CN105720983A - Error estimation method and device for time interleaving analog-digital conversion system - Google Patents

Error estimation method and device for time interleaving analog-digital conversion system Download PDF

Info

Publication number
CN105720983A
CN105720983A CN201610044047.7A CN201610044047A CN105720983A CN 105720983 A CN105720983 A CN 105720983A CN 201610044047 A CN201610044047 A CN 201610044047A CN 105720983 A CN105720983 A CN 105720983A
Authority
CN
China
Prior art keywords
error
function
frequency
time
fourier transformation
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN201610044047.7A
Other languages
Chinese (zh)
Inventor
李锐锐
谭洪舟
李宇
蔡彬
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
SYSU CMU Shunde International Joint Research Institute
National Sun Yat Sen University
Original Assignee
SYSU CMU Shunde International Joint Research Institute
National Sun Yat Sen University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by SYSU CMU Shunde International Joint Research Institute, National Sun Yat Sen University filed Critical SYSU CMU Shunde International Joint Research Institute
Priority to CN201610044047.7A priority Critical patent/CN105720983A/en
Publication of CN105720983A publication Critical patent/CN105720983A/en
Pending legal-status Critical Current

Links

Classifications

    • H—ELECTRICITY
    • H03—ELECTRONIC CIRCUITRY
    • H03M—CODING; DECODING; CODE CONVERSION IN GENERAL
    • H03M1/00—Analogue/digital conversion; Digital/analogue conversion
    • H03M1/10—Calibration or testing
    • H03M1/1071—Measuring or testing

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Analogue/Digital Conversion (AREA)

Abstract

The invention relates to an error estimation method and device for a time interleaving analog-digital conversion (TIADC) system.A test signal is input into the multi-channel TIADC system to obtain an error signal, then Fourier transform is conducted on the error signal to obtain frequency spectrum amplitude data, gain errors and biased errors of all sub-channels are estimated in the presence of time errors, then the gain errors and the biased errors are corrected and eliminated, and finally the time errors are estimated.Three mismatch errors can be estimated simultaneously online; when the outside environment changes, there is no need to adjust a filter system or redesign a hardware circuit, and people only need to input the test signal again to obtain the three mismatch errors; there is no limitation on the frequency of the input signal; the estimation precision of the three mismatch errors is high.

