Calibration apparatus and calibration method for signal generators

Abstract

The present invention provides a calibration device that can reduce the level of image signals generated within the quadrature modulation output bandwidth. [Solution] A calibration device 1 for a signal generator 10 equipped with a D / A converter and a quadrature modulator, comprising: a waveform data storage unit 20 that outputs I-waveform data and Q-waveform data of the IF frequency to the I channel and Q channel, respectively; a signal analysis unit 40 that captures the modulated signal output from the quadrature modulator and quadrature demodulates it into I-demodulated waveform data and Q-demodulated waveform data of the IF frequency; an error calculation unit 50 that calculates the difference between the I-phase and Q-phase of the amplitude IF frequency characteristics and the difference between the I-phase and Q-phase of the phase IF frequency characteristics from a plurality of I-demodulated waveform data and Q-demodulated waveform data of the IF frequency; and an error correction unit 30 having correction filters 33a and 33b that correct the difference between the I-phase and Q-phase of the amplitude IF frequency characteristics and the difference between the I-phase and Q-phase of the phase IF frequency characteristics.

Description

[Technical Field]

[0001] The present invention relates to a calibration device and calibration method for a signal generator. [Background technology]

[0002] In order to develop and test wireless receivers and devices, a signal generator is used to input a modulated signal to the object under test, and the RF frequency characteristics and IF frequency characteristics of the object under test are measured, as well as to confirm whether the object under test has achieved the desired characteristics.

[0003] Furthermore, in recent years, the bandwidth of modulated signals has been increasing in order to enable high-speed transmission of data and other information.

[0004] In signal generators using quadrature modulators, an image signal is generated due to the gain difference (amplitude difference) between the I-phase and Q-phase, the delay difference (phase difference) between the I-phase and Q-phase, and the quadrature error of the quadrature modulator, all of which occur in the path from the D / A converter to the quadrature modulator. When an image signal is generated, the measured value becomes degraded compared to the actual value in measurements such as EVM (Error Vector Magnitude) of the object being measured. Therefore, it is desirable to minimize the amount of image signal generated as much as possible. [Prior art documents] [Patent Documents]

[0005] [Patent Document 1] Patent No. 6209239 [Overview of the project] [Problems that the invention aims to solve]

[0006] However, in the conventional technology described in Patent Document 1, the gain difference and delay difference between the I-phase and Q-phase were measured at a certain IF frequency, and the image signal was reduced by generating a signal from the D / A converter to pull back that difference. In other words, the IF frequency characteristics with respect to the gain difference and delay difference between the I-phase and Q-phase were considered flat, and measured and corrected at a single IF frequency.

[0007] Furthermore, since orthogonality is not related to the IF frequency, it was measured and corrected with an IF frequency of 0 Hz.

[0008] However, in recent years, with the widening of the modulation signal bandwidth, the IF frequency vs. amplitude characteristics and IF frequency vs. phase characteristics cannot be considered flat. Therefore, correction using a correction value measured at a single IF frequency may not satisfy the target value of the IMRR (Image Rejection Ratio), such as 50 dB, for the level of the image signal generated within the modulation output bandwidth, potentially reducing the reliability of the measurement results of the object being measured.

[0009] Here, IMRR is defined as (power of the desired CW signal [dBm]) - (power of the generated image signal [dBm]), and the smaller the generated image signal, the larger the IMRR value.

[0010] The present invention was made to solve the above-mentioned problems, and aims to provide a calibration device and calibration method for a signal generator that can reduce the level of the image signal generated within the quadrature modulation output band. [Means for solving the problem]

[0011] The calibration apparatus according to the present invention is a calibration apparatus (1) for a signal generator (10) that, in order to achieve the above objective, comprises a D / A converter (11a, 11b) that performs D / A conversion of I waveform data and Q waveform data, respectively, and a quadrature modulator (12) that quadrature modulates the signal after D / A conversion and outputs it as a modulated signal, and further comprises a waveform data storage unit (20) that outputs the I waveform data and Q waveform data of the IF frequency to the I channel and Q channel, respectively, and captures the modulated signal output from the quadrature modulator, and outputs the captured modulated signal as I demodulated waveform data and Q demodulated waveform data of the IF frequency. The device is characterized by comprising: a signal analysis unit (40) that quadrature demodulates the harmonic waveform data; an error calculation unit (50) that calculates the difference between the I-phase and Q-phase of the amplitude IF frequency characteristics and the difference between the I-phase and Q-phase of the phase IF frequency characteristics for the signal generator from the I-demodulated waveform data and the Q-demodulated waveform data acquired for a plurality of IF frequencies; and an error correction unit (30) disposed between the waveform data storage unit and the D / A converter, having correction filters (33a, 33b) that correct the difference between the I-phase and Q-phase of the amplitude IF frequency characteristics and the difference between the I-phase and Q-phase of the phase IF frequency characteristics.

[0012] As described above, the error calculation unit calculates the difference between the I-phase and Q-phase of the amplitude IF frequency characteristics and the difference between the I-phase and Q-phase of the phase IF frequency characteristics for the signal generator from the I-demodulated waveform data and Q-demodulated waveform data acquired for multiple IF frequencies. The error correction unit then corrects the difference by applying, for example, an inverse characteristic filter to the difference between the I-phase and Q-phase of the amplitude IF frequency characteristics and the difference between the I-phase and Q-phase of the phase IF frequency characteristics. With this configuration, the difference between the I-phase and Q-phase of the amplitude IF frequency characteristics and the difference between the I-phase and Q-phase of the phase IF frequency characteristics, which were factors in the image signal in the signal path from the D / A converter to the quadrature modulator, can be appropriately corrected across the entire IF frequency, thereby reducing the level of the image signal generated within the quadrature modulation output band.

