KR101659402B1 - Multi-channel adc and channel distortion compensation method of the same - Google Patents

Abstract

A method of measuring a channel response of an analog-to-digital converter of a time interleaving scheme according to the present invention is a method of measuring a channel response of an analog-to-digital converter by applying a first signal of a first frequency and a second signal of a reference frequency to the analog- Measuring a second response by applying a third signal of a second frequency and the second signal to the analog-to-digital converter and measuring a second response, measuring a channel sequence of the analog-to-digital converter And determining response characteristics of each of the plurality of analog-to-digital converters interleaved with reference to the estimated channel sequence, the first and second responses.
Separating the sampled original measurement value into each of the plurality of channels with reference to the measured channel pressure characteristic, generating a separation signal Ymeas by filling a vacated time slot with '0' in each channel , Generating a vector rotation amount (E) in a frequency domain for a separation signal corresponding to each of the channels, and referring to the response characteristic (H) and the vector rotation amount (E) for each of the channels And calculating a restoration signal (Yreconst) of each of the channels.

Description

TECHNICAL FIELD [0001] The present invention relates to a multi-channel analog-to-digital converter, and more particularly,

The present invention relates to an electronic device, and more particularly, to a method for measuring a channel response characteristic of a multi-channel analog-to-digital converter that converts an analog signal to a digital signal and a method for compensating for channel distortion thereof.

With recent advances in technology and user demand, there is a growing demand for analog-to-digital converters (ADCs) with higher resolutions of 8 bits or more, even at high operating speeds of tens GS / s or more. While the sampling rate is increasing, the speed of the ADC is not keeping up. Therefore, a multi-channel TI (Time-Interleaving) ADC that connects pipelined ADCs in parallel has been proposed. TI ADCs can simultaneously provide high resolution and fast operation speed while reducing power consumption by paralleling multiple ADCs operating at relatively low speeds.

Typical TI ADCs connect low-speed sub-ADCs in parallel to improve overall ADC operation speed while optimizing power consumption without process constraints. However, there are problems in that the performance of the entire ADC is degraded due to mismatches such as offset mismatch, gain mismatch, phase, and sampling timing among multiple channels formed by each of the parallel-connected sub-ADCs. These mismatches have been studied in low frequency measurement signals and various solutions have been proposed. However, in a measurement signal of a high frequency band (for example, an RF band), it is difficult to solve the problem with these methods. In particular, the mismatch between the gain and phase of the channel is represented by a different value depending on the frequency. Therefore, there is an urgent need for a technique for measuring interchannel characteristics according to a change in frequency and compensating for effects due to mismatching between channels by referring to measured characteristics.

Therefore, the present invention is to provide a method capable of measuring frequency-dependent channel-to-channel mismatch, and an analog-to-digital converter capable of compensating for problems caused by such mismatching and a method for compensating for channel mismatching thereof.

According to an aspect of the present invention, there is provided a method of measuring a channel response of an analog-to-digital converter of a time interleaving method, the method comprising: averaging a first signal of a first frequency and a second signal of a reference frequency, Measuring a first response, applying a third signal of a second frequency and the second signal to the analog-to-digital converter and measuring a second response, measuring the second response based on the second signal, Estimating a channel sequence of the plurality of analog-to-digital converters, and determining a response characteristic of each of the plurality of analog-to-digital converters interleaved with reference to the estimated channel sequence, the first and second responses.

According to an aspect of the present invention, there is provided a method of compensating distortion of an analog-to-digital converter of a time interleaving method, the method comprising: measuring a response characteristic (H) of each of a plurality of channels of the interleaved analog- Separating the sampled original measurement values into each of the plurality of channels, generating a separation signal (Ymeas) by filling the empty time slot in each channel with '0' Generating a vector rotation amount E in a frequency domain with respect to a signal and a restoration signal Yreconst of each of the channels with reference to a response characteristic H and a vector rotation amount E for each of the channels, ). ≪ / RTI >

The TI ADC according to the embodiment of the present invention can easily estimate a sequence of channels to be selected at random. In the ADC of the present invention, the vector rotation amount of each of the channels can be used to compensate for distortion caused by different response characteristics between channels.