Description

Error estimation and device for time-interleaved A/D conversion system
Technical field
The present invention relates to technical field of signal sampling, particularly relate to a kind of error estimation for time-interleaved A/D conversion system and device.
Background technology
Along with the development of integrated circuit technique, the popularization of digitizing technique, the sampling rate of modulus switching device ADC and the requirement of sampling precision is more and more higher, and not requiring nothing more than data collecting system has high sample rate, also to have high sampling precision.In actual utilization, real-time sampling speed and sampling precision there is high dependency.But the maximum sampling rate of ADC is limited to its resolution, being conflict body between resolution and sampling rate, high sampling rate requires the shorter conversion time, and high-resolution then requires the longer conversion time.Design (IntegratedCircuitDesign, IC design) technique according to current IC, sampling rate more at a high speed will be realized, it would be desirable to explore a kind of ADC based on new construction and new method.A kind of important method realizing ultra-high speed sampling utilizes the time-interleaved A/D conversion system TIADC (Time-interleavedAnalog-to-DigitalConverter) of multichannel exactly.
As it is shown in figure 1, TIADC utilizes M sheet to have identical sample rate fsSingle ADC, adopt parallel structure, every ADC is with 1/ (M*f of being separated bys) interval sample, to reach sample rate for M*fs(total sampling rate f=M*fs) effect.In theory, the ADC structure that this time-interleaved for M passage is sampled enables to whole system sampling rate to M times of single DAC.But due to manufacturing process shortcoming inherently, it is impossible to make every a piece of ADC completely the same, so will necessarily make there is mismatch error between each passage ADC, thus seriously reducing the signal to noise ratio of whole ADC system.
Main mismatch error is time error (Time-skewerror), gain error (Gainerror) and biased error (offseterror).Biased error is equivalent to the error of an additivity, and it can cause and spurious spectra occurs on the frequency of single channel sample rate integral multiple on frequency spectrum.Gain error is equivalent to the error of a property taken advantage of, and the gain error between each passage can cause the amplitude modulation(PAM) of sampled signal.Time error is an only small amount relative to the sampling period, but it can cause that signal exists bigger spurious components, and it can cause an original signal spectrum spectrum offset on the integral multiple frequency of single channel sample rate on frequency spectrum.
It is usually utilization trimming front-end circuit based on early the mismatch correction of the time-interleaved ADC system of multichannel both at home and abroad, is reduced the impact of mismatch error by the circuit of meticulous layout.The shortcoming of this method is exactly when As time goes on, the change of temperature, the aging meeting of electric elements makes the correction effect of circuit lose efficacy.For the method overcoming this front end to revise, it is possible to use the method for back-end processing.
The correction algorithm of the mismatch error and digital back-end process thereof that are currently based on the time-interleaved ADC system of multichannel is the key of future development.Research institution both domestic and external is for time error, the correction algorithm of gain error and biased error has a lot, but traditional correction algorithm is all time error, the separately independent correction of gain error and biased error, even if correcting while multiple error can be realized, it is also required to the test signal of input multi-frequency, with the mode recording error of data base, or system need to be inputted the test signal of a fixed frequency, this kind of method not only needs extra memory space but also the test semaphore request of input is strict, so three kinds of errors are estimated by research one simultaneously simultaneously, the mismatch modification method for the time-interleaved ADC system of multichannel that input test signal restriction is few is of practical significance very much.
Summary of the invention
Based on this, it is necessary to for the problems referred to above, it is provided that a kind of error estimation for time-interleaved A/D conversion system and device, it is possible to estimate while realizing unbalance and test signal is not limited.
In order to achieve the above object, the technical scheme that the present invention takes is as follows:
A kind of error estimation for time-interleaved A/D conversion system, including step:
Input test signal in the time-interleaved A/D conversion system of multichannel, obtains the error signal of output；
Described error signal is carried out Fourier transformation, it is thus achieved that after Fourier transformation gain error occur the first frequency and biased error occur the second frequency；
Fourier transformation according to described error signal is in the value of the first frequency, it is thus achieved that the Fourier transformation of gain error；Fourier transformation according to described error signal is in the value of the second frequency, it is thus achieved that the Fourier transformation of biased error；
Fourier transformation according to gain error and the Fourier transformation of biased error, it is thus achieved that the gain error of each subchannel and biased error；
The new sampling function of each subchannel is obtained according to former sampling function, gain error and biased error；
The new sampling function of each subchannel is carried out Fourier transformation, obtains the first function；
Carry out Fourier transformation after being merged by the new sampling function of each subchannel, obtain the second function；
According to pre-conditioned, described first function and described second function are processed；
According to the first function after processing and the second function, it is thus achieved that the 3rd frequency of not mixing in first passage, and the passage corresponding with the 3rd frequency；
According to the passage that the 3rd frequency and the 3rd frequency are corresponding, it is thus achieved that the time error of each subchannel.
A kind of error estimation device for time-interleaved A/D conversion system, including:
Error signal acquisition module, for input test signal in the time-interleaved A/D conversion system of multichannel, obtains the error signal of output；
Frequency determines module, for described error signal is carried out Fourier transformation, it is thus achieved that the first frequency that after Fourier transformation, gain error occurs and the second frequency that biased error occurs；
Fourier transformation module, is used for the value at the first frequency of the Fourier transformation according to described error signal, it is thus achieved that the Fourier transformation of gain error；Fourier transformation according to described error signal is in the value of the second frequency, it is thus achieved that the Fourier transformation of biased error；
Gain and biased error determine module, for the Fourier transformation of the Fourier transformation according to gain error and biased error, it is thus achieved that the gain error of each subchannel and biased error；
Sampling function is new module more, for obtaining the new sampling function of each subchannel according to former sampling function, gain error and biased error；
First function determination module, for the new sampling function of each subchannel is carried out Fourier transformation, obtains the first function；