[0013] In the calibration device according to the present invention, the waveform data storage unit alternately outputs a multi-tone signal in which two or more tone signals having different IF frequencies are combined to the I channel and the Q channel, and while the multi-tone signal is being output to one of the I channel and the Q channel, no signal is output to the other of the I channel and the Q channel. The signal analysis unit may be configured to capture the modulation signal output from the quadrature modulator and perform quadrature demodulation on the captured modulation signal for each of the IF frequencies of the tone signals.

[0014] With this configuration, by using the single-sided multi-tone signal as calibration waveform data, it is possible to quickly and accurately measure the amplitude and phase IF frequency characteristics for each of the I channel and the Q channel.

[0015] In the calibration device according to the present invention, the error calculation unit calculates an IQ orthogonality error, and the error correction unit may include an orthogonality correction unit (32) that corrects the IQ orthogonality error with respect to the I waveform data and the Q waveform data output from the waveform data storage unit.

[0016] With this configuration, it is possible to appropriately correct the IQ orthogonality error, which was a cause of the image signal in the quadrature modulator, so that the level of the image signal generated within the quadrature modulation output band can be reduced.

[0017] The calibration device according to the present invention further includes a correction value calculation unit that uses the calculated IQ orthogonality error as α [rad] and calculates the first correction value A and the second correction value B used in the orthogonality correction unit by A = cos α / cos 2α and B = -tan α. The orthogonality correction unit includes a first multiplier (34a) that multiplies the I waveform data by the first correction value A, a second multiplier (34b) that multiplies the Q waveform data by the first correction value A, a third multiplier (34c) that multiplies the multiplication result of the second multiplier by the second correction value B, a fourth multiplier (34d) that multiplies the multiplication result of the first multiplier by the second correction value B, a first adder (35a) that adds the multiplication result of the first multiplier and the multiplication result of the third multiplier and outputs the result as the corrected I waveform data, and a second adder (35b) that adds the multiplication result of the second multiplier and the multiplication result of the fourth multiplier and outputs the result as the corrected Q waveform data. Such a configuration may be adopted.

[0018] With this configuration, it is possible to appropriately correct the IQ orthogonality error that has been a cause of the image signal in the quadrature modulator, so that the level of the image signal generated within the quadrature modulation output band can be reduced.

[0019] The calibration device according to the present invention further includes a control unit (70). The error calculation unit calculates the IQ gain balance. The control unit controls the waveform data storage unit, the signal analysis unit, the error calculation unit, and the error correction unit to correct the difference between the I phase and the Q phase of the amplitude IF frequency characteristic and the difference between the I phase and the Q phase of the phase IF frequency characteristic. Then, while the signal analysis unit measures the level of the image signal, a tracking correction of the IQ orthogonality and the IQ gain balance may be performed.

[0020] This configuration allows for the appropriate correction (rough adjustment) of the difference between the I-phase and Q-phase of the amplitude IF frequency characteristics, and the difference between the I-phase and Q-phase of the phase IF frequency characteristics, across the entire IF frequency. Subsequently, by measuring the image signal level and performing fine-tuning of the IQ orthogonality and IQ gain balance, the IQ orthogonality and IQ gain balance can be quickly set to optimal values.

[0021] The calibration method according to the present invention is a calibration method for a signal generator (10) comprising a D / A converter (11a, 11b) that performs D / A conversion of I waveform data and Q waveform data, respectively, and a quadrature modulator (12) that quadrature modulates the D / A converted signal and outputs it as a modulated signal, wherein the calibration method comprises a waveform data output step of outputting the I waveform data and Q waveform data of the IF frequency to the I channel and Q channel, respectively, and capturing the modulated signal output from the quadrature modulator, and demodulating the captured modulated signal into I demodulated waveform data and Q demodulated waveform data of the IF frequency. The method is characterized by including: a signal analysis step of quadrature demodulating waveform data; an error calculation step of calculating the difference between the I-phase and Q-phase of the amplitude IF frequency characteristics and the difference between the I-phase and Q-phase of the phase IF frequency characteristics for the signal generator from the I-demodulated waveform data and the Q-demodulated waveform data acquired for a plurality of IF frequencies; and an error correction step performed between the waveform data output step and the D / A conversion by the D / A converter, in which the difference between the I-phase and Q-phase of the amplitude IF frequency characteristics and the difference between the I-phase and Q-phase of the phase IF frequency characteristics are corrected by a correction filter.

[0022] This configuration allows for appropriate correction across the entire IF frequency to address the differences between the I-phase and Q-phase amplitude IF frequency characteristics, and the differences between the I-phase and Q-phase phase IF frequency characteristics, which were contributing factors to the image signal in the signal path from the D / A converter to the quadrature modulator. As a result, the level of the image signal generated within the quadrature modulation output bandwidth can be reduced.

[0023] The calibration method according to the present invention may also be configured such that, in the waveform data output step, a multitone signal obtained by combining two or more tone signals with different IF frequencies is alternately output to the I channel and the Q channel, and while the multitone signal is output to one of the I channel and the Q channel, no signal is output to the other of the I channel and the Q channel, and in the signal analysis step, the modulated signal output from the quadrature modulator is captured, and the captured modulated signal is quadrature demodulated for each of the IF frequencies of the tone signal.

[0024] This configuration allows for rapid and accurate measurement of the amplitude and phase IF frequency characteristics for each I channel and Q channel by using a single-sided multitone signal as calibration waveform data.

[0025] The calibration method according to the present invention may be configured such that, in the error calculation step, the IQ orthogonality error is calculated, and in the error correction step, an orthogonality correction step is performed to correct the I waveform data and Q waveform data output in the waveform data output step for the IQ orthogonality error.

[0026] This configuration allows for proper correction of the IQ orthogonality error, which was a factor in the image signal in the quadrature modulator, thereby reducing the level of the image signal generated within the quadrature modulation output bandwidth.

