CN108020803B - A kind of arbitrary waveform generator amplitude frequency sweep estimation calibration method - Google Patents

Abstract

The present invention provides a method for estimating and calibrating amplitude sweep frequency of an arbitrary waveform generator. By cooperating with the existing method of obtaining calibration sample points at equal intervals, the frequency response curve is obtained by linear sweep mode, and the frequency response curve is intensively sampled. , to obtain each extreme point, these points are the points that must be selected for the calibration frequency points, and then insert some non-extreme points in the extreme points, and the number of calibration sample frequency points obtained is the same as the number of equally spaced calibration samples, so in Compared with the original technology, the time complexity does not increase, and it is guaranteed that no "key point" of the frequency response curve is missed, which not only realizes the fully automatic operation of the amplitude calibration, but also improves the accuracy of the amplitude calibration. The invention greatly improves the accuracy of the calibration data without increasing the time complexity, and is especially suitable for an arbitrary waveform generator circuit with unsatisfactory analog channels and insufficient amplitude flatness.

Description

A Calibration Method for Amplitude Sweep Estimation of Arbitrary Waveform Generator

technical field

The invention relates to the field of amplitude calibration of arbitrary waveform generators, in particular to a method for estimating and calibrating amplitude sweep frequency of arbitrary waveform generators.

Background technique

Arbitrary waveform generator is an editable multifunctional signal generating device, which can generate regular waveforms such as sine, square wave, pulse wave, etc., and can also generate special application waveforms such as electrocardiogram, multi-tone, Gaussian pulse, etc., which can be applied to integrated Circuit testing, complex electromagnetic environment construction, satellite navigation communication testing and other fields.

Arbitrary waveform generators have high operating bandwidth performance, wide dynamic output range and high control precision requirements, which put forward extremely high requirements for the design of signal analog channels. When the hardware circuit is relatively fixed, designing an efficient and reliable calibration scheme will also greatly improve the accuracy of the output amplitude of the arbitrary waveform generator.

The time complexity of the calibration scheme depends on two factors: (1) the number of amplifiers and attenuators controlled by the hardware circuit; (2) the number of calibration samples selected within the working bandwidth.

The output end of the arbitrary waveform generator is connected to the input end of the receiver through the RF coaxial cable, and the back end is connected to the corresponding interface of the receiver through LAN, GPIB or serial port. When starting the calibration, the arbitrary waveform generator sends a connection establishment message to the receiver through the interface and waits for the receiver's feedback. After the connection is established successfully, the arbitrary waveform generator realizes channel state switching by adjusting the amplifier and attenuator, and adjusts the signal output through the circuit fine-tuning controller magnitude. When the channel state changes or the fine-tuning controller changes, the arbitrary waveform generator reads the measurement data of the receiver through the interface and stores it as a file for recall.

A commonly used calibration method is to preset calibration sample frequency points f ₀ , f ₁ ... f _n-1 and amplitude points (amplitude points for short) V ₀ , V ₁ ... V _m-1 at the instrument end. Frequency points and amplitude points are selected at equal intervals, that is, f _k2 -f _k1 = (k ₂ -k ₁ )×BW/n, Vk ₂ -Vk ₁ =(k ₂ -k ₁ )×V _pp /m, BW means The operating bandwidth of the instrument, V _pp represents the output peak-to-peak value. Assume that the adjustable range of the circuit fine-tuning controller is q gears from D ₀ to D _q-1 .

Usually the calibration process is as follows:

(1) Switch the channel status to CH _x , set the output frequency to the calibration sample frequency point;

(2) Change the gear position of the fine-tuning controller to D _t (0≤t≤q), and read back the measured value DV _t of the receiver at this time;

(3) If |DV _t -V _y |≤error threshold (0≤y≤m), that is, the absolute value of the error is less than the preset error threshold, record the fine-tuning control amount DV _t at this time, and set the effective flag of the data;

(4) Traverse all channel states and calibration sample frequency points, and store all corresponding DV _t .

Its data storage structure is shown in Figure 1:

The last column is the identifier. If it is 0, it means that the error is greater than the threshold and cannot be used as valid calibration data; if it is 1, it means that the error is within the threshold and the calibration data is valid.

The commonly used calibration method adopts the manual preset definition method in the selection of calibration frequency points, and the frequency points selected by all arbitrary waveform generators are the same, without taking into account the individual differences of hardware circuits, nor the frequency response curve The influence of unevenness on the selection of calibration frequency points.