Second function determination module, for carrying out Fourier transformation after being merged by the new sampling function of each subchannel, obtains the second function；
Function processing module, for processing described first function and described second function according to pre-conditioned；
Frequency and passage determine module, for according to the first function after processing and the second function, it is thus achieved that the 3rd frequency of not mixing in first passage, and the passage corresponding with the 3rd frequency；
Time error determines module, for the passage corresponding according to the 3rd frequency and the 3rd frequency, it is thus achieved that the time error of each subchannel.
The present invention is used for error estimation and the device of time-interleaved A/D conversion system, error signal is obtained by input test signal in multichannel TIADC system, then error signal is carried out Fourier transformation and obtains its spectral magnitude data, deposit gain error and biased error to each subchannel in case at time error to estimate, then gain error and biased error are corrected, eliminate gain error and biased error, finally time error is estimated.The present invention can estimate three kinds of mismatch errors online simultaneously；When extraneous environment change, it is not necessary to adjust filter system or redesign hardware circuit, only need to re-enter test signal and can obtain three kinds of mismatch errors；The frequency of input signal is not limited；The three kinds of mismatch error precision estimated are high.
Accompanying drawing explanation
Fig. 1 is the structural representation of the time-interleaved A/D conversion system embodiment one of multichannel；
Fig. 2 is the present invention schematic flow sheet for the error estimation embodiment of time-interleaved A/D conversion system；
Fig. 3 is the structural representation of the time-interleaved A/D conversion system embodiment two of multichannel；
Fig. 4 is three kinds of mismatch error schematic diagrams in single passage；
Fig. 5 is the schematic diagram of input signal and the output signal (time domain) that there are three kinds of mismatch errors；
Fig. 6 is the schematic diagram of input signal and the output signal (frequency domain) that there are three kinds of mismatch errors；
Fig. 7 is the schematic diagram of the signal after correct for gain error of the present invention and biased error and error signal (time domain)；
Fig. 8 is that the present invention inputs the schematic diagram of signal (frequency domain) after signal and correct for gain error and the correction of biased error；
Fig. 9 is the schematic diagram of RMSE of the present invention (dB) and time error；
Figure 10 is the schematic diagram of RMSE of the present invention (dB) and gain error；
Figure 11 is the schematic diagram of RMSE of the present invention (dB) and biased error；
Figure 12 is the present invention structural representation for the error estimation device embodiment of time-interleaved A/D conversion system；
Figure 13 is the structural representation of function processing module embodiment of the present invention；
Figure 14 is the structural representation that frequency of the present invention and passage determine module embodiments；
Figure 15 is the structural representation that time error of the present invention determines module embodiments.
Detailed description of the invention
For further setting forth the effect of technological means that the present invention takes and acquirement, below in conjunction with accompanying drawing and preferred embodiment, to technical scheme, carry out clear and complete description.
As it is shown in figure 1, a kind of error estimation for time-interleaved A/D conversion system, including step:
S101, in the time-interleaved A/D conversion system of multichannel input test signal, obtain output error signal；
S102, described error signal is carried out Fourier transformation, it is thus achieved that the first frequency that after Fourier transformation, gain error occurs and the second frequency that biased error occurs；
S103, according to value at the first frequency of the Fourier transformation of described error signal, it is thus achieved that the Fourier transformation of gain error；Fourier transformation according to described error signal is in the value of the second frequency, it is thus achieved that the Fourier transformation of biased error；
S104, Fourier transformation according to the Fourier transformation of gain error and biased error, it is thus achieved that the gain error of each subchannel and biased error；
S105, obtain the new sampling function of each subchannel according to former sampling function, gain error and biased error；
S106, the new sampling function of each subchannel is carried out Fourier transformation, obtain the first function；
S107, the new sampling function of each subchannel is merged after carry out Fourier transformation, obtain the second function；
S108, according to pre-conditioned, described first function and described second function are processed；
S109, according to the first function after processing and the second function, it is thus achieved that the 3rd frequency of not mixing in first passage, and the passage corresponding with the 3rd frequency；
S110, according to passage corresponding to the 3rd frequency and the 3rd frequency, it is thus achieved that the time error of each subchannel.
In step S101, the test signal of input can be cosine signal Acos (2 π finT+ θ), wherein A is the amplitude of cosine signal, finFor frequency, θ is initial phase, and t is the time.The test signal of input can also be other signal, for instance sinusoidal signal etc..
When testing signal and being cosine signal, the error signal of output is:
e i ( t ) = [ g i A c o s ( 2 πf i n t + θ ) + a i ] × Σ n = - ∞ + ∞ δ ( t - ( n + i - 1 M ) T s + Δt i )
Wherein, eiT () is error signal, giRepresent the gain error of the i-th passage, aiRepresent the biased error of the i-th passage, Δ tiRepresenting the time error of the i-th passage, M represents total port number, and i=(0,1.....M-1), δ () represents Dirac delta function, Ts=1/fs, fsRepresenting each passage and the sample frequency of monolithic ADC, n represents scope, for the expression way of digital signal in analog digital conversion.
It should be noted that above-mentioned error signal is just to carrying out example to the situation that test signal is cosine signal, concrete error signal not being made restriction, when testing signal and being other signal, error signal makes corresponding change.The signal of follow-up formula be test signal be cosine signal basis on obtain, concrete form is not limited, after no longer illustrate one by one.
In step s 102, to error signal eiT () carries out Fourier transformation, for instance carry out fast Fourier transform FFT.It should be noted that FFT has only been illustrated by follow-up formula, but the concrete form of Fourier transformation is not limited, follow-up no longer illustrates one by one.
The Fourier transformation of error signal is chosen gain error and the frequency of biased error appearance:
Gain error giFrequency domain occurs in the first frequency f k 1 = ( k - 1 ) × f s ± f i n , k = ( 1..... M ) . Biased error aiFrequency domain occurs in the second frequency
In step s 103, error function e is utilizediT the Fourier transformation of () is at the first frequency With the second frequencyValue, respectively obtain:
Gain error giFourier transformation be:
Y g [ f k 1 ] = Σ k = 1 M g k [ A c o s ( 2 πf i n t + θ ) ] × Σ n = - ∞ + ∞ δ ( t - ( n + i - 1 M ) T s + Δt i ) × e - i ( 2 π / M ) j ( k - 1 ) , i = ( 0 , 1..... M - 1 ) , J is imaginary unit；
Biased error aiFourier transformation be:
Y a [ f k 2 ] = Σ k = 1 M a k e - i ( 2 π / M ) j ( k - 1 ) , i = ( 0 , 1..... M - 1 ) .