1 is a block diagram illustrating a channel response measurement system for measuring channel response characteristics of an analog-to-digital converter according to an embodiment of the present invention.
Figure 2 is a block diagram illustrating an exemplary structure of the TI-ADC of Figure 1;
3 is a block diagram illustrating an exemplary configuration of the ADC_1 of FIG.
FIG. 4 is a diagram illustrating a random channel sequence of the TI-ADC 130. FIG.
5 is a diagram illustrating a method of matching a channel sequence of the present invention.
6 is a flowchart briefly showing a method of measuring channel characteristics in a time interleaving (TI) ADC of the present invention.
7 is an exemplary diagram illustrating the response characteristics of each of the channels measured with reference to the channel selection sequence of the present invention.
8 is a block diagram illustrating an ADC according to another technical aspect of the present invention.
9 is a diagram illustrating a method of separating measured values of the present invention by channel.
10 shows the determinant of Equation 2 for all channels.
FIG. 11 is a flowchart briefly showing a signal distortion compensation method of the present invention.
12A and 12B are spectrum charts showing the effect of the present invention.

Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings, so that those skilled in the art can easily carry out the technical idea of the present invention. The same elements will be referred to using the same reference numerals. Similar components will be referred to using similar reference numerals. The analog-to-digital converter according to the present invention to be described below and the operation performed thereby are only described for example, and various changes and modifications are possible without departing from the technical idea of the present invention.

1 is a block diagram illustrating a channel response measurement system for measuring channel response characteristics of an analog-to-digital converter according to an embodiment of the present invention. Referring to FIG. 1, a channel response measurement system 100 may include a signal generator 110, a mixer 120, a TI-ADC 130, and a channel response measurement unit 140.

The signal generator 110 includes a first signal generator 112 and a second signal generator 114. The first signal generator 112 generates a first signal {s (t)} which is a continuous wave that can vary the frequency. The second signal generator 114 generates a second signal r (t), which is used as a reference signal for determining a subsequent random channel sequence. The frequency f of the first signal {s (t)} generated by the first signal generator 112 may be varied under the control of the channel response measuring unit 140. [ In addition, the amplitude or phase of the continuous signal {s (t)} generated in the first signal generator 112 under the control of the channel response measuring unit 140 may also be controlled. In addition, it is possible to control the amplitude, the reference frequency fref, and the phase of the second signal {r (t)} which is the reference signal generated by the second signal generator 114.

The mixer 120 combines the first signal {s (t)} provided by the signal generator 110 and the second signal {r (t)}. The mixer 120 generates an input signal {x (t)} including the first signal {s (t)} and the second signal {r (t)}.

The TI-ADC 130 samples the input signal {x (t)} in a time-interleaving manner and outputs the sampled signal as a digital signal Yn. The TI-ADC 130 will sample the input signal {x (t)} and distribute the sampled values to a plurality of ADCs that comprise each channel with different time slots. In each of the plurality of distributed channels, the data distributed through the relatively low-speed ADC is processed and recovered as a digital signal. The specific structure of the TI-ADC 130 will be described in detail in the following drawings.

The channel response measuring unit 140 may measure a response characteristic for each of the channels from the output signal Yn of the TI-ADC 130. [ The channel response measuring unit 140 can measure the offset, gain, and phase difference of each of the channels with reference to the output signal Yn of the TI-ADC 130. [ The channel response measuring unit 140 may calculate the mismatch magnitude of each of these channels and use the result to correct the non-linear mismatch distortion for the channels.

In an interleaving operation generally occurring in a high-speed analog-to-digital converter, the start channel of a plurality of ADCs is arbitrarily determined. Therefore, even if the offset, gain, and phase difference of each of the channels are measured several times, it is not easy to specify the channel corresponding to the measured value. This arbitrary start channel (or start ADC) variation has difficulties in determining the correct characteristics using measured values.