When the frequency response curve is relatively flat, or presents a complete linearity, the frequency point error of equal interval calibration is very small and close to the actual frequency response curve. If the frequency response curve has more concave-convex "pits" or "bulges", using equal interval calibration may miss some key points, resulting in large errors. As shown in Figure 2, the left half is the frequency response curve obtained by the actual frequency sweep, the upper right part is the frequency response curve connected with equal interval calibration frequency points, and the shaded area of the lower right part reflects the actual frequency response curve and the equal interval calibration frequency points. The error size of the point frequency response curve.

In the case of a fixed analog channel, how to select fewer calibration frequency points and at the same time meet the requirements of the control accuracy of the arbitrary waveform generator will be the problem to be solved by the present invention.

Contents of the invention

Aiming at the problem of missing "key points" when the existing arbitrary waveform generator calibrates the frequency points at equal intervals, the present invention provides an arbitrary waveform generator amplitude sweep estimation calibration method, by first obtaining the extreme points of the frequency response curve The method ensures that the key points of the frequency response curve are not missed, and the accuracy of the amplitude calibration is improved.

The present invention adopts following technical scheme:

A method for estimating and calibrating the amplitude sweep frequency of an arbitrary waveform generator, which cooperates with the existing method for obtaining calibration samples at equal intervals, includes the following steps:

Step 1: The output terminal of the arbitrary waveform generator is connected to the input terminal of the spectrum analyzer through the RF coaxial cable, and the rear end of the arbitrary waveform generator is connected to the corresponding interface of the spectrum analyzer through LAN, GPIB or serial port;

Step 2: The arbitrary waveform generator performs a linear frequency sweep, the spectrum analyzer maintains the maximum frequency response curve, reads the frequency response data through LAN, GPIB or serial port, and obtains the frequency response curve;

Step 3: From the frequency response curve, obtain n1 extreme frequency points f ₀ , f ₁ ... f _n1-1 from small to large in numerical value through intensive sampling, and form an array of extreme frequency points;

Step 4: Insert the bandwidth start f _start and stop frequency point f _stop into the extreme value frequency point array to form [f _start , f ₀ , f ₁ ,... f _n1-1 , f _stop ] a total of n1+2 calibration frequency points point;

The method of obtaining calibration samples at equal intervals to obtain the number of calibration samples is n, and it is necessary to insert n-n1-2 intermediate calibration samples into n1+2 calibration frequency points to obtain complete n calibration samples. Insert The operation steps include:

a. Find the largest distance between two adjacent calibration frequency points, that is, the maximum value in f _p -f _p-1 . Δf is the original equal interval calibration sample point interval size. After finding the maximum value , the maximum value is divided by the interval size, that is (f _p -f _p-1 )/Δf, and the result is rounded down to Int;

b. If Int>1, insert Int-1 calibration samples between f p _- 1 and f _p , namely f _p-1 +Δf, f _p-1 +2Δf..., f _p-1 +( Int-1) Δf;

c. If Int≤1, insert one calibration sample point between f _p-1 and f _p , namely (f _p-1 + f _p )/2;

d. Return to step a until n-n1-2 intermediate calibration sample points are inserted;

Step 5: The arbitrary waveform generator starts the calibration thread and stores the calibration data.

Preferably, the n is greater than n1.

Preferably, said 1≤p≤n1+1.

The beneficial effects that the present invention has are:

The arbitrary waveform generator amplitude sweep frequency estimation calibration method provided by the present invention uses a linear frequency sweep method to obtain a frequency response curve, intensively samples the frequency response curve, and obtains each extreme point, which is a must-selected point for the calibration frequency point, and then Insert some non-extreme points in the extreme points, and the number of calibration sample frequency points obtained is the same as the number of equally spaced calibration samples, so the time complexity does not increase compared with the original technology, and it is guaranteed that there is no omission The "key point" of the frequency response curve not only realizes the fully automatic operation of the amplitude calibration, but also improves the accuracy of the amplitude calibration.

Description of drawings

Fig. 1 is a schematic diagram of calibration data storage structure.

Fig. 2 is a schematic diagram of frequency point error calibration with uneven frequency response curves at equal intervals.

Figure 3 is a dense sampling diagram of the frequency response curve.

Figure 4 is an error analysis diagram of the calibration method for amplitude sweep estimation of the arbitrary waveform generator.

Detailed ways

The specific embodiment of the present invention will be further described below in conjunction with accompanying drawing and specific embodiment:

Combining Figure 3 and Figure 4, a calibration method for amplitude sweep estimation of an arbitrary waveform generator, which cooperates with the existing method of obtaining calibration samples at equal intervals, includes the following steps:

Step 1: The output terminal of the arbitrary waveform generator is connected to the input terminal of the spectrum analyzer through the RF coaxial cable, and the rear end of the arbitrary waveform generator is connected to the corresponding interface of the spectrum analyzer through LAN, GPIB or serial port.