In step S104, estimate gain error and the biased error of each subchannel according to below equation:
Gain error [ g ] = 2 T s / Ae j θ × I F F T ( Y g [ f k 1 ] ) ;
Biased error [ a ] = T s × I F F T ( Y a [ f k 2 ] ) .
Wherein, IFFT is the inverse transformation of FFT.
In step S105, obtain the new sampling function of each subchannel according to following formula:
y ‾ A D C i = y A D C i - a i g i
Wherein, i=(0.....M-1), aiFor biased error, giFor gain error,For the new sampling function of each subchannel, yADCiFor former sampling function.
Original sampling function is updated to new sampling function, it is possible to eliminate gain error and the biased error of each subchannel.
In step s 106, to new sampling functionCarry out Fourier transformation, for instance fast Fourier transform, then the first function Y i = F F T ( y ‾ A D C i ) , i = ( 0..... M - 1 ) .
In step s 107, by new sampling functionIt is merged into functionAgain to functionCarry out fast Fourier transform, then the second function
In step S108, in one embodiment, may include that according to the pre-conditioned step that described first function and described second function are processed
By the first function YiIn less than ε × MAX (Yi) value be set to 0, its residual value is constant, and wherein ε is preset constant, for instance ε could be arranged to 0.06 etc., YiBeing the function of each passage in the first function, i=(0.....M-1), M is total port number, and MAX is maximum；
Value less than ε × MAX (Y) in second function Y is set to 0, and its residual value is constant.
In step S109, in one embodiment, described test signal is cosine signal, according to the first function after processing and the second function, it is thus achieved that the 3rd frequency of not mixing in first passage, and the step of the passage corresponding with the 3rd frequency may include that
Obtain Y0(0, π) makes amplitude | Y0(ejω) | the frequency ω being not zerok1, wherein Y0For the function of first passage in the first function after processing, ω is angular frequency, and e is natural Exponents, and j is imaginary unit；
From k=1 ..., M chooses and makesThe k being not zero, using the k that chooses as kp1, wherein M is total port number, and Y is the second function after processing；
Obtain Y0(π, 2 π) make amplitude | Y0(ejω) | the frequency ω being not zerok2；
From k=1 ..., M chooses and makesThe k being not zero, using the k that chooses as kp2。
Find Y in passage 00(i.e. YiMiddle i=0) in do not have mixing lack sampling frequency.Due to Y0For the Fourier transformation of scattered time, the cycle is 2 π, so being divided into (0, π) and (π, 2 π) two parts to find Y0In do not have mixing lack sampling frequency.
Search Y0(0, π) makes amplitude | Y0(ejω) | will not be the frequency ω of zerok1.Owing to input function is cosine signal, have and only one of which ωkCorresponding to making the magnitude function of Y
Search Y0The lack sampling frequency of the non-mixing in (π, 2 π), wherein finding method is in like manner, and now the magnitude function of Y becomesIn like manner find lack sampling frequency ωk2With corresponding kp2。
In step s 110, in one embodiment, according to the passage that the 3rd frequency and the 3rd frequency are corresponding, it is thus achieved that the step of the time error of each subchannel may include that
According to ωk1And kp1, and expression formula:Obtain Δ t1, wherein i=(0.....M-1), YiFor the function of each passage in the first function after processing, ω<2π>=((ω+π) mod2 π)-π, mod are complementation computing, and T=1/f, f is total sampling rate, i.e. f=fs× M；
According to ωk2And kp2, and expression formula:Obtain Δ t2；
According to Δ t1With Δ t2Meansigma methods obtain the time error of each subchannel.
Above-mentioned Δ t1With Δ t2Calculating process be: by ωk1Replace the ω in expression formulak, kp1Replace the K in expression formulap, obtain Δ t1；By ωk2Replace the ω in expression formulak, kp2Replace the K in expression formulap, obtain Δ t1。
In order to be better understood from embodiments of the present invention, the specific embodiment below in conjunction with the TIADC of a M=4 passage is described in detail.
As shown in Figure 3, the time-interleaved system schematic of multichannel TIADC for the present invention, the TIADC of input signal input M passage, every high-rate input signals is sampled by passage with identical sample rate but different sampling instants (adjacency channel difference T moment), finally merge output signal, realize the analog-to-digital conversion of high-speed sampling with this.
As shown in Figure 4, for the mismatch error situation in each subchannel.Owing in reality, the reason such as wiring can produce time mismatch error delta ti, the mismatch biased error a of additivity can be produced simultaneouslyiWith the property taken advantage of mismatch gain error gi。
Input test signal is cosine function Acos (2 π finT+ θ), A=1 is set,F is total sampling rate, and m is positive integer, and nfft is the positive integer power of 2, and nfft is exactly total hits.Here m=10027, nfft=2 are set16, f=16 × 109.The biased error a=[15,20,25,30] of every sub-channels, gain error g=[6,5,6,5], time error &Delta; t T = &lsqb; 0 , 0.03 , 0.01 , - 0.02 &rsqb; .
When not taking the present invention program, simulation result as it is shown in figure 5, when input cosine signal time, containing three kinds of mismatch errors (biasing mismatch error in 4 passages, gain mismatch errors, time mismatch error), output signal has had a strong impact on the waveform of original signal.By Fig. 6 it can be clearly seen that on frequency domain the frequency spectrum of original signal occur in that skew and occur in that a lot of spurious spectra.
Main idea is that: first estimate gain error and biased error, then again to signal correction, eliminate gain error and biased error, finally estimate time error.
First to error signal e i ( t ) = &lsqb; g i c o s ( 2 &pi;f i n t ) + a i &rsqb; &times; &Sigma; n = - &infin; + &infin; &delta; ( t - ( n + i - 1 M ) T s + &Delta;t i ) Carry out the FFT fast Fourier transform of 65536 points, fixing frequency obtainsWithBy formula &lsqb; g &rsqb; = 2 T s &times; I F F T ( Y g &lsqb; f k 1 &rsqb; ) , &lsqb; a &rsqb; = T s &times; I F F T ( Y a &lsqb; f k 2 &rsqb; ) Just it is estimated that the biased error a of every sub-channelsk=[15.00,20.00,25.00,30.00] and gain error gk=[6.00,5.00,6.00,5.00].Then every sub-channels is through formulaThe gain error in every sub-channels and biased error can be eliminated.
As it is shown in fig. 7, be correct for gain error with biased error after signal expression in time domain and original signal and correction after the expression in time domain of the difference (i.e. error signal) of signal.It can be seen that error signal has controlled within ± 2%.
As shown in Figure 8, the signal expression on frequency domain after the correction with biased error for input signal expression on frequency domain and correct for gain error.Relative to the original error existed on three kinds of mismatch error signal spectrums simultaneously, it can be seen that eliminate most spurious spectra and frequency line is moved, there remains the error owing to time mismatch error produces.
Then more every sub-channels is carried out the FFT fast Fourier transform of 16384 points, obtains Yi, i=(0,1,2,3), then the time-domain signal of 4 passages is merged into a signal, then the FFT fast Fourier carrying out 65536 points changes, and obtains Y, then to YiWith Y pretreatment, wherein ε could be arranged to 0.06.Find the frequency ω of non-mixingk1With corresponding Kp1, ωk2With corresponding Kp2.According to corresponding ωkAnd Kp, it is possible to obtain Δ t1, Δ t2.Seek Δ t again1, Δ t2Average can be obtained by every sub-channels time error estimate.Finally emulate the time error obtained to be estimated asThe time error obtained is estimated to be eliminated by follow-up method, to reach to eliminate the purpose of three kinds of mismatch errors simultaneously.The present invention is not related to the elimination bearing calibration that time error is follow-up.