The channel response measuring unit 140 of the present invention uses a first signal {s (t) of variable frequency and a second signal {r (t) of fixed reference frequency to solve this problem. That is, the first signal {s (t)} which is a sinusoidal wave of a variable frequency and the second signal {r (t)} which is a sinusoidal wave having a fixed reference frequency are mixed and applied to the TI-ADC 130. The gain, phase, or offset can be measured at different frequencies while varying the frequency of the first signal {s (t)}. At this time, since only the frequency of the first signal {s (t)} has changed, the measurement value for the signal to which the first frequency f is applied and the measurement value for the signal to be provided at the second frequency fref, The gain, phase, offset, etc. corresponding to the second signal {r (t)} do not change greatly. Therefore, the start channel (or the starting ADC) can be estimated based on the response characteristic for the second signal r (t).

Here, the signal y _n , which is a mixture of the signals output from the signal generator 110 under the control of the channel response measuring unit 140, can be expressed by Equation (1) below.

Here, n corresponds to the number of channels. And A _n and φ _n represent the amplitude and phase of the first signal {s (t)} component for each of the channels. A _{ref, n} and? _{Ref, n} represent the amplitude and phase response characteristics of the second signal {r (t)} component for each of the channels. Also, offset _n represents the DC offset response characteristic of each of the channels.

When the first frequency f is varied to provide the measurement signal {x (t)}, substantially frequency dependent characteristics will appear for the first signal {s (t)}. However, the response components for the second signal {r (t)} provided at a fixed frequency will maintain a relatively constant value. Considering this characteristic, the channel sequences of the data measured by the variable frequency can be easily distinguished by referring to the response to the second signal {r (t)} which is the reference signal. After the restoration of the channel sequence, the characteristics of each of the channels can be easily calculated through the measured gain, phase, and offset.

FIG. 2 is a block diagram illustrating a specific configuration of the TI-ADC of FIG. 1; FIG. Referring to FIG. 2, the TI-ADC 130 may include a circulator 131, a plurality of ADCs 132 to 135, and a multiplexer 136. Here, the number of channels can be arbitrarily determined according to the sampling frequency and various requirements, but for convenience of description, a case of configuring four channels will be exemplarily described.

The circulator 131 selects the input sampling signal x (t) and delivers it to any one of the channels. That is, the circulator 131 will sequentially assign the input signal {x (t)} starting from any one of the plurality of channels CH1, CH2, CH3, and CH4. X (t)} in the order of the third channel CH3, the fourth channel CH4 and the first channel CH1 if the channel first selected by the circulator 131 is the second channel CH2, Will be delivered separately.

The plurality of ADCs 132 to 135 constitute respective channels and convert logical values for the level of the transmitted input signal {x (t)}, respectively. The plurality of ADCs 132 to 135 can be set to a relatively low frequency relative to the sampling frequency. That is, as the number of the plurality of ADCs 132 to 135 increases, the ADCs can be implemented as low frequency driven configurations.

The multiplexer 136 selects the outputs of the ADCs 132-135 corresponding to each of the channels. As a result, the multiplexer 136 will be synchronized to the time information of the circulator 131 to select the output of the plurality of ADCs 132-135. The logic values sequentially selected by the multiplexer 136 will be provided as the output value Yn of the TI-ADC 136. [

Figure 3 is a block diagram illustrating an exemplary structure of the ADC of Figure 2; Referring to FIG. 3, a first one of the plurality of ADCs 132-135 includes a preamplifier 137, sample / holder circuit 138, and And an ADC circuit 139.

The preamplifier 137 pre-amplifies the input analog signal x (t) before processing. When the level of the input analog signal x (t) is low, it is necessary to perform sampling by amplifying the signal to a normalized level. Thus, the preamplifier 137 may be provided with an attenuator if the level of the analog signal x (t) is relatively high.

The sample / holder circuit 138 maintains the level of the amplified analog signal x (t) according to the sampling period until the next sampling time. For example, when the analog signal x (t) is input, the sample / holder circuit 138 maintains the voltage level of the analog signal x (t) up to the next rising edge at the rising edge of the sampling clock CLK, (t)}.

The ADC circuit 139 converts the sampling signal {x (t)} assigned to each of the channels into a digital signal. The ADC circuit 139 constituting each channel in the time-interleaved ADC can be driven at a clock frequency lower than the sampling frequency relatively. The sampling signal {x (t)} assigned by the ADC circuit 139 is output as the digital signal Yn.

