Semiconductor Circuits
The semiconductor circuit with integrated correction and test elements allows for direct inspection and diagnosis of correction elements, addressing the challenge of maintaining correction states during normal use, ensuring high accuracy and reliability in data converters.
Patent Information
- Application Number
- JP2021543087
- Authority / Receiving Office
- JP · JP
- Patent Type
- Patents
- Current Assignee / Owner
- Priority Date
- 2019-08-29
- Filing Date
- 2020-08-28
- Publication Date
- 2024-12-06
- Estimated Expiration
- 2040-08-28
AI Technical Summary
Existing data converters face challenges in maintaining correction states during normal use after manufacturing and shipping, as correction amounts can change, necessitating inspection methods before shipping and self-diagnosis functions to ensure accuracy and reliability, particularly in applications like in-vehicle components where internal failure detection and location identification are critical.
A semiconductor circuit is designed with a data converter that includes a correction element group and a test element group, allowing for direct inspection and diagnosis of correction elements using a digital circuit to test and diagnose the correction elements, ensuring all correction states are maintained and failures can be identified accurately.
The solution enables comprehensive inspection and diagnosis of correction elements, ensuring high accuracy and reliability of data converters by identifying and correcting failures in correction elements, thereby maintaining performance and reducing the risk of environmental fluctuations affecting correction states.
Smart Images

Figure 0007595575000001 
Figure 0007595575000002 
Figure 0007595575000003
Abstract
Description
[Technical field]
[0001] The present disclosure relates to semiconductor circuits, and more particularly to testing and diagnosis circuits capable of testing and diagnosing correction elements within data converters. [Background technology]
[0002] In various applications, data converters such as AD converters (ADC, analog-to-digital converters) and TD converters (TDC, time-to-digital converters) are used to measure physical quantities.
[0003] Examples include successive AD converters built into microcomputers, high-precision AD converters used in sensor AFEs (analog front ends), AD converters for detecting remaining battery capacity, wideband AD converters that realize software radios, high-precision TD converters that measure time on the order of picoseconds for ToF (Time-Of-Flight) applications, and TD converters used as elements of all-digital ADPLLs (ALL-Digital-PLLs) used in wireless applications.
[0004] In the field of data converters, there is a demand for higher performance, such as higher speed, higher accuracy, lower power consumption, and smaller space, as well as lower costs.
[0005] For example, in the case of a general-purpose successive approximation type AD converter (hereinafter referred to as a SAR-ADC), the resolution is limited to about 10 to 12 bits due to the limit of element precision.
[0006] In response to this, in recent years, techniques have been disclosed for actively correcting various element variations within analog circuits using digital circuits, and the digitization of circuits is progressing.
[0007] Examples of correction techniques disclosed in patent and non-patent documents are given below.
[0008] The ADC disclosed in Patent Document 1 is provided with a correction capacitance array in order to correct the offset and the upper capacitance of the internal DAC, and the capacitance value is set so as to compensate for nonlinearity.
[0009] Patent Document 2 describes an offset compensation circuit for a comparator, which is an element of an ADC.
[0010] In this comparator offset compensation circuit, an offset voltage compensation circuit is provided at the output of the differential preamplifier circuit, and a variable capacitance is provided at each of the differential outputs.
[0011] The offset of the comparator is made zero by adjusting the balance of the capacitance values connected to the positive and negative sides of the differential output so as to cancel out the circuit offset.
[0012] The comparator offset correction technique is related to the linearity (accuracy) of the ADC.
[0013] As shown in Patent Documents 1 and 2 above, by incorporating a correction element inside an analog circuit and performing correction using a digital circuit to offset element variations due to manufacturing variances, it is possible to reduce the built-in size of the element, reduce the circuit area, and improve accuracy.
[0014] Examples of inspection techniques disclosed in patent and non-patent literature are given below.
[0015] Patent Documents 3 and 4 disclose techniques for inspecting the DAC capacitance inside an AD.
[0016] The inspection method disclosed in Patent Document 3 includes a plurality of inspection capacitors, and judges the accuracy of the capacitors by comparing the potential of an internal capacitive DAC with the potential of the inspection capacitors.
[0017] The testing method disclosed in Patent Document 4 arranges test MOS transistors in an array at nodes of an internal capacitive DAC to form AND and OR logic, thereby detecting individual failures in the capacitive DAC (main element).