The effect that time error is estimated can be estimated, it is contemplated that RMSE (root-mean-squareerror), as standard, estimates the error effects between actual time error and estimation difference.Consistent with instance data before, time error increases with the step-length of each subchannel 0.01, observes the change of its RMSE.Can it is seen from figure 9 that, overall error RMSE is lower than-114dB.And the increase of the precision estimated mismatch error over time and increase, it was demonstrated that effectiveness of the invention.
In like manner the effect of the gain error estimated with biased error can also be estimated.When assessing the estimation effect of gain error, the biased error a=[15,20,25,30] of every sub-channels, time error Δ t/T=[0,0.03,0.01 ,-0.02], gain error g=[6,5,6,5], the gain error of each subchannel increases with step-length for 5.As shown in Figure 10, along with the increase of gain error, its RMSE also increases its result, but is generally lower than-250dB.When assessing the estimation effect of biased error, the gain error g=[6,5,6,5] of every sub-channels, time error Δ t/T=[0,0.03,0.01 ,-0.02], biased error a=[15,20,25,30], the biased error of each subchannel increases with step-length for 5.As shown in figure 11, overall RMSE is lower than-276dB for its result.
Can be seen that the present invention is to the effectiveness of the method for estimation of the gain error of the time-interleaved A/D conversion system of multichannel, biased error and time error and high accuracy from above experimental result.
Based on same inventive concept, the present invention also provides for a kind of error estimation device for time-interleaved A/D conversion system, below the detailed description of the invention of apparatus of the present invention is described in detail.
As shown in figure 12, a kind of error estimation device for time-interleaved A/D conversion system, including:
Error signal acquisition module 101, for input test signal in the time-interleaved A/D conversion system of multichannel, obtains the error signal of output；
Frequency determines module 102, for described error signal is carried out Fourier transformation, it is thus achieved that the first frequency that after Fourier transformation, gain error occurs and the second frequency that biased error occurs；
Fourier transformation module 103, is used for the value at the first frequency of the Fourier transformation according to described error signal, it is thus achieved that the Fourier transformation of gain error；Fourier transformation according to described error signal is in the value of the second frequency, it is thus achieved that the Fourier transformation of biased error；
Gain and biased error determine module 104, for the Fourier transformation of the Fourier transformation according to gain error and biased error, it is thus achieved that the gain error of each subchannel and biased error；
Sampling function is new module 105 more, for obtaining the new sampling function of each subchannel according to former sampling function, gain error and biased error；
First function determination module 106, for the new sampling function of each subchannel is carried out Fourier transformation, obtains the first function；
Second function determination module 107, for carrying out Fourier transformation after being merged by the new sampling function of each subchannel, obtains the second function；
Function processing module 108, for processing described first function and described second function according to pre-conditioned；
Frequency and passage determine module 109, for according to the first function after processing and the second function, it is thus achieved that the 3rd frequency of not mixing in first passage, and the passage corresponding with the 3rd frequency；
Time error determines module 110, for the passage corresponding according to the 3rd frequency and the 3rd frequency, it is thus achieved that the time error of each subchannel.
In one embodiment, described sampling function more new module 105 can obtain the new sampling function of each subchannel according to following formula:
y &OverBar; A D C i = y A D C i - a i g i
Wherein, i=(0.....M-1), aiFor biased error, giFor gain error,For the new sampling function of each subchannel, yADCiFor former sampling function.
In one embodiment, as shown in figure 13, described function processing module 108 includes:
First function processing unit 1081, for by the first function less than ε × MAX (Yi) value be set to 0, wherein ε is preset constant, YiBeing the function of each passage in the first function, i=(0.....M-1), M is total port number, and MAX is maximum；
Second function processing unit 1082, for value less than ε × MAX (Y) in the second function is set to 0, wherein Y is the second function.
In one embodiment, described test signal is cosine signal, and as shown in figure 14, described frequency and passage determine that module 109 may include that
First acquiring unit 1091, is used for obtaining Y0(0, π) makes amplitude | Y0(ejω) | the frequency ω being not zerok1, wherein Y0For the function of first passage in the first function after processing, ω is angular frequency, and e is natural Exponents, and j is imaginary unit；
First passage determines unit 1092, is used for from k=1 ..., M chooses and makesThe k being not zero, using the k that chooses as kp1, wherein M is total port number, and Y is the second function after processing；
Second acquisition unit 1093, is used for obtaining Y0(π, 2 π) make amplitude | Y0(ejω) | the frequency ω being not zerok2；
Unit 1094 determined by second channel, is used for from k=1 ..., M chooses and makesThe k being not zero, using the k that chooses as kp2。
In one embodiment, as shown in figure 15, described time error determines that module 110 may include that
Very first time error determines unit 1101, for according to ωk1And kp1, and expression formula: Obtain Δ t1, wherein i=(0.....M-1), YiFor the function of each passage in the first function after processing,ω<2π>=((ω+π) mod2 π)-π, mod are complementation computing, and T=1/f, f is total sampling rate；
Second time error determines unit 1102, for according to ωk2And kp2, and expression formula: &Delta; t = ( l n Y i ( e j&omega; k ) Y 0 ( e j&omega; k ) j ( &omega; k M + 2 ( K p - 1 ) &pi; M ) < 2 &pi; > - i ) T , Obtain Δ t2；
3rd time error determines unit 1103, for according to Δ t1With Δ t2Meansigma methods obtain the time error of each subchannel.
Other technical characteristic of apparatus of the present invention is identical with the inventive method, does not repeat them here.
Each technical characteristic of embodiment described above can combine arbitrarily, for making description succinct, the all possible combination of each technical characteristic in above-described embodiment is not all described, but, as long as the combination of these technical characteristics is absent from contradiction, all it is considered to be the scope that this specification is recorded.
Embodiment described above only have expressed the several embodiments of the present invention, and it describes comparatively concrete and detailed, but can not therefore be construed as limiting the scope of the patent.It should be pointed out that, for the person of ordinary skill of the art, without departing from the inventive concept of the premise, it is also possible to making some deformation and improvement, these broadly fall into protection scope of the present invention.Therefore, the protection domain of patent of the present invention should be as the criterion with claims.