The specific configuration of the ADC constituting any one of the channels described above has been described. Each channel will include at least the preamplifier 137, sample / holder circuit 138, and ADC circuit 139 described above. Thus, due to the property mismatch of the preamplifier 137, each of the channels may have a mismatch of amplitude characteristics. And due to the difference in characteristics of the sample / holder circuit 138, each of the channels may have phase mismatch. Due to the difference in characteristics of the ADC circuit 139, different DC offsets may appear in the output signals of the respective channels.

FIG. 4 is a diagram illustrating a random channel sequence of the TI-ADC 136. FIG. Referring to FIG. 4, in general, the channel response measurement system 100 randomly selects a channel to be initially selected in a sampling operation.

For example, in the first measurement, the selection order of the channels may be (CH1? CH2? CH3? CH4? CH1? CH2? ...). However, in the second measurement, the start channel may be different. That is, the fourth channel CH4 may be selected as a start channel. In this case, the channel selection sequence becomes (CH4? CH1? CH2? CH3? CH4? CH1? ...). Therefore, the measurement result is obtained by providing the frequency of the first signal {s (t)} at the first frequency f1 and the frequency of the first signal {s (t)} at the second frequency f2, The resulting data may be out of time. It is meaningless to measure the response characteristics of each of the channels in this state.

5 is a diagram illustrating a method for matching the channel sequences of the present invention. Referring to FIG. 5, a channel sequence of an entire signal can be estimated by referring to a response characteristic of a second signal {r (t)} provided at a fixed frequency. Then, the response characteristic of each channel according to each frequency band can be obtained with reference to the estimated channel sequence. More detailed description is as follows.

First, in step (a), the response characteristics of the channels for the different frequencies f1 and f2 are measured. However, at this time, the response characteristics for the different frequencies f1 and f2 are randomly output values without determining a sequence of any channel. Illustratively, a second signal {r (t)}, which is a sinusoidal signal of the reference frequency fref, is input to the TI- And is applied to the ADC 130. And the response characteristics of each of the channels through the output signal will be measured. At this time, it is assumed that the response characteristic of each channel measures the amplitude as a response characteristic. Then, the response characteristic for the first signal {s ₁ (t)} can be measured in the order of (0.2, 0.3, 0.4, 0.5), although the sequence of the channel is unknown. And the response characteristic of each of the channels for the second signal {r (t)} is measured as (0.4, 0.3, 0.2, 0.1). Then, a third signal {s ₂ (t)} which is a sinusoidal wave of the second frequency f 2 and a second signal {r (t)} which is a sinusoidal signal of the reference frequency fref are applied to the TI-ADC 130 . And the response characteristics of each of the channels through the output signal will be measured. Then, the sequence of the correct channel is unknown, but the response characteristic for the third signal {s ₂ (t)} can be measured in the order of (0.7, 0.6, 0.2, 0.0). And the response characteristic of each of the channels for the second signal {r (t)} is measured to be (0.2, 0.1, 0.4, 0.3).

In step (b), the sequence of the channels is recovered by comparing the response characteristics of the channels for the second signal {r (t)} freely changeable in frequency with reference to at least two measurement results. That is, the first response characteristic for the second signal {r (t)} is (0.4, 0.3, 0.2, 0.1) and the second response characteristic is (0.2, 0.1, 0.4, 0.3). Thus, if the channels are rearranged in the same order of the response characteristics, the channel sequence can be pulled forward in the second measurement.

In step (c), the channel response measuring unit 140 adjusts the arrangement of the response characteristics for the first signals {s ₁ (t), s ₂ (t) applied while varying the frequency. That is, when the channel selection sequence recovered in step (b) is applied, the response characteristics of the first channel CH1 will be 0.2 and 0.2 at the first and second frequencies, respectively. The response characteristics of the second channel (CH2) will be (0.3, 0.0) at the first and second frequencies, respectively. The response characteristics of the third channel CH1 will be (0.4, 0.7) at the first and second frequencies, respectively. The response characteristic of the fourth channel CH4 will be (0.5, 0.6) at the first frequency and the second frequency, respectively.