Step 2: The arbitrary waveform generator performs a linear frequency sweep, the spectrum analyzer maintains the maximum frequency response curve, and reads the frequency response data through LAN, GPIB or serial port to obtain the frequency response curve.

Step 3: Obtain n1 extreme frequency points f ₀ , f ₁ . . . f _n1-1 from the frequency response curve through intensive sampling to form an array of extreme frequency points. As shown in Figure 3.

Step 4: Insert the bandwidth start f _start and stop frequency point f _stop into the extreme value frequency point array to form [f _start , f ₀ , f ₁ ,... f _n1-1 , f _stop ] a total of n1+2 calibration frequency points point;

The original method of obtaining calibration samples at equal intervals obtains the number of calibration samples as n, so it is necessary to insert n-n1-2 intermediate calibration samples in n1+2 calibration frequency points to obtain a complete n calibration Sample point, insert operation steps include:

a. Find the largest distance between two adjacent calibration frequency points, that is, the maximum value in f _p -f _p-1 . Δf is the original equal interval calibration sample point interval size. After finding the maximum value , the maximum value is divided by the interval size, that is (f _p -f _p-1 )/Δf, and the result is rounded down to Int;

b. If Int>1, insert Int-1 calibration samples between f p _- 1 and f _p , namely f _p-1 +Δf, f _p-1 +2Δf..., f _p-1 +( Int-1) Δf;

c. If Int≤1, insert one calibration sample point between f _p-1 and f _p , namely (f _p-1 + f _p )/2;

d. Return to step a until n-n1-2 intermediate calibration sample points are inserted.

Step 5: The arbitrary waveform generator starts the calibration thread and stores the calibration data.

The shaded area in FIG. 4 is obviously much smaller than the shaded area in FIG. 2 , and the present invention greatly improves the accuracy of the calibration data without increasing the time complexity.

The present invention adopts the method of pre-estimating the extremum point by sweeping the frequency in advance, without increasing the frequency point of the calibration sample, and improves the accuracy of the calibration data, especially suitable for arbitrary waveform generators with unsatisfactory analog channels and insufficient amplitude flatness circuit.

Of course, the above descriptions are not intended to limit the present invention, and the present invention is not limited to the above examples. Changes, modifications, additions or replacements made by those skilled in the art within the scope of the present invention shall also belong to the present invention. protection scope of the invention.

Claims (2)

1. an arbitrary waveform generator amplitude sweep frequency estimation calibration method, is characterized in that, cooperates with the method that existing equal interval obtains calibration sample point, comprises the following steps: Step 1: The output terminal of the arbitrary waveform generator is connected to the input terminal of the spectrum analyzer through the RF coaxial cable, and the rear end of the arbitrary waveform generator is connected to the corresponding interface of the spectrum analyzer through LAN, GPIB or serial port; Step 2: The arbitrary waveform generator performs a linear frequency sweep, the spectrum analyzer maintains the maximum frequency response curve, reads the frequency response data through LAN, GPIB or serial port, and obtains the frequency response curve; Step 3: From the frequency response curve, obtain n1 extreme frequency points f ₀ , f ₁ ... f _n1-1 from small to large in numerical value through intensive sampling, and form an array of extreme frequency points; Step 4: Insert the bandwidth start f _start and stop frequency point f _stop into the extreme value frequency point array to form [f _start , f ₀ , f ₁ ,... f _n1-1 , f _stop ] a total of n1+2 calibration frequency points point; The method of obtaining calibration samples at equal intervals to obtain the number of calibration samples is n, and it is necessary to insert n-n1-2 intermediate calibration samples into n1+2 calibration frequency points to obtain complete n calibration samples. Insert The operation steps include: a. Find the largest distance between two adjacent calibration frequency points, that is, the maximum value in f _p -f _p-1 . Δf is the original equal interval calibration sample point interval size. After finding the maximum value , the maximum value is divided by the interval size, that is (f _p -f _p-1 )/Δf, and the result is rounded down to Int; b. If Int>1, insert Int-1 calibration samples between f p _- 1 and f _p , namely f _p-1 +Δf, f _p-1 +2Δf..., f _p-1 +( Int-1) Δf; c. If Int≤1, insert one calibration sample point between f _p-1 and f _p , namely (f _p-1 + f _p )/2; d. Return to step a until n-n1-2 intermediate calibration sample points are inserted; Step 5: The arbitrary waveform generator starts the calibration thread and stores the calibration data. 2. a kind of arbitrary waveform generator amplitude sweep estimation calibration method according to claim 1 is characterized in that, described n is greater than n1.