Claims (10)

1. the error estimation for time-interleaved A/D conversion system, it is characterised in that include step:
Input test signal in the time-interleaved A/D conversion system of multichannel, obtains the error signal of output；
Described error signal is carried out Fourier transformation, it is thus achieved that after Fourier transformation gain error occur the first frequency and biased error occur the second frequency；
Fourier transformation according to described error signal is in the value of the first frequency, it is thus achieved that the Fourier transformation of gain error；Fourier transformation according to described error signal is in the value of the second frequency, it is thus achieved that the Fourier transformation of biased error；
Fourier transformation according to gain error and the Fourier transformation of biased error, it is thus achieved that the gain error of each subchannel and biased error；
The new sampling function of each subchannel is obtained according to former sampling function, gain error and biased error；
The new sampling function of each subchannel is carried out Fourier transformation, obtains the first function；
Carry out Fourier transformation after being merged by the new sampling function of each subchannel, obtain the second function；
According to pre-conditioned, described first function and described second function are processed；
According to the first function after processing and the second function, it is thus achieved that the 3rd frequency of not mixing in first passage, and the passage corresponding with the 3rd frequency；
According to the passage that the 3rd frequency and the 3rd frequency are corresponding, it is thus achieved that the time error of each subchannel.
2. the error estimation for time-interleaved A/D conversion system according to claim 1, it is characterised in that include according to the pre-conditioned step that described first function and described second function are processed:
By in the first function less than ε × MAX (Yi) value be set to 0, wherein ε is preset constant, YiBeing the function of each passage in the first function, i=(0 ... ..M-1), M is total port number, and MAX is maximum；
Value less than ε × MAX (Y) in second function is set to 0, and wherein Y is the second function.
3. the error estimation for time-interleaved A/D conversion system according to claim 1, it is characterized in that, described test signal is cosine signal, according to the first function after processing and the second function, obtain the 3rd frequency of not mixing in first passage, and the step of the passage corresponding with the 3rd frequency include:
Obtain Y0(0, π) makes amplitude | Y0(ejω) | the frequency ω being not zerok1, wherein Y0For the function of first passage in the first function after processing, ω is angular frequency, and e is natural Exponents, and j is imaginary unit；
From k=1 ..., M chooses and makesThe k being not zero, using the k that chooses as kp1, wherein M is total port number, and Y is the second function after processing；
Obtain Y0(π, 2 π) make amplitude | Y0(ejω) | the frequency ω being not zerok2；
From k=1 ..., M chooses and makesThe k being not zero, using the k that chooses as kp2。
4. the error estimation for time-interleaved A/D conversion system according to claim 3, it is characterised in that according to the passage that the 3rd frequency and the 3rd frequency are corresponding, it is thus achieved that the step of the time error of each subchannel includes:
According to ωk1And kp1, and expression formula:Obtain Δ t1, wherein i=(0 ... ..M-1), YiFor the function of each passage in the first function after processing, ω< 2 π >=((ω+π) mod2 π)-π, mod are complementation computing, and Y=1/f, f is total sampling rate；
According to ωk2And kp2, and expression formula:Obtain Δ t2；
According to Δ t1With Δ t2Meansigma methods obtain the time error of each subchannel.
5. the error estimation for time-interleaved A/D conversion system according to Claims 1-4 any one, it is characterised in that according to expression formulaObtain the new sampling function of each subchannel, wherein i=(0 ... ..M-1), aiFor biased error, giFor gain error,For the new sampling function of each subchannel, yADCiFor former sampling function.
6. the error estimation device for time-interleaved A/D conversion system, it is characterised in that including:
Error signal acquisition module, for input test signal in the time-interleaved A/D conversion system of multichannel, obtains the error signal of output；
Frequency determines module, for described error signal is carried out Fourier transformation, it is thus achieved that the first frequency that after Fourier transformation, gain error occurs and the second frequency that biased error occurs；
Fourier transformation module, is used for the value at the first frequency of the Fourier transformation according to described error signal, it is thus achieved that the Fourier transformation of gain error；Fourier transformation according to described error signal is in the value of the second frequency, it is thus achieved that the Fourier transformation of biased error；
Gain and biased error determine module, for the Fourier transformation of the Fourier transformation according to gain error and biased error, it is thus achieved that the gain error of each subchannel and biased error；
Sampling function is new module more, for obtaining the new sampling function of each subchannel according to former sampling function, gain error and biased error；
First function determination module, for the new sampling function of each subchannel is carried out Fourier transformation, obtains the first function；
Second function determination module, for carrying out Fourier transformation after being merged by the new sampling function of each subchannel, obtains the second function；
Function processing module, for processing described first function and described second function according to pre-conditioned；
Frequency and passage determine module, for according to the first function after processing and the second function, it is thus achieved that the 3rd frequency of not mixing in first passage, and the passage corresponding with the 3rd frequency；
Time error determines module, for the passage corresponding according to the 3rd frequency and the 3rd frequency, it is thus achieved that the time error of each subchannel.
7. the error estimation device for time-interleaved A/D conversion system according to claim 6, it is characterised in that described function processing module includes:
First function processing unit, for by the first function less than ε × MAX (Yi) value be set to 0, wherein ε is preset constant, YiBeing the function of each passage in the first function, i=(0 ... ..M-1), M is total port number, and MAX is maximum；
Second function processing unit, for value less than ε × MAX (Y) in the second function is set to 0, wherein Y is the second function.
8. the error estimation device for time-interleaved A/D conversion system according to claim 6, it is characterised in that described test signal is cosine signal, and described frequency and passage determine that module includes:
First acquiring unit, is used for obtaining Y0(0, π) makes amplitude | Y0(ejω) | the frequency ω being not zerok1, wherein Y0For the function of first passage in the first function after processing, ω is angular frequency, and e is natural Exponents, and j is imaginary unit；
First passage determines unit, is used for from k=1 ..., M chooses and makesThe k being not zero, using the k that chooses as kp1, wherein M is total port number, and Y is the second function after processing；
Second acquisition unit, is used for obtaining Y0(π, 2 π) make amplitude | Y0(ejω) | the frequency ω being not zerok2；
Unit determined by second channel, is used for from k=1 ..., M chooses and makesThe k being not zero, using the k that chooses as kp2。
9. the error estimation device for time-interleaved A/D conversion system according to claim 8, it is characterised in that described time error determines that module includes:
Very first time error determines unit, for according to ωk1And kp1, and expression formula: Obtain Δ t1, wherein i=(0 ... ..M-1), YiFor the function of each passage in the first function after processing,ω< 2 π >=((ω+π) mod2 π)-π, mod are complementation computing, and T=1/f, f is total sampling rate；
Second time error determines unit, for according to ωk2And kp2, and expression formula: &Delta; t = ( l n Y i ( e j &omega; k ) Y 0 ( e j &omega; k ) j ( &omega; k M + 2 ( K p - 1 ) &pi; M ) < 2 &pi; > - i ) T , Obtain Δ t2；
3rd time error determines unit, for according to Δ t1With Δ t2Meansigma methods obtain the time error of each subchannel.
10. the error estimation device for time-interleaved A/D conversion system according to claim 6 to 9 any one, it is characterised in that described sampling function more new module is according to expression formulaObtain the new sampling function of each subchannel, wherein i=(0 ... ..M-1), aiFor biased error, giFor gain error,For the new sampling function of each subchannel, yADCiFor former sampling function.
CN201610044047.7A 2016-01-22 2016-01-22 Error estimation method and device for time interleaving analog-digital conversion system Pending CN105720983A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201610044047.7A CN105720983A (en) 2016-01-22 2016-01-22 Error estimation method and device for time interleaving analog-digital conversion system