It has been described that frequency dependent response characteristics can be measured through the steps described above for each of a plurality of channels. By restoring the channel selection sequence of the TI-ADC 130, it is possible to easily measure the response characteristic for each channel even if the start channel is arbitrarily selected. The response characteristic for each of the plurality of channels may be calculated through the operation of recovering the channel selection sequence, and the calculated value may be used as a value for compensating the mismatch of each of the channels.

6 is a flowchart briefly showing a method of measuring channel characteristics in a time interleaving (TI) ADC of the present invention. Referring to FIG. 6, when a measurement signal is applied, a reference signal of a fixed frequency component is added, and a response characteristic to the reference signal can be used as data for recovering a channel selection sequence.

In step S110, the first signal {s ₁ (t)} which is a sine wave of the first frequency f _{1 and} the second signal {r (t)} which is the sine wave signal of the reference frequency fref are supplied to the TI- A combined input signal is applied to the TI-ADC 130. And the response characteristic for each of the applied first signal {s ₁ (t)} and the second signal {r (t)} will be measured.

In step S120, the third signal {s ₂ (t)} which is a sinusoidal wave of the second frequency f 2 and the second signal {r (t)} which is a sinusoidal signal of the reference frequency fref are supplied to the TI- And is applied to the TI-ADC 130. And the response characteristic for each of the applied third signal {s ₃ (t)} and the second signal {r (t)} will be measured.

In step S130, the channel selection sequence of each of steps S110 and S120 is estimated by referring to the array of response characteristics of the second signal {r (t)} which is the reference signal. Since the reference signal component is provided at a fixed reference frequency (fref), it will provide a relatively stable response characteristic for each of a plurality of frequency dependent channels. Therefore, the order of an arbitrary start channel can be easily estimated by referring to the response characteristics of the second signal {r (t)}, that is, amplitude or phase.

In step S140, the channel response characteristic measured a plurality of times will be calculated with reference to the estimated channel selection sequence.

In the above, a method of determining the selection sequence of the interleaved channel has been described. Here, the features of the present invention are described in the manner that the first signal and the third signal {s ₁ (t), s ₂ (t)} of _two different frequencies are provided, but this is merely an example. That is, to measure the accurate channel response characteristics, signals of different frequencies {s ₁ (t), s ₂ (t),. , s _i (t)}.

7 is an exemplary diagram illustrating the response characteristics of each of the channels measured with reference to the channel selection sequence of the present invention. Referring to FIG. 7, amplitude response characteristics and phase response characteristics of channels CH1, CH2, CH3, and CH4 according to frequencies are shown.

The change of the response characteristic can be consistently observed for all the channels CH1, CH2, CH3 and CH4 by applying a variable frequency and recovering the channel selection sequence with reference to the reference signal provided at the fixed frequency. The calculation of the response characteristic can also be easily applied to the phase response characteristic of each of the channels CH1, CH2, CH3, and CH4. In addition, it will be appreciated that the DC offset response characteristics of each of the channels CH1, CH2, CH3, and CH4 can be calculated by the method described above.

8 is a block diagram illustrating an ADC according to another technical aspect of the present invention. Referring to FIG. 8, the ADC 200 of the present invention may include a TI-ADC 210 and a distortion compensator 220.

The TI-ADC 210 samples the input signal {x (t)} using a time interleaving method and outputs the sampled signal as a digital signal Yn. The TI-ADC 210 will sample the input signal {x (t)} and separate the sampled values into a plurality of ADCs constituting each channel with different time slots. In each of the plurality of distributed channels, the data distributed through the relatively low-speed ADC is processed and recovered as a digital signal. The measurement signal Yn, meas separated by the channel through the TI-ADC 210 may be outputted.

The distortion compensator 220 processes the separated signals Yn and meas for each channel in the frequency domain. At this time, according to the signal Yn, meas separated for each channel, a value of '0' is filled in an empty slot. A signal to which '0' is added to an empty slot in a signal separated for each channel is equal to the sum of vectors shifted by the number of channels in the frequency domain. At this time, the respective shifted vectors will have different amounts of rotation (E) depending on the number of times of transition and the channel. In the present invention, a distorted signal can be recovered using the number of shifted times and the amount of rotation according to the number of channels in the frequency domain. The restored signal is output as (Yreconst).