[0027] The calibration method according to the present invention further includes a correction value calculation step in which the calculated IQ orthogonality error is α [rad], and a first correction value A and a second correction value B used in the orthogonality correction step are calculated by A = cosα / cos2α and B = -tanα, wherein the orthogonality correction step may include a first multiplication step of multiplying the I waveform data by the first correction value A, a second multiplication step of multiplying the Q waveform data by the first correction value A, a third multiplication step of multiplying the result of the second multiplication step by the second correction value B, a fourth multiplication step of multiplying the result of the first multiplication step by the second correction value B, a first addition step of outputting the corrected I waveform data by adding the multiplication result of the first multiplication step and the multiplication result of the third multiplication step, and a second addition step of outputting the corrected Q waveform data by adding the multiplication result of the second multiplication step and the multiplication result of the fourth multiplication step.

[0028] This configuration allows for proper correction of the IQ orthogonality error, which was a factor in the image signal in the quadrature modulator, thereby reducing the level of the image signal generated within the quadrature modulation output bandwidth.

[0029] The calibration method according to the present invention may be configured such that, in the error calculation step, the IQ gain balance is calculated, and the execution of the waveform data output step, the signal analysis step, the error calculation step, and the error correction step is controlled to correct the difference between the I phase and Q phase of the amplitude IF frequency characteristics and the difference between the I phase and Q phase of the phase IF frequency characteristics, and then the IQ orthogonality and the IQ gain balance are fine-tuned while measuring the level of the image signal.

[0030] This configuration allows for the appropriate correction (rough adjustment) of the difference between the I-phase and Q-phase of the amplitude IF frequency characteristics, and the difference between the I-phase and Q-phase of the phase IF frequency characteristics, across the entire IF frequency. Subsequently, by measuring the image signal level and performing fine-tuning of the IQ orthogonality and IQ gain balance, the IQ orthogonality and IQ gain balance can be quickly set to optimal values. [Effects of the Invention]

[0031] According to the present invention, a calibration device and calibration method for a signal generator can be provided that can reduce the level of the image signal generated within the quadrature modulation output band. [Brief explanation of the drawing]

[0032] [Figure 1] This is a block diagram showing the configuration of a calibration device for a signal generator according to one embodiment of the present invention. [Figure 2] This is an explanatory diagram showing the configuration of the calibration signals output to the I channel and Q channel during calibration. [Figure 3] This is an explanatory diagram showing the configuration of the modulated signal captured during calibration. [Figure 4] This is an explanatory diagram for explaining IQ gain balance and IQ orthogonality error. [Figure 5] This is a flowchart explaining the calibration method for signal generators. [Modes for carrying out the invention]

[0033] Embodiments of the present invention will be described below with reference to the drawings.

[0034] (composition) Figure 1 is a block diagram showing the configuration of a calibration device 1 according to an embodiment of the present invention. As shown in Figure 1, the calibration device 1 according to this embodiment performs calibration on a signal generator 10 and includes a waveform data storage unit 20, an error correction unit 30, a signal analysis unit 40, an error calculation unit 50, a correction value calculation unit 60, a control unit 70, a storage unit 80, a display unit 81, an operation unit 82, and a switch 19. The calibration device 1 and the signal generator 10 constitute a signal generator 100 with a calibration function.

[0035] <Signal Generator> The signal generator 10 to be calibrated comprises digital-to-analog converters (hereinafter referred to as "DACs") 11a and 11b, a quadrature modulator 12, and a local oscillator 18. The quadrature modulator 12 comprises multipliers (mixers) 13a and 13b, a 90-degree phase shifter 14, and an adder 16.

[0036] DAC11a performs D / A conversion of the waveform data of the common-mode component (I component) in the baseband or IF frequency band (hereinafter referred to as "I waveform data") into analog waveform data and outputs it to the quadrature modulator 12. DAC11b performs D / A conversion of the waveform data of the quadrature component (Q component) in the baseband or IF frequency band (hereinafter referred to as "Q waveform data") into analog waveform data and outputs it to the quadrature modulator 12.

[0037] The local oscillator 18 generates a carrier signal of a predetermined carrier frequency under the control of the control unit 70 and outputs it to the quadrature modulator 12.

[0038] In the quadrature modulator 12, multiplier 13a multiplies the I component signal output from DAC 11a with the carrier signal output from local oscillator 18. Multiplier 13b multiplies the Q component signal output from DAC 11b with the carrier signal shifted by 90 degrees by the 90-degree phase shifter 14. Adder 16 adds the multiplication result of multiplier 13a and the multiplication result of multiplier 13b to output a modulated signal (RF signal). In this way, the quadrature modulator 12 quadrature modulates the carrier signal with analog I waveform data and Q waveform data and outputs it as a modulated signal.

[0039] <Waveform Data Storage Unit> The waveform data storage unit 20 stores digital waveform data and is configured to pre-store waveform data of the I and Q components of the baseband or IF frequency band prepared for testing the device under test (not shown) (hereinafter referred to as "test waveform data"). The waveform data storage unit 20 is also configured to pre-store waveform data of the I and Q components of the baseband prepared for calibrating the signal generator 10 (hereinafter referred to as "calibration waveform data"). The test waveform data and calibration waveform data are generated, for example, by a DSP (Digital Signal Processor) not shown.

[0040] When the signal generator 10 is being calibrated, the waveform data storage unit 20 outputs I-waveform data and Q-waveform data of the IF frequency to the I-channel and Q-channel respectively as calibration waveform data under the control of the control unit 70.

[0041] For example, the waveform data storage unit 20 alternately outputs a multitone signal, which is obtained by combining two or more tone signals with different IF frequencies, to the I channel and the Q channel as calibration waveform data. While a multitone signal is output to one of the I channel and the Q channel, no signal is output to the other of the I channel and the Q channel, or a zero signal is output to the other channel. Each tone signal is, for example, a sine wave. The above calibration waveform data is also called a "single-phase multitone signal".