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201610044047.7A CN105720983A (en) 2016-01-22 2016-01-22 Error estimation method and device for time interleaving analog-digital conversion system

Publications (1)

Publication Number Publication Date
CN105720983A true CN105720983A (en) 2016-06-29

Family

ID=56153720

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201610044047.7A Pending CN105720983A (en) 2016-01-22 2016-01-22 Error estimation method and device for time interleaving analog-digital conversion system

Country Status (1)

Country Link
CN (1) CN105720983A (en)

Cited By (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105846822A (en) * 2016-03-24 2016-08-10 广东顺德中山大学卡内基梅隆大学国际联合研究院 Method and system for measuring mismatch error of time-interleaved analog-to-digital conversion system
CN106209103A (en) * 2016-07-29 2016-12-07 电子科技大学 TIADC gain based on spectrum analysis and the bearing calibration of time error
CN106341132A (en) * 2016-08-08 2017-01-18 中国工程物理研究院电子工程研究所 Error blind correction method for time interleaved sampling ADC (Analog-to-Digital Converter)
CN106375252A (en) * 2016-09-05 2017-02-01 广东顺德中山大学卡内基梅隆大学国际联合研究院 Method and system for correcting IQ and TIADC joint frequency distortion
CN106374920A (en) * 2016-09-05 2017-02-01 中山大学 Estimation and compensation method of TIADC system based on polynomial model
CN107046425A (en) * 2017-04-18 2017-08-15 中国电子科技集团公司第二十四研究所 The sampling time error detecting system counted based on sampling interval difference
CN112636754A (en) * 2020-11-06 2021-04-09 中山大学 Channel mismatch estimation method, system, device and medium for broadband cross sampling system
CN113595553A (en) * 2021-07-22 2021-11-02 华中科技大学 Data acquisition system and method based on time-interleaved parallel sampling
CN114267407A (en) * 2022-03-03 2022-04-01 合肥悦芯半导体科技有限公司 Precision correction method, device and system and precision correction equipment

Citations (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20110063149A1 (en) * 2008-12-29 2011-03-17 Intersil Americas Inc. Calibration of offset, gain and phase errors in m-channel time-interleaved analog-to-digital converters
CN102075464A (en) * 2011-01-18 2011-05-25 电子科技大学 Joint estimation and real-time correction method for channel error of TIADC system
CN102761335A (en) * 2011-04-29 2012-10-31 英特希尔美国公司 Multiplier-free algorithms for sample-time and gain mismatch error estimation in a two-channel time-interleaved analog-to-digital converter
CN103812506A (en) * 2012-11-13 2014-05-21 邹月娴 TIADC (Time-Interleaved Analog-to-Digital Converter) time mismatch parameter blind measuring method based on signal frequency domain sparsity
US20140266823A1 (en) * 2013-03-15 2014-09-18 Integrated Device Technology, Inc. Systems and methods for estimation of offset and gain errors in a time-interleaved analog-to-digital converter
US20150070198A1 (en) * 2013-09-06 2015-03-12 Broadcom Corporation Flexible ADC Calibration Technique Using ADC Capture Memory
US9088293B1 (en) * 2014-09-04 2015-07-21 Semtech Corporation Calibration of a time-interleaved analog-to-digital converter (ADC) circuit
US9143147B1 (en) * 2014-07-03 2015-09-22 Keysight Technologies, Inc. Calibration of inter-slice gain and offset errors in time-interleaved analog-to- digital converter
US20150303934A1 (en) * 2014-04-17 2015-10-22 The Board Of Regents, The University Of Texas System System and method for dynamic path-mismatch equalization in time-interleaved adc
CN105024696A (en) * 2015-07-02 2015-11-04 大唐微电子技术有限公司 Sampling time error calibrating device and method of multi-channel parallel analog-to-digital conversion system