9 is a diagram illustrating a method of separating measured values of the present invention by channel. Referring to FIG. 9, the original measured value of an input signal may be divided into a plurality of channels CH1, CH2, CH3, and CH4. In each of the separate channels, a '0' is added between the original measurement values.

In the first channel CH1, the original measured values 1, 2 and 1 are allocated separately. And '0' s filled in empty slots between the original measurements. Accordingly, when '0' is filled between the original measured values (1, 2, 1) and the original measured values, the measured value (Y _{1 , meas} ) formed in the first channel CH ₁ is set to '100020001000' .

In the second channel (CH2), the original measured values (3, 3, 1) are allocated separately. And '0' s filled in empty slots between the original measurements. Accordingly, when '0' is filled between the original measured values (3, 3, 1) and the original measured values, the measured value (Y _{2, meas} ) formed in the second channel (CH2) is set to '030003000100' .

In the third channel CH3, the original measured values 2, 4 and 8 are allocated separately. And '0' s filled in empty slots between the original measurements. Therefore, when '0' is filled between the original measured values (2, 4, 8) and the original measured values, the measured value Y _{3, meas} formed on the third channel CH ₃ is set to '002000400080' .

In the fourth channel (CH4), the original measured values (5, 1, 4) are allocated separately. And '0' s filled in empty slots between the original measurements. Therefore, when '0' is filled between the original measured values (5, 1 and 4) and the original measured values, the measured value (Y _{4, meas} ) formed on the fourth channel CH4 is set to '000500010004' .

The measured values that are separated into the respective channels CH1, CH2, CH3, and CH4 and filled with '0' in the vacant slot can be represented by the sum of vectors shifted by the number of channels M in the frequency domain. In addition, each of the transited vectors will have a different amount of rotation depending on the number of transitions k and the corresponding channel. This relationship can be expressed by the following equation (2).

Here, M is the number of channels, Fs is a sampling frequency, and H ₁ (f) is the channel response function, and E represents the vector rotation amount of the shift.

The distortion compensator 220 of the present invention can generate a distortion compensated signal in consideration of the original measured value and the vector rotation amount for a value obtained by adding '0' to an empty time slot in each of the channels described above .

FIG. 10 shows a determinant expressing Equation (2) for all channels. Referring to FIG. 10, the matrix multiplication of the channel response characteristic H, the vector rotation amount E, and the restoration signal Yreconst corresponds to the separated measurement signal Ymeas for each channel. Here, it is assumed that the original measured values are separated by M channels.

The channel response characteristic matrix H can be calculated accurately and easily through the configuration and method described above with reference to FIG. The channel response characteristic H will be generated for each channel by a plurality of frequencies f + m x Fs / M. Here, m is an integer having a range of 0? M? M-1. That is, the channel response characteristic for the first channel (CH1) may be analyzed as many as the number of separated channels. This channel response characteristic is applied to the second channel (CH2), the third channel (CH3), and the M channel (CHM).

The vector rotation amount matrix E may be provided as a complex component for each element of the channel response characteristic matrix H. [ That is, it is a vector for providing an effect of shifting the phase by the number of '0' inserted in an empty slot or by the number of channels. The vector rotation amount matrix E defines a channel response characteristic corresponding to each of the channels, and a vector rotation amount corresponding to each of the shifted vectors.

The reconstructed signal matrix Yreconst may be regarded as an output signal when a channel measurement response is applied to a signal generated by assigning a circle measurement value to each of the channels and adding '0' to an empty slot. And the channel-by-channel separated measurement signal matrix Ymeas corresponds to an input signal including distortion before applying the response characteristic corresponding to each of the channels.

The reconstructed signal matrix Y can be calculated according to the following equation (3) by the above-mentioned determinant.

A reconstruction signal matrix (Yreconst) in which distortion is compensated can be obtained by applying an inverse matrix as shown in Equation (3). This operation procedure may be performed by the distortion compensator 220 described above. Or such an arithmetic procedure may be provided by an arithmetic algorithm and processed by various processors.

FIG. 11 is a flowchart briefly showing a signal distortion compensation method of the present invention. Referring to FIG. 11, it is possible to obtain the number of shifts and the amount of vector rotation provided at a different amount of rotation with respect to the channel, using the measured values separated by channels and the already calculated channel response characteristics. Then, the restoration signal can be obtained for each channel by referring to the vector rotation amount.

In step S210, the response characteristic for each channel of the TI-ADC will be measured. This response characteristic can be measured by using the reference signal of the fixed frequency described above and the response characteristic of each of the channels with high accuracy.

In step S220, the original measurement value will be divided into a plurality of channels. The signal separation into the channels for the original measurements has already been described in detail in Fig. That is, signal separation is performed through a process of sequentially dividing the original measured values into the respective channels and filling the empty slots in each of the channels with '0'.

In step S230, a matrix equation corresponding to equation (2) is defined in the frequency domain. That is, the principle that the signals separated into the plurality of channels are equal to the sum of the values shifted by the number of channels in the frequency domain is applied. At this time, each of the transitioned vectors will have different amounts of rotation depending on the number of transitions and corresponding channels. By applying this principle, the vector rotation amount matrix E can be obtained.

In step S240, a restoration signal is obtained using the matrices H and E obtained in the previous steps S210 and S230. In this way, the distortion-compensated ADC signal can be output.

12A and 12B are spectra showing the effect of the present invention. 12A is a result of measurement using a TI-ADC for continuous tone signals (2 tones) having two frequency components. Since the characteristics of each channel are slightly different, many interference signals are observed in addition to the original signal of -35 to -40 dB (f = 1656, f = 3007 MHz band). These interfering signals are due to interchannel mismatch (different characteristics between channels) and these interfering signals must be removed in order to obtain a precise measurement signal.

FIG. 12B shows characteristics of the channel response characteristic of the present invention and the continuous tone signal (2 tones) having two frequency components when the distortion compensation method of the present invention is applied. Here, it can be seen that all the interference signals between -75 and -90 dB have been removed.

The embodiments have been disclosed in the drawings and specification as described above. Although specific terms have been employed herein, they are used for purposes of illustration only and are not intended to limit the scope of the invention as defined in the claims or the claims. Therefore, those skilled in the art will appreciate that various modifications and equivalent embodiments are possible without departing from the scope of the present invention. Accordingly, the true scope of the present invention should be determined by the technical idea of the appended claims.

Claims (8)

A channel response measurement method of a time-interleaved analog-to-digital converter comprising:
Applying a signal mixed with a first signal of a first frequency and a second signal of a reference frequency to the analog-to-digital converter and measuring a first response;
Applying a third signal of a second frequency and the second signal to the analog-to-digital converter and measuring a second response;
Estimating a channel sequence of the analog-to-digital converter based on the second signal; And
Determining a response characteristic of each of a plurality of analog-to-digital converters to be interleaved with reference to the estimated channel sequence, the first to second responses,
Wherein the channel sequence comprises sequential selection orders by a circulator for channels corresponding to each of the plurality of analog-to-digital converters, or information about a first selected starting channel. delete delete The method according to claim 1,
Wherein the first response or the second response comprises a response characteristic to at least one of amplitude, phase, and direct current offset for the first through third signals. The method according to claim 1,
Wherein the step of estimating the channel sequence comprises matching the relative order of the response characteristics of the plurality of analog-to-digital converters based on response characteristics such as amplitude, phase or DC offset of the second signal. A distortion compensating method of a time-interleaving type analog-to-digital converter comprising:
Measuring a response characteristic (H) for each of a plurality of channels of the analog-to-digital converter being interleaved;
Separating the sampled original measurement value into each of the plurality of channels, and filling the empty time slot of each channel with '0' to generate a separation signal Ymeas;
Generating a vector rotation amount (E) in a frequency domain for a separation signal corresponding to each of the channels; And
Calculating a reconstruction signal (Yreconst) of each of the channels with reference to a response characteristic (H) and a vector rotation amount (E) for each of the channels. The method according to claim 6,
The method of claim 1, wherein in generating the vector rotation amount (E), the transitioned vectors corresponding to the time slot filled with '0' are provided as a number of transitions and a determinant defined for each of the plurality of channels. 8. The method of claim 7,
The restoration signal (Yreconst)

Is calculated from the determinant of the frequency domain satisfying the relation of < EMI ID = 1.0 >

Images