[0042] Figure 2 is an explanatory diagram showing the configuration of calibration waveform data (calibration signals) output from the waveform data storage unit 20 to the I channel and Q channel during calibration. As shown in Figure 2, the waveform data storage unit 20 outputs a multitone signal to the I channel for a predetermined time (a predetermined number of samples), and then outputs a multitone signal to the Q channel for a predetermined time (a predetermined number of samples). While the waveform data storage unit 20 is outputting a multitone signal to the I channel for a predetermined time, it outputs a zero-value signal to the Q channel or no signal at all. Conversely, while the multitone signal is outputting to the Q channel for a predetermined time, it outputs a zero-value signal to the I channel or no signal at all. These signals will hereafter be referred to as "I-phase multitone" and "Q-phase multitone" using the channel on which the signal is output.

[0043] The calibration waveform data may be modified as appropriate to facilitate the measurement and calculation of correction items (such as IQ amplitude characteristic difference, IQ phase characteristic difference, IQ orthogonality error, and IQ gain balance) used to suppress the image signal.

[0044] <Error correction section> The error correction unit 30 is positioned between the waveform data storage unit 20 and the DACs 11a and 11b, and includes a first correction unit 31 and a second correction unit 32. The first correction unit 31 includes an I-phase correction filter 33a and a Q-phase correction filter 33b, and is configured to correct the difference between the I-phase and Q-phase of the amplitude IF frequency characteristics, and the difference between the I-phase and Q-phase of the phase IF frequency characteristics. The difference between the I-phase and Q-phase of the amplitude IF frequency characteristics, and the difference between the I-phase and Q-phase of the phase IF frequency characteristics, mainly arise from the IQ signal path difference from the DACs 11a and 11b to the quadrature modulator 12. The correction filters 33a and 33b are digital filters such as FIR (Finite Impulse Response) filters. The first correction unit 31 can also correct the IQ gain balance.

[0045] The second correction unit 32 of the error correction unit 30, also called the orthogonality correction unit, corrects the IQ orthogonality error in the I waveform data and Q waveform data output from the waveform data storage unit 20.

[0046] Specifically, the second correction unit 32 comprises a first multiplier 34a, a second multiplier 34b, a third multiplier 34c, a fourth multiplier 34d, a first adder 35a, and a second adder 35b. The first multiplier 34a multiplies the I waveform data by a first correction value A calculated by the correction value calculation unit 60 from the IQ orthogonality error. The second multiplier 34b multiplies the Q waveform data by the first correction value A. The third multiplier 34c multiplies the result of the multiplication by the second multiplier 34b by a second correction value B calculated by the correction value calculation unit 60 from the IQ orthogonality error. The fourth multiplier 34d multiplies the result of the multiplication by the first multiplier 34a by the second correction value B. The first adder 35a outputs corrected I waveform data by adding the multiplication result of the first multiplier 34a and the multiplication result of the third multiplier 34c. The second adder 35b outputs corrected Q waveform data by adding the multiplication result of the second multiplier 34b and the multiplication result of the fourth multiplier 34d.

[0047] Here, the first correction value A is calculated by A = cosα / cos2α, and the second correction value B is calculated by B = -tanα. α [rad] is the IQ orthogonality error.

[0048] The error correction unit 30 may be configured, for example, using an FPGA (Field Programmable Gate Array), or it may be implemented by software signal processing on a computer device.

[0049] <Switch> Switch 19 is connected to the adder 16 of the quadrature modulator 12 and switches between test mode and calibration mode under the control of the control unit 70. In test mode, switch 19 is connected to contact a, and a modulated signal (RF signal) with errors causing the image signal corrected is output to the device under test as a test signal. In calibration mode, switch 19 is connected to contact b, and the modulated signal is sent to the signal analysis unit 40.

[0050] <Signal analysis section> The signal analysis unit 40 captures the modulated signal output from the quadrature modulator 12 and quadrature demodulates the captured modulated signal into I demodulated waveform data and Q demodulated waveform data at the IF frequency. For example, when the above-mentioned single-phase multitone signal is used as calibration waveform data, the signal analysis unit 40 captures the modulated signal output from the quadrature modulator 12 and quadrature demodulates the captured modulated signal for each IF frequency of the tone signal.

[0051] The signal analysis unit 40 can, for example, perform A / D conversion, FFT (Fast Fourier Transform) processing, quadrature demodulation, power measurement, and other operations on the modulated signal output from the quadrature modulator 12 as needed. The signal analysis unit 40 may also be configured to include, for example, a vector signal analyzer.

[0052] Figure 3 is an explanatory diagram showing the configuration of the modulated signal captured by the analog-to-digital converter (ADC) of the signal analysis unit 40 during calibration. As shown in Figure 3, the data captured by the ADC of the signal analysis unit 40 and stored in the digital data storage unit is arranged alternately with data corresponding to I-phase multitone and data corresponding to Q-phase multitone. In the example in Figure 3, the data captured for a capture time of 200 μs triggered by an external trigger signal corresponds to the I-phase multitone, and the data captured for the next 200 μs corresponds to the Q-phase multitone. For the I-phase multitone data, the first and last predetermined length of data are not used, and the remaining data is used for FFT processing. The same applies to the Q-phase multitone data.

[0053] <Error calculation section> The error calculation unit 50 includes an IQ amplitude characteristic difference calculation unit 51, an IQ phase characteristic difference calculation unit 52, an IQ gain balance calculation unit 53, and an IQ orthogonality error calculation unit 54.

[0054] The IQ amplitude characteristic difference calculation unit 51 calculates the difference between the I-phase and Q-phase of the amplitude IF frequency characteristics for the signal generator 10 from the I-demodulated waveform data and Q-demodulated waveform data acquired for multiple IF frequencies.

[0055] The IQ phase characteristic difference calculation unit 52 calculates the difference between the I phase and Q phase of the IF frequency characteristics for the signal generator 10 from the I demodulated waveform data and Q demodulated waveform data acquired for multiple IF frequencies.

[0056] The IQ gain balance calculation unit 53 calculates the IQ gain balance based on the power of the modulated signal obtained by the signal analysis unit 40. Alternatively, the IQ gain balance calculation unit 53 may calculate the IQ gain balance based on the IF frequency characteristics of the amplitude calculated by the IQ amplitude characteristic difference calculation unit 51.

[0057] The IQ orthogonality error calculation unit 54 calculates the IQ orthogonality error based on the power of the modulated signal obtained by the signal analysis unit 40.

[0058] <Correction Value Calculation Unit> The correction value calculation unit 60 calculates correction values ​​used by the error correction unit 30 to reduce the image signal. For example, the correction value calculation unit 60 calculates the first correction value A and the second correction value B used by the second correction unit 32 of the error correction unit 30 from the IQ orthogonality error α [rad] calculated by the IQ orthogonality error calculation unit 54, using the formulas A = cosα / cos2α and B = -tanα.

[0059] <Department Head> The control unit 70 is composed of a computer including, for example, a CPU, ROM, RAM, etc., and controls the entire device of the calibration device 1 or the signal generator with calibration function 100. The control unit 70 controls the operation of, for example, the waveform data storage unit 20, the error correction unit 30, the signal analysis unit 40, the error calculation unit 50, the correction value calculation unit 60, the switch 19, etc. For example, the control unit 70 outputs a waveform specification signal to the waveform data storage unit 20 and determines the waveform data to be output from the waveform data storage unit 20.

[0060] The control unit 70 controls the waveform data storage unit 20, the signal analysis unit 40, the error calculation unit 50, and the error correction unit 30 to correct the difference between the I-phase and Q-phase of the amplitude IF frequency characteristics and the difference between the I-phase and Q-phase of the phase IF frequency characteristics (coarse adjustment). Then, the signal analysis unit 40 measures the level of the image signal and performs fine adjustment of the IQ orthogonality error and IQ gain balance.

[0061] The calibration device 1 may include a display unit 81, an operation unit 82, and a storage unit 80.

[0062] The display unit 81 is composed of a display device such as an LCD or CRT, and displays various content in accordance with control signals from the control unit 70. This content may include, for example, a list of multiple communication standards or a list of multiple waveform data.

[0063] The operation unit 82 is for user input and includes an input device such as a keyboard, touch panel, or mouse. Alternatively, the operation unit 82 may be configured so that the target of operation, such as buttons, soft keys, pull-down menus, or text boxes, is displayed on the display unit 81.

[0064] The memory unit 80 stores the errors and correction values ​​calculated by the error calculation unit 50 and the correction value calculation unit 60.

[0065] The control unit 70, signal analysis unit 40, error calculation unit 50, correction value calculation unit 60, and error correction unit 30 may each be configured in part or in whole by a computer.

[0066] (Factors in the image signal) The image signal is affected by factors such as the difference in IQ amplitude characteristics and IQ phase characteristics that occur in the IQ signal path from DAC11a,11b to quadrature modulator12, and the IQ gain balance error and IQ orthogonality error that occur in quadrature modulator12.

[0067] The IQ amplitude characteristic difference is the difference between the I-phase and Q-phase with respect to the frequency-amplitude characteristic in the IF frequency range. The IQ phase characteristic difference is the difference between the I-phase and Q-phase with respect to the frequency-phase characteristic in the IF frequency range.

[0068] The IQ gain balance error is the error when the signal P1 input to the quadrature modulator 12 is output as the signal P2, as shown in Fig. 4(a). Specifically, when the level of the I component of the signal P1 is V I and the level of the Q component is V Q , when the level of the I component of the signal P2 is V I ·d I and the level of the Q component is V Q ·d Q (d I ≠d Q ), the d I and d Q are referred to as the IQ gain balance error. Note that when d I =d Q , that is, when the levels of the I component and the Q component are balanced, there is no particular problem because level adjustment can be performed in the subsequent circuit.

[0069] The IQ orthogonality error is the error indicating that the intersection angle between the I-phase axis and the Q-phase axis deviates from π / 2 [rad]. For example, when the angle formed by the Q-phase axis with respect to the I-phase axis is (π / 2 + α) [rad], this α is referred to as the orthogonality error.

[0070] The error calculation unit 50 measures the power components of the modulation signal output from the quadrature modulator 12 by the signal analysis unit 40 when using appropriate calibration waveform data, thereby obtaining the gain balance error d I of the I-phase and the gain balance error d Q of the Q-phase, as well as the orthogonality error α.

[0071] (Calibration method) Next, the calibration method will be described.

[0072] The difference between the I-phase and Q-phase of the frequency-amplitude characteristics and frequency-phase characteristics across the required IF frequency range (not just at a single point) in the signal path from DACs 11a and 11b to the quadrature modulator 12 is obtained and used as a correction value. In addition, the IQ quadrature error of the quadrature modulator 12 is measured and used as a correction value. The level of the image signal generated within the modulation output band is reduced mainly by eliminating the difference in frequency characteristics (aligning them to one frequency characteristic) and reducing the IQ quadrature error across the entire required IF frequency range.

[0073] Specifically, for example, the level of the image signal generated within the modulation output bandwidth can be reduced by correcting it as follows.

[0074] (1) The frequency characteristic difference of the signal path from DAC11a,11b to the quadrature modulator 12, that is, the difference between IF frequency vs. amplitude characteristic and IF frequency vs. phase characteristic, is corrected by an inverse characteristic filter on the waveform data input to DAC11a,11b so that the difference between the I path and the Q path is eliminated (coarse adjustment).

[0075] (2) Correct the IQ orthogonality error of the quadrature modulator 12 so that it is eliminated. For example, the IQ orthogonality error is fine-tuned so that the level of the image signal is optimized, and the result of the fine-tuning is used as the correction value and recorded (fine-tuning).

[0076] Correction for frequency characteristic differences in the I / Q signal path is performed before correcting the IQ orthogonality error.

[0077] Figure 5 is a flowchart showing an example of a calibration method.

[0078] First, set the calibration waveform data and the various conditions of the calibration device 1 (step S1).

[0079] Specifically, the initial phase of the single-phase multitone signal is set so that the PAPR (Peak Power to Average Power Ratio) of the signal is minimized. The single-phase multitone signal is used to determine the difference in frequency characteristics of the IQ signal path. The multitone signal is output sequentially from the I-phase DAC11a and the Q-phase DAC11b. The duration of the multitone signal is, for example, 200 [μs]. A duration that is easy to handle is selected according to the output bandwidth of the signal generator 10 (< the analysis bandwidth of the signal analysis unit 40).

[0080] For the N tone signals that make up a multitone signal, the initial phase θ k [rad] can be set, for example, by the following formula. θ k =(k(k-1) / (N-1))π,(0≦k≦N-1)

[0081] For the tone signal frequency, for example, start at 5 MHz and go up to 170 MHz, with intervals of 5 MHz. This frequency interval will be the same as the interval of the frequency points that will be saved as correction values.

[0082] Regarding the selection of frequency intervals, considering that the waveform data storage unit 20, which is an arbitrary waveform generator, repeatedly outputs waveform data, it is preferable that the connection from the end to the beginning of the signal is continuous. Conversely, if it is not continuous, unwanted waves will be generated during repeated output. In the case of waveform data with a time length of 200 [μs], if the tone signal has a frequency that is an integer multiple of 1 / (200 [μs]) = 50 kHz (integer division of a period of 200 [μs]), it will be continuous even when repeatedly output. For example, it is possible to intentionally make the frequency intervals non-uniform in order to avoid hardware-induced spurious signals during measurement. However, even in that case, it is preferable to select the frequency so that the center frequency of the FFT bin in the signal analysis unit 40 matches the frequency of the tone signal.

[0083] Next, rough adjustments are made. In rough adjustments, the FIR filters that make up correction filters 33a and 33b are determined.

[0084] Calibration waveform data output from the waveform data storage unit 20 is input to DACs 11a and 11b, and the modulated signal (RF signal) output from the quadrature modulator 12 is captured or digitized by the signal analysis unit 40 (step S2). At this time, a trigger signal is output at the beginning of the waveform data to trigger the capture. The calibration waveform data is a single-phase multitone signal.

[0085] The signal analysis unit 40 performs FFT processing on the captured data. In doing so, it performs FFT processing only on the necessary parts that it wants to observe. The signal analysis unit 40 performs quadrature demodulation as needed to obtain I-waveform demodulated data and Q-waveform demodulated data of the IF frequency.

[0086] Next, the error calculation unit 50 analyzes the data obtained by the signal analysis unit 40 to acquire the IF frequency vs. amplitude characteristics and the IF frequency vs. phase characteristics (step S3).

[0087] Specifically, the error calculation unit 50 extracts only the frequency points of the multitone signal from the FFT processing results. Then, it calculates the complex number of the deviation angle corresponding to each frequency point of the multitone signal, which is the extracted result for the I-phase and Q-phase sides, respectively. Since the initial phase of the calibration waveform data is the same for the I-phase and Q-phase, the difference in the calculated deviation angle is the difference between the "IF frequency point vs. phase characteristic" of the signal path. Cycle slip correction is performed on this difference. The inverse characteristic of the calculated cycle slip correction is used as the phase characteristic correction value.

[0088] Furthermore, the amplitude is calculated in the same manner for both the I-phase and Q-phase. The reciprocal of the amplitude ratio between the I-phase and Q-phase is used as the correction value for the amplitude characteristics.

[0089] Specifically, as correction value data, A(ω B ),Δt(ω B ) can be obtained.

[0090] Here, a certain IF frequency ω B For A(ω B ) and Δt(ω B) indicates the amplitude difference and phase (delay time) difference between phase I and phase Q. B [rad / s] represents the IF frequency (baseband frequency) expressed in terms of angular frequency. The difference in frequency characteristics between the I phase and the Q phase can be expressed as a single complex number: A(ω B )exp(-jω B ×Δt(ω B )) where j represents the imaginary unit.

[0091] The characteristic difference on the negative IF frequency side can also be calculated from the characteristic difference on the positive IF frequency side. The characteristic difference on the negative IF frequency side is the complex conjugate of the frequency characteristic difference at the positive IF frequency.

[0092] The IF frequency characteristics at locations where the acquisition point and application point of the correction value are different are calculated using interpolation.

[0093] Next, the correction value calculation unit 60 generates FIR tap information for correction filters 33a and 33b, which are FIR filters, from the IF frequency vs. amplitude characteristics and IF frequency vs. phase characteristics (step S4). The FIR filters are applied to the calibration waveform data during the fine-tuning stage.

[0094] Specifically, for example, if the frequency response of an FIR filter is set to 2.5 MHz intervals, and the frequency points for acquiring correction values ​​are set to 5 MHz intervals, the frequency response data for frequency points for which correction values ​​have not been acquired is calculated by interpolation. Also, since only the frequency response at positive IF frequencies is measured, the negative IF frequency response data necessary for generating the FIR filter is calculated from the frequency response data at positive IF frequencies.

[0095] The FIR filter coefficients are set by applying the correction values ​​obtained from the rough adjustment.

[0096] Next, fine-tuning is performed. This fine-tuning involves correcting the IQ orthogonality error and the IQ gain balance.

[0097] The calibration waveform data output from the waveform data storage unit 20 is input to the error correction unit 30, and the modulation signal (RF signal) output from the quadrature modulator 12 is captured by the signal analysis unit 40 (step S5). The calibration waveform data does not have to be a single-phase multi-tone signal, and any waveform data can be used as long as the IQ gain balance, IQ orthogonality error, and level of the image signal can be obtained.

[0098] Next, the error calculation unit 50 analyzes the data of the modulation signal captured by the signal analysis unit 40 to obtain the IQ gain balance, IQ orthogonality error, and level of the image signal (step S6).

[0099] Next, the control unit 70 compares the IMRR calculated from the level of the image signal with a predetermined reference value. If the IMRR is greater than the reference value (NO in step S7), the correction value calculation unit 60 changes the correction value and / or resets the correction filters 33a and 33b (step S8), and returns to step S5. The correction value is changed in a direction that allows the tracking to end promptly.

[0100] If the IMRR is less than or equal to the reference value (YES in step S7), the current correction value is stored in the storage unit 80 and set in the error correction unit 30 (step S9).

[0101] Calibration is completed by the above operations.

[0102] s As described above, in the tracking during fine adjustment, the correction values of the IQ orthogonality error and IQ gain balance are chased to the optimum values. While applying the correction value in the error correction unit 30 to the waveform data output from the waveform data storage unit 20 and measuring the IMRR with the signal analysis unit 40, the optimum value is searched for.

[0103] <Calculation and Application of FIR Filter Coefficients> The purpose of applying the correction value in coarse adjustment is to suppress the difference in the IF band frequency characteristics in the section from the DACs 11a and 11b to the quadrature modulator 12. Therefore, a FIR filter that cancels the difference from the I-phase signal path side may be applied to the Q-phase signal path side.

[0104] Since the correction value is calculated by taking the difference between the frequency characteristics of the I-phase signal path and the Q-phase signal path, the frequency characteristics that form the basis of the Q-phase FIR filter coefficients should be the inverse characteristics of the correction value data (including interpolation).

[0105] The discrete frequency characteristics determined through the above process are converted into an impulse response using IDFT (Inverse Discrete Fourier Transform). The columns are rearranged to create a causal system, that is, so that the impulse response starts from time 0. Finally, since this is an FIR filter applied only to the Q phase, only the real components are extracted. In this way, the filter coefficients of the FIR filter for the Q phase signal path can be determined.

[0106] On the other hand, the I-phase signal path is designed to output a signal with the same delay time as the Q-phase signal path, without changing the IF frequency-amplitude characteristics. (Since the phase will change by the fixed delay time, the IF frequency-phase characteristics will fluctuate accordingly. It is important that the relative values ​​and difference values ​​between the I-phase and Q-phase sides remain constant with respect to the IF frequency.)

[0107] In this case, the IF frequency characteristics of the full-pass filter are converted into an impulse response using IDFT. By rearranging the columns to create a causal system, the signal is output with the same delay time as the Q-phase signal path.

[0108] <Correlation error correction value> The correction value calculation unit 60 calculates angle correction value A and angle correction value B as follows, when the IQ orthogonality error value at one IF frequency is α [rad]. A = cosα / cos2α, B = -tanα

[0109] In the second correction unit 32 of Figure 1, angle correction value A is applied to the first multiplier 34a and the second multiplier 34b, and angle correction value B is applied to the third multiplier 34c and the fourth multiplier 34d. The IQ gain balance can be corrected by changing the filter coefficients of the correction filters 33a and 33b.

[0110] (effect) For example, across the entire bandwidth of 160 MHz (+ / - 80 MHz), the minimum IMRR within the bandwidth was conventionally around 34 dB, but by applying the embodiment of the present invention, it can be reduced to around 50 dB, an improvement of approximately 16 dB. This makes it possible to generate high-quality wireless signals and improve the reliability of, for example, EVM measurement results of the object under test. [Industrial applicability]

[0111] As described above, the present invention has the effect of reducing the level of the image signal generated within the quadrature modulation output band, and is useful as a calibration device and calibration method for a signal generator. [Explanation of Symbols]

[0112] 1 Calibration device 10 Signal Generator 11a, 11b Digital-to-Analog Converter (DAC) 12. Quadrature modulator 13a, 13b Multiplier (Mixer) 14 90 degree phase shifter 16 Adder 18 Local Oscillator 19 switches 20 Waveform data storage unit 30 Error correction section 31. First Correction Section 32 Second Correction Section 33a, 33b Correction filters 34a, 34b, 34c, 34d multipliers 35a, 35b Adder 40 Signal analysis section 50 Error calculation section 51 IQ Amplitude Characteristic Difference Calculation Unit 52 IQ phase characteristic difference calculation section 53 IQ Gain Balance Calculation Unit 54 IQ Orthogonality Error Calculation Unit 60 Correction Value Calculation Unit 70 Control Unit 80 Storage section 81 Display section 82 Control section 100 Signal generator with calibration function

Claims

1. A calibration device (1) for a signal generator (10) comprising: D / A converters (11a, 11b) that perform D / A conversion of I waveform data and Q waveform data, respectively; and a quadrature modulator (12) that quadrature modulates the signal after D / A conversion and outputs it as a modulated signal, A waveform data storage unit (20) outputs the I-waveform data and Q-waveform data of the IF frequency to the I-channel and Q-channel, respectively. A signal analysis unit (40) captures the modulated signal output from the quadrature modulator and quadrature demodulates the captured modulated signal into I demodulated waveform data and Q demodulated waveform data of the IF frequency, An error calculation unit (50) calculates the difference between the I-phase and Q-phase of the amplitude IF frequency characteristics and the difference between the I-phase and Q-phase of the phase IF frequency characteristics for the signal generator from the I-demodulated waveform data and the Q-demodulated waveform data acquired for a plurality of IF frequencies, An error correction unit (30) is disposed between the waveform data storage unit and the D / A converter and has correction filters (33a, 33b) that correct the difference between the I phase and Q phase of the amplitude IF frequency characteristics and the difference between the I phase and Q phase of the phase IF frequency characteristics. A calibration device characterized by being equipped with the following features.

2. The waveform data storage unit alternately outputs a multitone signal, which is obtained by combining two or more tone signals with different IF frequencies, to the I channel and the Q channel, and while the multitone signal is being output to one of the I channel and the Q channel, it does not output a signal to the other of the I channel and the Q channel. The signal analysis unit captures the modulated signal output from the quadrature modulator and quadrature demodulates the captured modulated signal for each IF frequency of the tone signal. The calibration apparatus according to claim 1.

3. The error calculation unit calculates the IQ orthogonality error, The calibration apparatus according to claim 2, wherein the error correction unit includes an orthogonality correction unit (32) that corrects the IQ orthogonality error with respect to the I waveform data and Q waveform data output from the waveform data storage unit.

4. The system further includes a correction value calculation unit that calculates a first correction value A and a second correction value B used in the orthogonality correction unit, where A = cosα / cos2α and B = -tanα, with the calculated IQ orthogonality error being α. The orthogonality correction unit is, A first multiplier (34a) multiplies the I waveform data by the first correction value A, A second multiplier (34b) multiplies the Q waveform data by the first correction value A, A third multiplier (34c) multiplies the result of the second multiplier by the second correction value B, A fourth multiplier (34d) multiplies the multiplication result of the first multiplier by the second correction value B, A first adder (35a) outputs the corrected I waveform data by adding the multiplication result of the first multiplier and the multiplication result of the third multiplier, The calibration apparatus according to claim 3, further comprising a second adder (35b) that outputs corrected Q waveform data by adding the multiplication result of the second multiplier and the multiplication result of the fourth multiplier.

5. The control unit (70) is further provided, The error calculation unit calculates the IQ gain balance, The control unit controls the waveform data storage unit, the signal analysis unit, the error calculation unit, and the error correction unit to correct the difference between the I-phase and Q-phase of the amplitude IF frequency characteristics, and the difference between the I-phase and Q-phase of the phase IF frequency characteristics. Then, the signal analysis unit measures the level of the image signal and performs fine-tuning of the IQ orthogonality and the IQ gain balance. The calibration apparatus according to claim 4.

6. A calibration method for a signal generator (10) comprising D / A converters (11a, 11b) that perform D / A conversion on I-waveform data and Q-waveform data, respectively, and a quadrature modulator (12) that quadrature modulates the D / A converted signal and outputs it as a modulated signal, wherein A waveform data output step that outputs the I-waveform data and Q-waveform data of the IF frequency to the I-channel and Q-channel, respectively, A signal analysis step in which the modulated signal output from the quadrature modulator is captured, and the captured modulated signal is quadrature-demodulated into I-demodulated waveform data and Q-demodulated waveform data of the IF frequency, An error calculation step is performed to calculate the difference between the I-phase and Q-phase of the amplitude IF frequency characteristics and the difference between the I-phase and Q-phase of the phase IF frequency characteristics for the signal generator from the I-demodulated waveform data and the Q-demodulated waveform data obtained for a plurality of IF frequencies. An error correction step is performed between the waveform data output step and the D / A conversion by the D / A converter, in which the difference between the I-phase and Q-phase of the amplitude IF frequency characteristics and the difference between the I-phase and Q-phase of the phase IF frequency characteristics are corrected by a correction filter. A calibration method characterized by including the following.

7. In the waveform data output step, a multitone signal obtained by combining two or more tone signals with different IF frequencies is alternately output to the I channel and the Q channel, and while the multitone signal is being output to one of the I channel and the Q channel, no signal is output to the other of the I channel and the Q channel. In the signal analysis step, the modulated signal output from the quadrature modulator is captured, and the captured modulated signal is quadrature-demodulated for each IF frequency of the tone signal. The calibration method according to claim 6.

8. In the error calculation step described above, the IQ orthogonality error is calculated, The calibration method according to claim 7, wherein the error correction step involves performing an orthogonality correction step to correct the I-waveform data and Q-waveform data output in the waveform data output step by correcting the I-Q orthogonality error.

9. The method further includes a correction value calculation step in which the calculated IQ orthogonality error is denoted as α, and the first correction value A and the second correction value B used in the orthogonality correction step are calculated as A = cosα / cos2α and B = -tanα, In the orthogonality correction step, A first multiplication step in which the I waveform data is multiplied by the first correction value A, A second multiplication step in which the Q waveform data is multiplied by the first correction value A, A third multiplication step in which the multiplication result of the second multiplication step is multiplied by the second correction value B, A fourth multiplication step in which the multiplication result of the first multiplication step is multiplied by the second correction value B, A first addition step that outputs the corrected I waveform data by adding the multiplication result of the first multiplication step and the multiplication result of the third multiplication step, The calibration method according to claim 8, further comprising: a second addition step of adding the multiplication result of the second multiplication step and the multiplication result of the fourth multiplication step and outputting the corrected Q waveform data.

10. In the error calculation step, the IQ gain balance is calculated, The execution of the waveform data output step, the signal analysis step, the error calculation step, and the error correction step is controlled to correct the difference between the I-phase and Q-phase of the amplitude IF frequency characteristics and the difference between the I-phase and Q-phase of the phase IF frequency characteristics, and then the IQ orthogonality and the IQ gain balance are fine-tuned while measuring the level of the image signal. The calibration method according to claim 9.

Images