Patent Citations (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20110063149A1 (en) * 2008-12-29 2011-03-17 Intersil Americas Inc. Calibration of offset, gain and phase errors in m-channel time-interleaved analog-to-digital converters
CN102075464A (en) * 2011-01-18 2011-05-25 电子科技大学 Joint estimation and real-time correction method for channel error of TIADC system
CN102761335A (en) * 2011-04-29 2012-10-31 英特希尔美国公司 Multiplier-free algorithms for sample-time and gain mismatch error estimation in a two-channel time-interleaved analog-to-digital converter
CN103812506A (en) * 2012-11-13 2014-05-21 邹月娴 TIADC (Time-Interleaved Analog-to-Digital Converter) time mismatch parameter blind measuring method based on signal frequency domain sparsity
US20140266823A1 (en) * 2013-03-15 2014-09-18 Integrated Device Technology, Inc. Systems and methods for estimation of offset and gain errors in a time-interleaved analog-to-digital converter
US20150070198A1 (en) * 2013-09-06 2015-03-12 Broadcom Corporation Flexible ADC Calibration Technique Using ADC Capture Memory
US20150303934A1 (en) * 2014-04-17 2015-10-22 The Board Of Regents, The University Of Texas System System and method for dynamic path-mismatch equalization in time-interleaved adc
US9143147B1 (en) * 2014-07-03 2015-09-22 Keysight Technologies, Inc. Calibration of inter-slice gain and offset errors in time-interleaved analog-to- digital converter
US9088293B1 (en) * 2014-09-04 2015-07-21 Semtech Corporation Calibration of a time-interleaved analog-to-digital converter (ADC) circuit
CN105024696A (en) * 2015-07-02 2015-11-04 大唐微电子技术有限公司 Sampling time error calibrating device and method of multi-channel parallel analog-to-digital conversion system

Cited By (15)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105846822A (en) * 2016-03-24 2016-08-10 广东顺德中山大学卡内基梅隆大学国际联合研究院 Method and system for measuring mismatch error of time-interleaved analog-to-digital conversion system
CN106209103A (en) * 2016-07-29 2016-12-07 电子科技大学 TIADC gain based on spectrum analysis and the bearing calibration of time error
CN106209103B (en) * 2016-07-29 2019-09-24 电子科技大学 The bearing calibration of TIADC gain and time error based on spectrum analysis
CN106341132A (en) * 2016-08-08 2017-01-18 中国工程物理研究院电子工程研究所 Error blind correction method for time interleaved sampling ADC (Analog-to-Digital Converter)
CN106341132B (en) * 2016-08-08 2019-05-24 中国工程物理研究院电子工程研究所 The error blind correction method of time-interleaved sampling ADC
CN106375252B (en) * 2016-09-05 2019-05-03 广东顺德中山大学卡内基梅隆大学国际联合研究院 A kind of modification method and system of IQ and TIADC frequency Combine distortion
CN106374920A (en) * 2016-09-05 2017-02-01 中山大学 Estimation and compensation method of TIADC system based on polynomial model
CN106375252A (en) * 2016-09-05 2017-02-01 广东顺德中山大学卡内基梅隆大学国际联合研究院 Method and system for correcting IQ and TIADC joint frequency distortion
CN107046425A (en) * 2017-04-18 2017-08-15 中国电子科技集团公司第二十四研究所 The sampling time error detecting system counted based on sampling interval difference
CN107046425B (en) * 2017-04-18 2020-09-01 中国电子科技集团公司第二十四研究所 Sampling interval difference value statistics-based sampling time error detection system
CN112636754A (en) * 2020-11-06 2021-04-09 中山大学 Channel mismatch estimation method, system, device and medium for broadband cross sampling system
CN113595553A (en) * 2021-07-22 2021-11-02 华中科技大学 Data acquisition system and method based on time-interleaved parallel sampling
CN113595553B (en) * 2021-07-22 2023-10-20 华中科技大学 Data acquisition system and method based on time interleaving parallel sampling
CN114267407A (en) * 2022-03-03 2022-04-01 合肥悦芯半导体科技有限公司 Precision correction method, device and system and precision correction equipment
CN114267407B (en) * 2022-03-03 2022-06-10 合肥悦芯半导体科技有限公司 Precision correction method, device and system and precision correction equipment

Similar Documents

Publication Publication Date Title
CN105720983A (en) Error estimation method and device for time interleaving analog-digital conversion system
CN107102255B (en) Single ADC acquisition channel dynamic characteristic test method
Kang et al. Phase difference correction method for phase and frequency in spectral analysis
CN104007316B (en) A kind of High Precision Frequency method under lack sampling speed and measuring instrument thereof
CN103257271A (en) Device and method for detecting micro grid harmonic wave and inter-harmonics based on STM32F107VCT6
CN106483374A (en) A kind of harmonic wave harmonic detection method based on Nuttall double window whole phase FFT
CN101674087A (en) Method for obtaining channel mismatching error of time alternative ADC system
CN203287435U (en) A micro electrical network harmonic wave and inter-harmonic wave test apparatus based on an STM32F107VCT6
CN103647554A (en) Method for measuring bias and gain errors of time alternative ADC acquisition system
CN105915221A (en) TIADC system and method
CN105259410B (en) The frequency estimating methods and its device of a kind of lack sampling waveform under very noisy interference
CN109274372B (en) Inter-channel sampling time mismatch error extraction method for TIADC system
Li et al. A background correlation-based timing skew estimation method for time-interleaved ADCs
Liu et al. Adaptive semiblind background calibration of timing mismatches in a two-channel time-interleaved analog-to-digital converter
CN106341132A (en) Error blind correction method for time interleaved sampling ADC (Analog-to-Digital Converter)
Gao et al. An adaptive calibration technique of timing skew mismatch in time-interleaved analog-to-digital converters
Pan et al. Frequency response mismatch calibration in generalized time-interleaved systems
CN103812506A (en) TIADC (Time-Interleaved Analog-to-Digital Converter) time mismatch parameter blind measuring method based on signal frequency domain sparsity
Kuojun et al. A TIADC mismatch calibration method for digital storage oscilloscope
CN103746699B (en) Signal reconstruction method based on rotation matrix error estimation for alternative sampling system
CN110285881A (en) A kind of intensive spectral frequency estimation technique based on full phase filtering
US10498348B1 (en) Time error and gain offset estimation in interleaved analog-to-digital converters
CN105846822A (en) Method and system for measuring mismatch error of time-interleaved analog-to-digital conversion system
Asami et al. Technique to improve the performance of time-interleaved AD converters with mismatches of non-linearity
Song et al. Digital estimation and compensation of analog errors in frequency-interleaved ADCs

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination