone test signal is buffered and input to the other input of the quadrature mixer. Then, the quadrature mixer inside the RF transceiver performs quadrature mixing between the preset in-band single-tone test signal and the preset local oscillator signal, and outputs a frequency of... In addition to other intermodulation frequency signals, each frequency signal is filtered by an I-path filter and a Q-path filter, resulting in a frequency of [frequency value missing] in its respective path. The frequency signal is sampled and converted to digital by the corresponding first analog-to-digital converter and second analog-to-digital converter, respectively, and then sent to the digital gain compensation module and the cutoff frequency adjustment module for analysis and processing to obtain the in-band power value of the I-channel. and Q-band power value .

[0033] Thus, by injecting a single-tone signal within the low-frequency band, this embodiment makes the I / Q power measurement approximately equal to DC power, thereby eliminating the dependence on orthogonal calibration signals, significantly reducing the complexity of separating test signal generation and computation, and further expanding the applicable scenarios for broadband IQ imbalance calibration.

[0034] Step S20: Calculate the IQ difference between the in-band power value of the I-channel and the in-band power value of the Q-channel, and perform digital gain compensation on the I-channel and / or Q-channel of the RF transceiver based on the IQ difference, so that the gains of the I-channel and Q-channel of the RF transceiver are the same.

[0035] In this step, the IQ difference is used to indicate the low-frequency (DC) gain mismatch between the I and Q paths, specifically:

[0036] The difference in IQ. This represents the in-band power value of channel I. This represents the Q-channel in-band power value. Since the preset in-band single-tone test signal has a low frequency and is located in the flat region of the filter passband, this IQ difference essentially characterizes the gain mismatch between the I and Q channels under low-frequency (or DC) conditions. In the gain compensation control logic, it is... It is used as an error feedback quantity for closed-loop regulation. When When the gain is greater than zero, it indicates that the gain of channel I is higher than that of channel Q. In this case, you can decrease the digital gain of channel I alone, increase the digital gain of channel Q alone, or decrease the gain of channel I and increase the gain of channel Q simultaneously to make the gains of both channels equal. When the value is less than zero, it indicates that the gain of the I-path is lower than that of the Q-path. In this case, you can reduce the digital gain of the Q-path alone, increase the digital gain of the I-path alone, or reduce the gain of the Q-path and increase the gain of the I-path at the same time to make the gains of the two paths the same.

[0037] This step achieves amplitude balance in the I / Q paths, eliminates the degradation of image rejection ratio caused by the discreteness of analog devices, and ensures that the two filter adjustments have the same starting level conditions.

[0038] Step S30: Control the injection of a preset attenuation point test signal and a preset local oscillator signal. After processing by the receiving path, obtain the I-channel cutoff frequency power value output by the I-channel filter and the Q-channel cutoff frequency power value output by the Q-channel filter respectively. The I-channel filter and the Q-channel filter have the same preset cutoff frequency attenuation point, and the frequency of the preset attenuation point test signal is the sum of the third preset frequency and the first preset frequency. The third preset frequency is the frequency at the cutoff frequency attenuation point.

[0039] In this embodiment, a preset attenuation point test signal is controlled to be injected as the second calibration test signal. This preset attenuation point test signal also manifests as a discrete frequency point in the frequency domain, with its frequency being the sum of a third preset frequency and a first preset frequency. The second preset frequency is the frequency at the cutoff frequency attenuation point, corresponding to the theoretical target attenuation value of the filter's amplitude-frequency response. The third preset frequency is set to... This is the cutoff frequency of the I-channel filter / Q-channel filter at the -1dB point (or -3dB point). Therefore, the frequency of the preset attenuation point test signal is... .

[0040] In this embodiment, the generation frequency of the second phase-locked loop inside the RF transceiver is controlled to be... The preset local oscillator signal, after buffering, is output to one input of the quadrature mixer; the frequency generated by the first phase-locked loop inside the RF transceiver is controlled to be... The preset attenuation point test signal is buffered and then input to the other input of the quadrature mixer. The quadrature mixer inside the RF transceiver then performs quadrature mixing between the preset attenuation point test signal and the preset local oscillator signal, and outputs a frequency of... In addition to other intermodulation frequency signals, each frequency signal is filtered by an I-path filter and a Q-path filter, resulting in a frequency of [frequency value missing] in its respective path. The frequency signal is sampled and converted to digital by the corresponding first analog-to-digital converter and second analog-to-digital converter, respectively, and then sent to the cutoff frequency adjustment module for analysis and processing to obtain the in-band power value of the I-channel. and Q-band power value .

[0041] Step S40: Determine the power difference between the I-channel dual-frequency points based on the I-channel in-band power value and the I-channel cutoff frequency point power value, and determine the power difference between the Q-channel dual-frequency points based on the Q-channel in-band power value and the Q-channel cutoff frequency point power value.

[0042] It should be noted that the I / Q path dual-frequency power difference refers to the difference between the in-band reference frequency power and the cutoff frequency power within the same path, essentially characterizing the attenuation depth of the filter from the passband to the cutoff point. In this step, the I path dual-frequency power difference... satisfy: ; Q-channel dual-frequency power difference satisfy: .

[0043] Step S50: Obtain the target attenuation amount corresponding to the cutoff frequency attenuation point, and adjust the cutoff frequency of the I-channel filter according to the first deviation between the power difference of the I-channel dual-frequency points and the target attenuation amount; and adjust the cutoff frequency of the Q-channel filter according to the second deviation between the power difference of the Q-channel dual-frequency points and the target attenuation amount.

[0044] It should be noted that the target attenuation can be obtained by consulting the manual of the I / Q filter, which defines the theoretical attenuation value (e.g., 1dB) that should be achieved at the cutoff frequency (e.g., -1dB).

[0045] In another embodiment, step S50 includes steps S511 to S512: Step S511: If the power difference between the two frequency points of the I-channel is less than the target attenuation, then reduce the cutoff frequency of the I-channel filter.

[0046] Step S512: If the power difference between the two frequency points of the I-channel is greater than the target attenuation, then increase the cutoff frequency of the I-channel filter.

[0047] In this embodiment, if This indicates that the measured attenuation of the I-channel is less than the target attenuation (1dB), meaning too much signal is passing through, resulting in excessive bandwidth. In this case, the bandwidth needs to be reduced, i.e., the cutoff frequency of the I-channel filter needs to be decreased until... and They are equal or approximately equal. If This indicates that the measured attenuation of the I-channel is greater than the target attenuation (1dB), meaning too little signal is passing through and the bandwidth is too small. In this case, the bandwidth needs to be increased, i.e., the cutoff frequency of the I-channel filter needs to be increased until... and Equal or approximately equal.

[0048] In another embodiment, step S50 includes steps S521-S522: Step S521: If the power difference between the two frequency points of the Q-channel is less than the target attenuation, then increase the cutoff frequency of the Q-channel filter.

[0049] Step S522: If the power difference between the two frequency points of the Q-channel is greater than the target attenuation, then reduce the cutoff frequency of the Q-channel filter.

[0050] In this embodiment, if This indicates that the measured attenuation of the Q-channel is less than the target attenuation (1dB), meaning too much signal is passing through and the bandwidth is too large. In this case, the bandwidth needs to be reduced, i.e., the cutoff frequency of the Q-channel filter needs to be reduced until... and They are equal or approximately equal. If This indicates that the measured attenuation of the Q-channel is greater than the target attenuation (1dB), meaning too little signal is passing through and the bandwidth is too small. In this case, the bandwidth needs to be increased, i.e., the cutoff frequency of the Q-channel filter needs to be increased until... and Equal or approximately equal.

[0051] In summary, this embodiment can calibrate frequency-independent IQ imbalance in the digital domain and frequency-dependent IQ imbalance in the analog domain. It can simplify the complex broadband IQ balance calculation in the digital domain and achieve broadband IQ imbalance calibration of RF transceivers at low cost.

[0052] It should be noted that the signal transfer function of an Nth-order low-pass filter satisfies: (1-1) in, For signal transfer function, For complex frequency variables, Here, N represents the pole index, and N is the filter order. For the first An extreme point, is the gain constant.

[0053] Analysis based on formula (1-1) shows that the maximum amplitude and phase changes of the low-pass filter are located near the cutoff frequency. Therefore, amplitude and phase calibration can be equivalent to compensating for the cutoff frequency of the low-pass filter. When the bandwidth of the operational amplifier that makes up the low-pass filter is large enough relative to the filter bandwidth, the degree of image mismatch in the receiving channel can be controlled simply by compensating for the cutoff frequency of the low-pass filter in the I and Q paths.

[0054] like Figure 4 and Figure 5 As shown, the interpolation value between the test power at a specified bandwidth and the power at a low frequency is denoted as... (like or When the specified bandwidth is -1dB, if If the bandwidth is less than 1dB, it indicates that the bandwidth is too large and needs to be reduced. Figure 4 of The frequency response graph shows the following: the horizontal axis represents frequency (freq), the vertical axis represents amplitude (Amplitude), the solid line represents the ideal curve, and the dashed line represents the actual deviation curve. If the bandwidth is greater than 1dB, it indicates that the bandwidth is too small and needs to be increased. Figure 5 of As shown in the frequency response diagram, the horizontal axis represents frequency (freq), and the vertical axis represents amplitude (Amplitude). The solid line represents the ideal curve, and the dashed line represents the actual deviation curve.

[0055] It should be noted that bandwidth adjustment can be achieved by adjusting resistor or capacitor arrays. The adjustment array can be designed as an N-bit weighted array, calibrated using a successive approximation method. For an N-bit weighted capacitor array, calibration is completed in at most N cycles. Ultimately, this results in... It is close to 1dB, and the calibration error is 1LSB (Least Significant Bit).

[0056] In this embodiment, IQ imbalance is decoupled into two categories: "frequency-independent" and "frequency-dependent," and targeted calibration is performed in both the digital and analog domains respectively. This solves the problems of high hardware resource consumption, high latency, and limited application scenarios in the pure digital domain calibration method of the prior art. Specifically, for frequency-independent IQ gain imbalance, this embodiment performs calibration in the digital domain. By injecting a preset in-band single-tone test signal within the common passband of the I / Q filter, the in-band power values ​​of the I and Q paths are obtained, and the IQ difference is calculated. Based on this, digital gain compensation is performed in the digital domain to make the gains of the two paths consistent. For frequency-dependent IQ imbalance (i.e., broadband image rejection deterioration caused by inconsistent I / Q filter bandwidth), this embodiment performs calibration in the analog domain. By injecting a preset attenuation point test signal located at the attenuation point of the filter cutoff frequency, the dual-frequency power difference of the I and Q paths is calculated in combination with the aforementioned in-band power values. Then, the power difference between the I and Q paths is compared with the target attenuation amount, and the cutoff frequencies of the I and Q path filters are adjusted according to the deviation. Thus, by compensating for the cutoff frequency of the low-pass filters in the I and Q paths, the degree of image mismatch in the receiving channel can be directly controlled. Compared with existing technologies, this embodiment controls the injected single-tone test signal as the calibration test signal, eliminating the dependence on orthogonal calibration signals, significantly reducing the complexity of separating test signal generation and computation, and thus expanding the applicable scenarios for broadband IQ imbalance calibration. Furthermore, it can calibrate frequency-independent IQ imbalance in the digital domain and frequency-dependent IQ imbalance in the analog domain, simplifying the complex broadband IQ balance calculations in the digital domain and reducing the hardware resource consumption and latency required for broadband IQ imbalance calibration of RF transceivers.

[0057] It should be noted that existing analog domain calibration schemes suffer from insufficient adjustment accuracy.

[0058] Please see Figure 2 In the second embodiment of this application, the I-channel filter includes a first adjustable impedance array for adjusting the cutoff frequency of the I-channel filter, and the Q-channel filter includes a second adjustable impedance array for adjusting the cutoff frequency of the Q-channel filter. Step S50 includes step S531: Adjusting the equivalent capacitance and / or equivalent resistance of the first adjustable impedance array to change the RC time constant of the first adjustable impedance array; Adjust the equivalent capacitance and / or equivalent resistance of the second adjustable impedance array to change the RC time constant of the second adjustable impedance array.

[0059] In this embodiment, the first adjustable impedance array and the second adjustable impedance array are circuit modules integrated inside the I-channel filter and the Q-channel filter, respectively, capable of changing the RC time constant by adjusting the equivalent capacitance and / or equivalent resistance values. The RC time constant is a key physical parameter characterizing the filter's charging and discharging rate and frequency response, determined by the product of resistance (R) and capacitance (C). In this embodiment, by adjusting the equivalent capacitance and / or equivalent resistance values ​​of the first adjustable impedance array to change the RC time constant of the I-channel filter, and simultaneously by adjusting the equivalent capacitance and / or equivalent resistance values ​​of the second adjustable impedance array to change the RC time constant of the Q-channel filter, the cutoff frequencies of the I-channel and Q-channel filters can be flexibly and precisely controlled independently based on the inverse relationship between the RC time constant and the cutoff frequency. This effectively solves the problem that traditional fixed-parameter filters cannot adapt to multi-band requirements or compensate for the inconsistency in frequency response between the I and Q channels caused by process variations, ensuring the orthogonality of the signal processing link and system performance.

[0060] It is understandable that the higher the precision of the cutoff frequency adjustment array (number of control bits), the higher the image rejection level.

[0061] In one embodiment, the first adjustable impedance array / second adjustable impedance array is an N-bit binary weighted capacitor array, which includes: A basic branch consisting of a fixed resistor and a fixed capacitor connected in parallel; And a switchable capacitor array consisting of N-bit binary weighted capacitor units, where the value of the k-th capacitor is... Each capacitor unit is connected in parallel to both ends of the basic branch through an independent switch; The equivalent capacitance of the adjustable impedance array satisfies:

[0062] in, This represents the capacitance value of the capacitor unit. For a fixed capacitance value, 1 ≤ k ≤ N, 0 ≤ n ≤ And k and n are both integers; Adjusting the equivalent capacitance values ​​of the first adjustable impedance array / second adjustable impedance array includes: A successive approximation algorithm is used to control the on / off state of each independent switch in the N-bit binary weighted capacitor array in order from the most significant bit to the least significant bit.

[0063] It should be noted that for a unity-gain third-order low-pass Butterworth filter, the transfer function in equation (1-1) degenerates to:

[0064] The complex frequency response (transfer function) of the filter represents the complex ratio of the output signal to the input signal in the frequency domain, including amplitude and phase information; , These are the imaginary unit and the normalized angular frequency, respectively.

[0065] Then its magnitude transfer function is (1-2) This is the modulus of the transfer function.

[0066] Its phase transfer function for (1-3) Let be the phase transfer function (phase response) of the filter, representing the phase lag of the output signal relative to the input signal as the frequency of the output signal changes. The changing relationship.

[0067] Expression (1-3) describes the phase of a Butterworth third-order low-pass filter as a function of frequency, with the negative sign indicating that the output lags behind the input.

[0068] From the amplitude-frequency response and phase-frequency response of the third-order low-pass Butterworth filter above, we can see... and All are as they please It is a function that changes and is uniquely determined by it. However, in actual semiconductor circuit manufacturing, the inherent device mismatch must be considered, that is, there is a small mismatch between the I and Q paths. Errors, their effect on amplitude and phase.

[0069] right and Differentiation yields the rate of change when a small mismatch occurs in the I and Q pathways. (1-4) in, For the reason The resulting change in amplitude response (amplitude error) indicates the degree to which the filter gain deviates from the ideal value when frequency mismatch exists. For the reason The resulting change in phase response (phase error) indicates the degree to which the filter phase deviates from the ideal value when frequency mismatch exists.

[0070] This represents a tiny frequency error, typically determined by RC mismatch in the manufacturing process.

[0071] Expanding the above equation near the cutoff frequency yields... (1-5) The relationship between image suppression and phase / amplitude mismatch is as follows: (1-6) The above formula can be used to calculate if there is a frequency mismatch. hour, , The calculated IRR is -37.34 dB. Therefore, improving... The accuracy can improve the image suppression ratio.

[0072] The cutoff frequency of the filter is determined by the RC time constant. For example... Figure 6 As shown, in this embodiment, the N-bit binary weighted capacitor array consists of a basic branch with a fixed resistor and a fixed capacitor connected in parallel, and N capacitor values ​​arranged according to... A circuit structure consisting of regularly distributed switchable capacitor units connected in parallel to the basic branch via independent switches has an equivalent capacitance value that satisfies... (in This represents the capacitance value of the capacitor unit. For a fixed capacitance value, 0 ≤ n ≤ And n is an integer control code). The successive approximation algorithm used during adjustment is a search strategy that compares and determines the switch state sequentially from the most significant bit to the least significant bit. For example, when N=3, the most significant bit (the 3rd bit, 4th bit, etc.) is first compared. When the switch is closed and the others are open, determine whether the current capacitance value has reached or exceeded the target value. If it has not reached the target value, keep that bit closed; if it has exceeded the target value, open that bit. Then, follow this logic to determine the second bit (2). ) and the 1st position ( The on / off state of the control code n can be determined by only 3 comparisons.

[0073] The capacitance adjustment range, derived from the maximum and minimum values ​​of the capacitor array, is expressed by the following formula: (1-7) For capacitor adjustment range, This represents the maximum capacitance value of the capacitor array. This is the minimum capacitance value of the capacitor array.

[0074] Typically, R and C deviations caused by manufacturing process variations are approximately 30% (the mismatch value is 25% in TSMC's 22nm process). In an RC compensation array, the C array needs to cover the overall RC deviation; therefore, the capacitance coverage ratio under this process needs to reach 56.25%. The process-related capacitance coverage ratio parameter is defined as follows: The adjustable range of the capacitor array , The value of is determined by the following formula, where Desired capacitance value: ; (1-8) Substituting into formula (1-7) gives

[0075] When higher orders are ignored, we have (1-9) (1-7) and (1-9) are combined to form: (1-10) Equation (1-10) shows that the number of bits in the capacitor array and the process deviation coefficient determine the... ,Right now The calibration accuracy.

[0076] Based on the above analysis, in another embodiment, the adjustment accuracy of the first adjustable impedance array and / or the second adjustable impedance array is adjusted based on the capacitance coverage ratio parameter and the bit number N. In this embodiment, by coordinating the adjustment of the impedance array accuracy with the capacitance coverage ratio parameter and the bit number N, the decoupling optimization of the tuning range and calibration resolution is achieved, improving the locking accuracy of the filter cutoff frequency and the consistency of the amplitude-frequency characteristics.

[0077] Fit using the above formula At that time, different capacitor array bit positions correspond to The following table shows the size and the corresponding Mag and Pha mismatches when changing by one LSB (Least Significant Bit): Table 1. Formula for calculating mismatch and image suppression values ​​for different capacitor array bits.

[0078] Where Cw is the desired capacitance value, in pF; 1. Capacitor coverage ratio parameter, in %; 2. Bits, in Bits; 3. Cu / Cw, the normalized ratio of unit step capacitance to desired capacitance, reflecting the proportion of LSB step in total capacitance, in %; 4. Cu, the unit step capacitance value, in fF; 5. Mag, the relative amplitude change caused by changing one LSB, in percentage % / mdB; 6. Pha, the phase change caused by changing one LSB, in rad / deg; 7. IMR, the image rejection value, in dB.

[0079] By comparing the formula calculations with the actual circuit simulation, this embodiment... The good fit between the phase and amplitude mismatch values ​​proves the effectiveness of the formula fitting.

[0080] Thus, this embodiment achieves precise digital step adjustment of capacitance value through a binary weighted structure, which can reduce adjustment complexity, improve calibration efficiency and convergence speed, solve the problems of low adjustment accuracy and long search time of traditional linear arrays, and ensure that the cutoff frequency of the I / Q filter can quickly and accurately match the preset index and compensate for process deviations.

[0081] In another embodiment, the first adjustable impedance array / second adjustable impedance array is an N-bit binary weighted resistor array, and the N-bit binary weighted capacitor array includes: A basic branch consisting of a fixed resistor and a fixed capacitor connected in parallel; And a switchable capacitor array consisting of N-bit binary weighted resistor units, where the value of the k-th resistor is... Each resistor unit is connected in parallel to both ends of the basic branch through an independent switch; The equivalent resistance of the adjustable impedance array satisfies:

[0082] in, This represents the resistance value of the resistor unit. The resistance value is a fixed value, 1 ≤ k ≤ N, 0 ≤ n ≤ And k and n are both integers; Adjusting the equivalent resistance values ​​of the first adjustable impedance array / second adjustable impedance array includes: The successive approximation algorithm is used to control the on / off state of each independent switch in the N-bit binary weighted resistor array in order from the most significant bit to the least significant bit.

[0083] In this embodiment, an N-bit binary weighted resistor array refers to a basic branch consisting of a fixed resistor and a fixed capacitor connected in parallel, and N resistor values ​​arranged according to... A circuit structure consisting of regularly distributed switchable resistor units connected in parallel to both ends of a basic branch via independent switches has an equivalent resistance value that satisfies... A successive approximation algorithm can be used to adjust the on / off state of each independent switch sequentially from the most significant bit to the least significant bit. For example, when N=3, the most significant bit (4 bits) is determined first. ) Whether the switch can be connected to make the equivalent resistance closer to the target value, and then determine the second bit (2) in sequence. ) and the 1st position ( Only three comparisons are needed to determine the final control code n. Thus, precise digital step adjustment of the capacitance value is achieved through a binary weighted structure, reducing adjustment complexity, improving calibration efficiency and convergence speed, and solving the problems of low adjustment accuracy and long search time in traditional linear arrays. This ensures that the cutoff frequency of the I / Q path filter can quickly and accurately match the preset specifications and compensate for process deviations.

[0084] It's important to note that reducing bandwidth requires lowering the cutoff frequency, which necessitates increasing the RC time constant. Specifically, in adjusting an adjustable impedance array, a capacitor array requires increasing the equivalent capacitance, while a resistor array requires increasing the equivalent resistance. Conversely, increasing bandwidth requires raising the cutoff frequency, decreasing the RC constant, and correspondingly reducing the capacitance or resistance values ​​in the array. Thus, by digitally adjusting the value of n in the N-bit binary weighted array, the filter bandwidth can be precisely controlled in steps, enabling closed-loop calibration of the I / Q channel frequency response characteristics.

[0085] This application also proposes a broadband orthogonal calibration device for a radio frequency transceiver.

[0086] Please see Figure 3 In one embodiment of this application, the broadband orthogonal calibration device for the radio frequency transceiver includes: The injection control and power acquisition module is used to control the injection of a preset in-band single-tone test signal and a preset local oscillator signal. After processing by the receiving path of the RF transceiver, the module acquires the I-channel in-band power value output by the I-channel filter and the Q-channel in-band power value output by the Q-channel filter of the RF transceiver, respectively. The frequency of the preset local oscillator signal is a first preset frequency, and the frequency of the preset in-band single-tone test signal is the sum of the second preset frequency and the first preset frequency. The second preset frequency is within the common passband of the I-channel filter and the Q-channel filter. The digital gain compensation module is used to calculate the IQ difference between the in-band power value of the I-channel and the in-band power value of the Q-channel, and to perform digital gain compensation on the I-channel and / or Q-channel of the RF transceiver based on the IQ difference, so that the gain of the I-channel and Q-channel of the RF transceiver is the same. The injection control and power acquisition module is also used to control the injection of a preset attenuation point test signal and a preset local oscillator signal. After processing by the receiving path, the power values ​​at the I-channel cutoff frequency point output by the I-channel filter and the power values ​​at the Q-channel cutoff frequency point output by the Q-channel filter are obtained respectively. The I-channel filter and the Q-channel filter have the same preset cutoff frequency attenuation point, and the frequency of the preset attenuation point test signal is the sum of the third preset frequency and the first preset frequency. The third preset frequency is the frequency at the cutoff frequency attenuation point. The cutoff frequency adjustment module is used to determine the power difference of the I-channel dual-frequency point based on the I-channel in-band power value and the I-channel cutoff frequency point power value, and to determine the power difference of the Q-channel dual-frequency point based on the Q-channel in-band power value and the Q-channel cutoff frequency point power value. In addition, it is used to obtain the target attenuation corresponding to the cutoff frequency attenuation point, and adjust the cutoff frequency of the I-channel filter according to the first deviation between the power difference of the I-channel dual-frequency points and the target attenuation; and adjust the cutoff frequency of the Q-channel filter according to the second deviation between the power difference of the Q-channel dual-frequency points and the target attenuation.

[0087] In this embodiment, the injection control and power acquisition module may include a microcontroller unit (MCU) to control the injection timing of the test signal, switch paths, and read and transmit the power values ​​of the I / Q channels. The digital gain compensation module may include a digital signal processor (DSP) to calculate the in-band power difference between the I / Q channels and generate the corresponding digital gain compensation coefficients. The cutoff frequency adjustment module may include a received signal strength indicator (RSSI) and a calibration algorithm unit. The RSSI is used to acquire the power values ​​of the I / Q channels at the cutoff frequency attenuation point, and the calibration algorithm unit calculates and outputs the filter cutoff frequency adjustment word (e.g., based on the deviation between the power difference at the two frequency points and the target attenuation) according to the RSSI. Figure 3 (I_Wc<9:0> / Q_Wc<9:0>), thereby completing the closed-loop calibration of the cutoff frequency.

[0088] It should be noted that existing technologies typically require injecting calibration test signals with orthogonal characteristics for testing, in order to facilitate the separation of broadband IQ imbalance parameters through calculation, thus limiting their application scenarios. In this embodiment, a preset in-band single-tone test signal is controlled as the first calibration test signal. This preset in-band single-tone test signal is represented as a discrete frequency point in the frequency domain, and its frequency is the sum of the second preset frequency and the first preset frequency, with the second preset frequency falling within the common passband of the I-path filter and the Q-path filter. The first preset frequency is set to... The second preset frequency is The frequency of the preset in-band single-tone test signal is then... You can set The frequency value is relatively low, so that the obtained I-band power value and Q-band power value are approximately DC power.

[0089] In one embodiment, the quadrature mixer inside the RF transceiver is connected to the output of the low-noise amplifier via a switching circuit. The first analog-to-digital converter and the second analog-to-digital converter inside the RF transceiver are connected to the I-channel filter and the Q-channel filter, respectively, to convert the analog signals output from the I-channel filter and the Q-channel filter into digital signals for subsequent analysis and calculation. The injection control and power acquisition module is also used for: Control the disconnection switch circuit.

[0090] The first analog-to-digital converter and the second analog-to-digital converter are calibrated separately to remove their respective DC offset errors.

[0091] In this embodiment, the injection control and power acquisition module can first disconnect the low-noise amplifier and remove the DC bias error between the first and second analog-to-digital converters before controlling the second phase-locked loop inside the RF transceiver to generate a frequency of... The preset local oscillator signal, after buffering, is output to one input of the quadrature mixer; simultaneously, the frequency generated by the first phase-locked loop inside the RF transceiver is controlled to be... The preset in-band single-tone test signal is buffered and input to the other input of the quadrature mixer. Then, the quadrature mixer inside the RF transceiver performs quadrature mixing between the preset in-band single-tone test signal and the preset local oscillator signal, and outputs a frequency of... In addition to other intermodulation frequency signals, each frequency signal is filtered by an I-path filter and a Q-path filter, resulting in a frequency of [frequency value missing] in its respective path. The frequency signal is sampled and converted to digital by the corresponding first analog-to-digital converter and second analog-to-digital converter, respectively, and then sent to the digital gain compensation module and the cutoff frequency adjustment module (RSSI) for analysis and processing to obtain the in-band power value of the I-channel. and Q-band power value In this way, by injecting a single-tone signal within the low-frequency band, the power measurement of the I / Q channels can be approximated as DC power, thereby eliminating the dependence on orthogonal calibration signals, significantly reducing the complexity of separating test signal generation and calculation, and thus expanding the applicable scenarios for broadband IQ imbalance calibration.

[0092] In this embodiment, the IQ difference is used to indicate the low-frequency (DC) gain mismatch between the I and Q channels, specifically:

[0093] The difference in IQ. This represents the in-band power value of channel I. This represents the Q-channel in-band power value. Since the preset in-band single-tone test signal has a low frequency and is located in the flat region of the filter passband, this IQ difference essentially characterizes the gain mismatch between the I and Q channels under low-frequency (or DC) conditions. In the gain compensation control logic, it is... It is used as an error feedback quantity for closed-loop regulation. When When the gain is greater than zero, it indicates that the gain of channel I is higher than that of channel Q. In this case, you can decrease the digital gain of channel I alone, increase the digital gain of channel Q alone, or decrease the gain of channel I and increase the gain of channel Q simultaneously to make the gains of both channels equal. When the gain is less than zero, it indicates that the gain of the I-path is lower than that of the Q-path. In this case, the digital gain of the Q-path can be decreased individually, the digital gain of the I-path can be increased individually, or the gain of the Q-path can be decreased and the gain of the I-path can be increased simultaneously to make the gains of the two paths the same. In this way, the amplitude balance of the I / Q paths is achieved, eliminating the image rejection ratio degradation caused by the discreteness of analog devices, and ensuring that the two filter adjustments have the same starting level conditions.

[0094] In this embodiment, the preset attenuation point test signal is used as the second calibration test signal. In the frequency domain, it is also represented as a discrete frequency point, the frequency of which is the sum of the third preset frequency and the first preset frequency. The second preset frequency is the frequency at the cutoff frequency attenuation point, corresponding to the theoretical target attenuation value of the filter's amplitude-frequency response. The third preset frequency is set to... This is the cutoff frequency of the I-channel filter / Q-channel filter at the -1dB point (or -3dB point). Therefore, the frequency of the preset attenuation point test signal is... .

[0095] During operation, the frequency generated by the second phase-locked loop inside the RF transceiver is controlled to be... The preset local oscillator signal, after buffering, is output to one input of the quadrature mixer; the frequency generated by the first phase-locked loop inside the RF transceiver is controlled to be... The preset attenuation point test signal is buffered and then input to the other input of the quadrature mixer. The quadrature mixer inside the RF transceiver then performs quadrature mixing between the preset attenuation point test signal and the preset local oscillator signal, and outputs a frequency of... In addition to other intermodulation frequency signals, each frequency signal is filtered by an I-path filter and a Q-path filter, resulting in a frequency of [frequency value missing] in its respective path. The frequency signal is sampled and converted to digital by the corresponding first and second analog-to-digital converters, and then sent to the cutoff frequency adjustment module (RSSI) for analysis and processing to obtain the in-band power value of the I-channel. and Q-band power value .

[0096] In this embodiment, the I / Q path dual-frequency power difference refers to the difference between the in-band reference frequency power and the cutoff frequency power under the same path, essentially characterizing the attenuation depth of the filter from the passband to the cutoff point. In this step, the I path dual-frequency power difference... satisfy: ; Q-channel dual-frequency power difference satisfy: .

[0097] It should be noted that the target attenuation can be obtained by consulting the I / Q filter manual, which defines the theoretical attenuation value (e.g., 1dB) that should be achieved at the cutoff frequency (e.g., -1dB).

[0098] In one embodiment, when the cutoff frequency adjustment module adjusts the cutoff frequency of the I-channel filter based on the first deviation between the power difference of the I-channel dual-frequency points and the target attenuation, it is specifically used for: If the power difference between the two frequency points of the I-channel is less than the target attenuation, then reduce the cutoff frequency of the I-channel filter. If the power difference between the two frequency points of the I-channel is greater than the target attenuation, then increase the cutoff frequency of the I-channel filter. When adjusting the cutoff frequency of the Q-channel filter based on the second deviation between the power difference at the dual frequency points of the Q-channel and the target attenuation, the cutoff frequency adjustment module is specifically used for: If the power difference between the two frequency points of the Q-channel is less than the target attenuation, then increase the cutoff frequency of the Q-channel filter. If the power difference between the two frequency points of the Q-channel is greater than the target attenuation, then reduce the cutoff frequency of the Q-channel filter.

[0099] In this embodiment, if This indicates that the measured attenuation of the I-channel is less than the target attenuation (1dB), meaning too much signal is passing through, resulting in excessive bandwidth. In this case, the bandwidth needs to be reduced, i.e., the cutoff frequency of the I-channel filter needs to be decreased until... and They are equal or approximately equal. If This indicates that the measured attenuation of the I-channel is greater than the target attenuation (1dB), meaning too little signal is passing through and the bandwidth is too small. In this case, the bandwidth needs to be increased, i.e., the cutoff frequency of the I-channel filter needs to be increased until... and Equal or approximately equal.

[0100] like This indicates that the measured attenuation of the Q-channel is less than the target attenuation (1dB), meaning too much signal is passing through and the bandwidth is too large. In this case, the bandwidth needs to be reduced, i.e., the cutoff frequency of the Q-channel filter needs to be reduced until... and They are equal or approximately equal. If This indicates that the measured attenuation of the Q-channel is greater than the target attenuation (1dB), meaning too little signal is passing through and the bandwidth is too small. In this case, the bandwidth needs to be increased, i.e., the cutoff frequency of the Q-channel filter needs to be increased until... and Equal or approximately equal.

[0101] In summary, this embodiment can calibrate frequency-independent IQ imbalance in the digital domain and frequency-dependent IQ imbalance in the analog domain. It can simplify the complex broadband IQ balance calculation in the digital domain and achieve broadband IQ imbalance calibration of RF transceivers at low cost.

[0102] In this embodiment, bandwidth adjustment can be achieved by adjusting a resistor or capacitor array. The adjustment array can be designed as an N-bit weighted array, calibrated using a successive approximation method. For an N-bit weighted capacitor array, calibration is completed in at most N cycles. Ultimately, this results in... Very close to 1dB, with a calibration error of 1LSB (Least Significant Bit).

[0103] In this embodiment, the IQ imbalance is decoupled into two categories: "frequency-independent" and "frequency-dependent," and targeted calibration is performed in both the digital and analog domains respectively. This solves the problems of high hardware resource consumption, high latency, and limited application scenarios in the pure digital domain calibration method of the prior art. Specifically, for frequency-independent IQ gain imbalance, this embodiment performs calibration in the digital domain. A preset in-band single-tone test signal within the common passband of the I / Q path filters is injected through the injection control and power acquisition module, and the in-band power values ​​of the I and Q paths are obtained. The in-band power values ​​of the I and Q paths are obtained through the digital gain compensation module, and the IQ difference is calculated. Based on this, digital gain compensation is performed in the digital domain to make the gains of the two paths consistent. To address frequency-related IQ imbalance (i.e., deteriorated broadband image rejection due to inconsistent I / Q filter bandwidth), this embodiment performs calibration in the analog domain. A preset attenuation point test signal located at the filter cutoff frequency attenuation point is injected through an injection control and power acquisition module, and the power values ​​at the cutoff frequencies of the I and Q paths are acquired. The cutoff frequency adjustment module calculates the dual-frequency power difference between the I and Q paths based on the aforementioned in-band power values. Then, the power difference between the I and Q paths is compared with the target attenuation, and the cutoff frequencies of the I and Q path filters are adjusted according to the deviation. Thus, this embodiment can directly control the degree of image mismatch in the receiving channel by compensating for the cutoff frequencies of the low-pass filters in the I and Q paths. Compared with existing technologies, this embodiment uses an injection control and power acquisition module to control the injection of a single-tone test signal as the calibration test signal, eliminating the dependence on orthogonal calibration signals, significantly reducing the complexity of separating test signal generation and computation, and thus expanding the applicable scenarios for broadband IQ imbalance calibration. Furthermore, the digital gain compensation module can calibrate frequency-independent IQ imbalance in the digital domain, and the cutoff frequency adjustment module can calibrate frequency-dependent IQ imbalance in the analog domain. This simplifies the complex broadband IQ balance calculation in the digital domain and reduces the hardware resource consumption and latency required to achieve broadband IQ imbalance calibration of RF transceivers.

[0104] In another embodiment of this application, the I-channel filter includes a first adjustable impedance array for adjusting the cutoff frequency of the I-channel filter, and the Q-channel filter includes a second adjustable impedance array for adjusting the cutoff frequency of the Q-channel filter. When adjusting the cutoff frequency of the I-channel filter, the cutoff frequency adjustment module is specifically used to: adjust the equivalent capacitance and / or equivalent resistance of the first adjustable impedance array to change the RC time constant of the first adjustable impedance array. When adjusting the cutoff frequency of the Q-channel filter, the cutoff frequency adjustment module is specifically used to: adjust the equivalent capacitance and / or equivalent resistance of the second adjustable impedance array to change the RC time constant of the second adjustable impedance array.

[0105] In one embodiment, the first adjustable impedance array / second adjustable impedance array is an N-bit binary weighted capacitor array, which includes: A fixed resistor R cons With a fixed capacitor C cons Basic branches formed by parallel connections; And a switchable capacitor array consisting of N-bit binary weighted capacitor units, where the value of the k-th capacitor is... Each capacitor unit is connected via an independent switch (including SW). [0] ~SW [N] They are connected in parallel to both ends of the basic branch; The equivalent capacitance of the adjustable impedance array satisfies:

[0106] in, This represents the capacitance value of the capacitor unit. For a fixed capacitance value, 1 ≤ k ≤ N, 0 ≤ n ≤ And k and n are both integers; Adjusting the equivalent capacitance values ​​of the first adjustable impedance array / second adjustable impedance array includes: A successive approximation algorithm is used to control the on / off state of each independent switch in the N-bit binary weighted capacitor array in order from the most significant bit to the least significant bit.

[0107] In this embodiment, the first and second adjustable impedance arrays are integrated inside the I-channel and Q-channel filters, respectively. The cutoff frequency adjustment module is a circuit module that can change the RC time constant by adjusting the equivalent capacitance and / or equivalent resistance values. The RC time constant is a key physical parameter characterizing the filter's charging and discharging rate and frequency response, determined by the product of resistance (R) and capacitance (C). In this embodiment, by adjusting the equivalent capacitance and / or equivalent resistance values ​​of the first adjustable impedance array to change the RC time constant of the I-channel filter, and simultaneously by adjusting the equivalent capacitance and / or equivalent resistance values ​​of the second adjustable impedance array to change the RC time constant of the Q-channel filter, the cutoff frequencies of the I-channel and Q-channel filters can be flexibly and accurately controlled independently based on the inverse relationship between the RC time constant and the cutoff frequency. This effectively solves the problem that traditional fixed-parameter filters cannot adapt to multi-band requirements or compensate for the inconsistency in frequency response between the I and Q channels due to process variations, ensuring the orthogonality of the signal processing link and system performance.

[0108] like Figure 6 As shown, in this embodiment, the N-bit binary weighted capacitor array consists of a basic branch with a fixed resistor and a fixed capacitor connected in parallel, and N capacitor values ​​arranged according to... A circuit structure consisting of regularly distributed switchable capacitor units connected in parallel to the basic branch via independent switches has an equivalent capacitance value that satisfies... (in This represents the capacitance value of the capacitor unit. For a fixed capacitance value, 0 ≤ n ≤ And n is an integer control code). The successive approximation algorithm used during adjustment is a search strategy that compares and determines the switch state sequentially from the most significant bit to the least significant bit. For example, when N=3, the most significant bit (the 3rd bit, 4th bit, etc.) is first compared. When the switch is closed and the others are open, determine whether the current capacitance value has reached or exceeded the target value. If it has not reached the target value, keep that bit closed; if it has exceeded the target value, open that bit. Then, follow this logic to determine the second bit (2). ) and the 1st position ( The on / off state of the control code n can be determined by only 3 comparisons.

[0109] The capacitance adjustment range, derived from the maximum and minimum values ​​of the capacitor array, is expressed by the following formula: (1-7) For capacitor adjustment range, This represents the maximum capacitance value of the capacitor array. This is the minimum capacitance value of the capacitor array.

[0110] Typically, R and C deviations caused by manufacturing process variations are approximately 30% (the mismatch value is 25% in TSMC's 22nm process). In an RC compensation array, the C array needs to cover the overall RC deviation; therefore, the capacitance coverage ratio under this process needs to reach 56.25%. The process-related capacitance coverage ratio parameter is defined as follows: The adjustable range of the capacitor array , The value of is determined by the following formula, where Desired capacitance value: ; (1-8) Substituting into formula (1-7) gives

[0111] When higher orders are ignored, we have (1-9) (1-7) and (1-9) are combined to form: (1-10) Equation (1-10) shows that the number of bits in the capacitor array and the process deviation coefficient determine the... ,Right now The calibration accuracy.

[0112] Based on the above analysis, in another embodiment, the adjustment accuracy of the first adjustable impedance array and / or the second adjustable impedance array is adjusted based on the capacitance coverage ratio parameter and the bit number N. In this embodiment, by coordinating the adjustment of the impedance array accuracy with the capacitance coverage ratio parameter and the bit number N, the decoupling optimization of the tuning range and calibration resolution is achieved, improving the locking accuracy of the filter cutoff frequency and the consistency of the amplitude-frequency characteristics.

[0113] This application also provides an RF transceiver, please refer to... Figure 3 The radio frequency transceiver includes a first phase-locked loop (PLL), a second PLL, a quadrature mixer, an I-channel filter, a Q-channel filter, a first analog-to-digital converter (ADC), and a second ADC, as well as a broadband quadrature calibration device including the radio frequency transceiver described above. The quadrature mixer is connected to the output terminals of the first PLL, the second PLL, the input terminals of the I-channel filter, and the Q-channel filter, respectively. The output terminal of the I-channel filter is connected to the input terminal of the first ADC, and the output terminal of the Q-channel filter is connected to the input terminal of the second ADC. The output terminals of the first ADC, the second ADC, the controlled terminals of the first PLL, and the controlled terminals of the second PLL are respectively connected to an injection control and power acquisition module. The controlled terminals of the I-channel filter and the controlled terminals of the Q-channel filter are connected to a cutoff frequency adjustment module.

[0114] Compared with the prior art, the beneficial effects of the radio frequency transceiver provided in this application are the same as the beneficial effects of the broadband orthogonal calibration device of the radio frequency transceiver provided in the above embodiments, and other technical features in the radio frequency transceiver are the same as the features disclosed in the methods of the above embodiments, which will not be repeated here.

[0115] The above description is merely an exemplary embodiment of this application and does not limit the patent scope of this application. Any equivalent structural transformations made based on the technical concept of this application and the contents of the specification and drawings of this application, or direct / indirect applications in other related technical fields, are included within the patent protection scope of this application.

Claims

1. A broadband orthogonal calibration method for an RF transceiver, characterized in that, include: A preset in-band single-tone test signal and a preset local oscillator signal are injected and processed by the receiving path of the RF transceiver to obtain the I-channel in-band power value output by the I-channel filter and the Q-channel in-band power value output by the Q-channel filter of the RF transceiver, respectively. The frequency of the preset local oscillator signal is a first preset frequency, and the frequency of the preset in-band single-tone test signal is the sum of a second preset frequency and the first preset frequency. The second preset frequency is within the common passband of the I-channel filter and the Q-channel filter. Calculate the IQ difference between the in-band power value of the I-channel and the in-band power value of the Q-channel, and perform digital gain compensation on the I-channel and / or Q-channel of the RF transceiver based on the IQ difference, so that the gains of the I-channel and Q-channel of the RF transceiver are the same. The preset attenuation point test signal and the preset local oscillator signal are controlled and injected. After being processed by the receiving path, the power values ​​at the I-channel cutoff frequency point output by the I-channel filter and the power values ​​at the Q-channel cutoff frequency point output by the Q-channel filter are obtained respectively. The I-channel filter and the Q-channel filter have the same preset cutoff frequency attenuation point, and the frequency of the preset attenuation point test signal is the sum of the third preset frequency and the first preset frequency. The third preset frequency is the frequency at the cutoff frequency attenuation point. The power difference between the I-channel dual-frequency points is determined based on the I-channel in-band power value and the I-channel cutoff frequency point power value, and the power difference between the Q-channel dual-frequency points is determined based on the Q-channel in-band power value and the Q-channel cutoff frequency point power value. Obtain the target attenuation amount corresponding to the cutoff frequency attenuation point, and adjust the cutoff frequency of the I-channel filter according to the first deviation between the power difference of the I-channel dual-frequency points and the target attenuation amount; Furthermore, the cutoff frequency of the Q-channel filter is adjusted based on the second deviation between the power difference of the Q-channel dual-frequency points and the target attenuation.

2. The broadband orthogonal calibration method for an RF transceiver as described in claim 1, characterized in that, The step of adjusting the cutoff frequency of the I-channel filter based on the first deviation between the power difference of the I-channel dual-frequency points and the target attenuation includes: If the power difference between the two frequency points of the I-channel is less than the target attenuation, then the cutoff frequency of the I-channel filter is reduced. If the power difference between the two frequency points of the I-channel is greater than the target attenuation, then the cutoff frequency of the I-channel filter is increased.

3. The broadband orthogonal calibration method for an RF transceiver as described in claim 1, characterized in that, The step of adjusting the cutoff frequency of the Q-channel filter based on the second deviation between the power difference at the Q-channel dual-frequency points and the target attenuation includes: If the power difference between the two frequency points of the Q-channel is less than the target attenuation, then the cutoff frequency of the Q-channel filter is increased. If the power difference between the two frequency points of the Q-channel is greater than the target attenuation, then the cutoff frequency of the Q-channel filter is reduced.

4. The broadband orthogonal calibration method for an RF transceiver as described in claim 1, characterized in that, The I-channel filter includes a first adjustable impedance array for adjusting the cutoff frequency of the I-channel filter, and the Q-channel filter includes a second adjustable impedance array for adjusting the cutoff frequency of the Q-channel filter. Adjusting the cutoff frequency of the I-channel filter includes: Adjust the equivalent capacitance and / or equivalent resistance of the first adjustable impedance array to change the RC time constant of the first adjustable impedance array; Adjusting the cutoff frequency of the Q-channel filter includes: Adjust the equivalent capacitance and / or equivalent resistance of the second adjustable impedance array to change the RC time constant of the second adjustable impedance array.

5. The broadband orthogonal calibration method for an RF transceiver as described in claim 4, characterized in that, The first adjustable impedance array / the second adjustable impedance array is an N-bit binary weighted capacitor array, wherein the N-bit binary weighted capacitor array includes: A basic branch consisting of a fixed resistor and a fixed capacitor connected in parallel; And a switchable capacitor array consisting of N-bit binary weighted capacitor units, where the value of the k-th capacitor is... Each capacitor unit is connected in parallel to both ends of the basic branch through an independent switch; The equivalent capacitance value of the adjustable impedance array satisfies: in, This represents the capacitance value of the capacitor unit. For a fixed capacitance value, 1 ≤ k ≤ N, 0 ≤ n ≤ And k and n are both integers; Adjusting the equivalent capacitance value of the first adjustable impedance array / the second adjustable impedance array includes: A successive approximation algorithm is used to control the on / off state of each independent switch in the N-bit binary weighted capacitor array in order from the most significant bit to the least significant bit.

6. A broadband orthogonal calibration device for a radio frequency transceiver, characterized in that, include: The injection control and power acquisition module is used to control the injection of a preset in-band single-tone test signal and a preset local oscillator signal. After processing by the receiving path of the RF transceiver, the module acquires the I-channel in-band power value output by the I-channel filter and the Q-channel in-band power value output by the Q-channel filter of the RF transceiver, respectively. The frequency of the preset local oscillator signal is a first preset frequency, and the frequency of the preset in-band single-tone test signal is the sum of a second preset frequency and the first preset frequency. The second preset frequency is within the common passband of the I-channel filter and the Q-channel filter. A digital gain compensation module is used to calculate the IQ difference between the in-band power value of the I-channel and the in-band power value of the Q-channel, and to perform digital gain compensation on the I-channel and / or Q-channel of the RF transceiver based on the IQ difference, so that the gains of the I-channel and Q-channel of the RF transceiver are the same. The injection control and power acquisition module is also used to control the injection of a preset attenuation point test signal and the preset local oscillator signal. After processing by the receiving path, the power values ​​at the I-channel cutoff frequency point output by the I-channel filter and the power values ​​at the Q-channel cutoff frequency point output by the Q-channel filter are obtained respectively. The I-channel filter and the Q-channel filter have the same preset cutoff frequency attenuation point, and the frequency of the preset attenuation point test signal is the sum of a third preset frequency and a first preset frequency. The third preset frequency is the frequency at the cutoff frequency attenuation point. The cutoff frequency adjustment module is used to determine the power difference of the I-channel dual-frequency point based on the I-channel in-band power value and the I-channel cutoff frequency point power value, and to determine the power difference of the Q-channel dual-frequency point based on the Q-channel in-band power value and the Q-channel cutoff frequency point power value. In addition, it is used to obtain the target attenuation amount corresponding to the cutoff frequency attenuation point, and adjust the cutoff frequency of the I-channel filter according to the first deviation between the power difference of the I-channel dual frequency points and the target attenuation amount; Furthermore, the cutoff frequency of the Q-channel filter is adjusted based on the second deviation between the power difference of the Q-channel dual-frequency points and the target attenuation.

7. The broadband orthogonal calibration device for an RF transceiver as described in claim 6, characterized in that, When the cutoff frequency adjustment module adjusts the cutoff frequency of the I-channel filter based on the first deviation between the power difference of the I-channel dual-frequency points and the target attenuation, it is specifically used for: If the power difference between the two frequency points of the I-channel is less than the target attenuation, then the cutoff frequency of the I-channel filter is reduced. If the power difference between the two frequency points of the I-channel is greater than the target attenuation, then the cutoff frequency of the I-channel filter is increased. When the cutoff frequency adjustment module adjusts the cutoff frequency of the Q-channel filter based on the second deviation between the power difference of the Q-channel dual-frequency points and the target attenuation, it is specifically used for: If the power difference between the two frequency points of the Q-channel is less than the target attenuation, then the cutoff frequency of the Q-channel filter is increased. If the power difference between the two frequency points of the Q-channel is greater than the target attenuation, then the cutoff frequency of the Q-channel filter is reduced.

8. The broadband orthogonal calibration device for a radio frequency transceiver as described in claim 6, characterized in that, The I-channel filter includes a first adjustable impedance array for adjusting the cutoff frequency of the I-channel filter, and the Q-channel filter includes a second adjustable impedance array for adjusting the cutoff frequency of the Q-channel filter. When adjusting the cutoff frequency of the I-channel filter, the cutoff frequency adjustment module is specifically used to: adjust the equivalent capacitance value and / or equivalent resistance value of the first adjustable impedance array to change the RC time constant of the first adjustable impedance array. When adjusting the cutoff frequency of the Q-channel filter, the cutoff frequency adjustment module is specifically used to: adjust the equivalent capacitance and / or equivalent resistance of the second adjustable impedance array to change the RC time constant of the second adjustable impedance array.

9. The broadband orthogonal calibration device for a radio frequency transceiver as described in claim 8, characterized in that, The first adjustable impedance array / the second adjustable impedance array is an N-bit binary weighted capacitor array, wherein the N-bit binary weighted capacitor array includes: A basic branch consisting of a fixed resistor and a fixed capacitor connected in parallel; And a switchable capacitor array consisting of N-bit binary weighted capacitor units, where the value of the k-th capacitor is... Each capacitor unit is connected in parallel to both ends of the basic branch through an independent switch; The equivalent capacitance value of the adjustable impedance array satisfies: in, This represents the capacitance value of the capacitor unit. For a fixed capacitance value, 1 ≤ k ≤ N, 0 ≤ n ≤ And k and n are both integers; Adjusting the equivalent capacitance value of the first adjustable impedance array / the second adjustable impedance array includes: A successive approximation algorithm is used to control the on / off state of each independent switch in the N-bit binary weighted capacitor array in order from the most significant bit to the least significant bit.

10. A radio frequency transceiver, characterized in that, The transceiver includes a first phase-locked loop (PLL), a second PLL, a quadrature mixer, an I-channel filter, a Q-channel filter, a first analog-to-digital converter (ADC), and a second ADC, as well as a broadband quadrature calibration device including the RF transceiver as described in any one of claims 6 to 9. The quadrature mixer is connected to the output terminals of the first PLL, the second PLL, the I-channel filter, and the Q-channel filter, respectively. The output terminal of the I-channel filter is connected to the input terminal of the first ADC, and the output terminal of the Q-channel filter is connected to the input terminal of the second ADC. The output terminal of the first ADC, the controlled terminal of the first PLL, and the controlled terminal of the second PLL are connected to the injection control and power acquisition module, respectively. The controlled terminals of the I-channel filter and the Q-channel filter are connected to the cutoff frequency adjustment module.

Images