[0018] In the method described in Patent Document 3, there is a concern that the inspection accuracy may decrease if the multiple inspection capacitors themselves are broken or have poor accuracy, so it is necessary to guarantee the accuracy of the inspection capacitors.
[0019] Furthermore, in the technique described in Patent Document 4, the gate of an additional MOS transistor is hung from the analog node of the capacitive DAC, which may slow down the settling time of the capacitive DAC and reduce the conversion speed.
[0020] As described above, the testing method is usually a method for testing components such as the capacitance of a DAC that mainly contributes to AD conversion (hereinafter referred to as a main element), but is not a method for testing a correction element, and there is no method for testing the correction element itself.
[0021] In contrast to the analog inspection approaches of Patent Documents 3 and 4, Non-Patent Documents 2 and 5 disclose digital methods for diagnosing faults using digital values that reflect analog errors.
[0022] As described in Non-Patent Document 2, it is said that during inspection, it is a good idea to actively read the values in the self-calibration memory and use them for testing.
[0023] That is, since the register value indicates the amount of correction for the amount of error, if the register value is excessively large, it is indirectly determined that the error inside the analog is too large, and the analog can be rejected as a failure.
[0024] However, such digital methods are indirect methods and cannot distinguish whether the main element has deteriorated / failed, the correction element has deteriorated / failed, or the driver circuit that drives the correction element has deteriorated / failed, making it difficult to identify the true location of the failure.
[0025] Moreover, Patent Document 6 discloses a TD converter (TDC, time-to-digital converter) as a data converter.
[0026] Non-Patent Document 3 discloses an open-loop integrating amplifier for achieving low power consumption. [Prior art documents] [Patent documents]
[0027] [Patent Document 1] U.S. Patent No. 8,638,248 [Patent Document 2] Patent No. 5412369 [Patent Document 3] JP 2015-128203 A [Patent Document 4] U.S. Pat. No. 5,175,547 [Patent Document 5] JP 2017-38200 A [Patent Document 6] JP 2010-273118 A [Non-patent literature]
[0028] [Non-Patent Document 1] Bob Verbruggen, “A 2.1mW 11b 410MS / s Dynamic Pipelined SAR ADC with Background Calibration in 28nm Digital CMOS”, VLSI Symposium 2013 [Non-Patent Document 2] Digitally Assisted Analog Test Technology, Haruo Kobayashi, Institute of Electronics, Information and Communication Engineers, July 22, 2010 [Non-Patent Document 3] Badr Malki, “A complementary dynamic residue amplifier for a 67 dB SNDR 1.36 mW170 MS / s pipelined SAR ADC”, ESSCIRC 2014. Figure 31 shows the gain correction function. Summary of the Invention [Problem to be solved by the invention]
[0029] In data converters that incorporate correction elements to achieve higher accuracy, higher speeds, and lower power consumption, if the amount of correction, such as background correction, can change during normal use after shipment, inspection methods before shipment and self-diagnosis functions after shipment are required to guarantee all correction conditions.
[0030] The present disclosure provides a semiconductor circuit for testing a correction function of a data converter in a semiconductor circuit having a converter. [Means for solving the problem]
[0031] The disclosure disclosed in this disclosure is roughly configured as follows as a means for solving the problem.
[0032] In the following configuration, the numbers and symbols in parentheses () indicate the corresponding numbers and symbols among the disclosed embodiments, and are intended merely to clarify the correspondence relationship and not to limit the present disclosure.
[0033] A semiconductor device according to one embodiment of the present disclosure is a semiconductor circuit (1) including a data converter (100) that outputs a digital output signal corresponding to a signal under measurement, and includes an analog circuit (101) that inputs the signal under measurement and a digital circuit (102) that outputs the digital output signal, the analog circuit (101) has a correction element group (10) including correction elements for correcting nonlinearity that occurs in the process of converting the signal under measurement to the digital output signal in the analog circuit, and a test element group (20) including test elements for testing the correction elements, and the digital circuit (102) tests the correction element group (10) using the test element group (20). Effect of the Invention
[0034] According to the present disclosure, it is possible to test the correction function in a semiconductor circuit having a data converter. [Brief description of the drawings]
[0035] [Figure 1] FIG. 1 is a block diagram showing an example of the configuration of a semiconductor circuit including a data converter according to a first embodiment. [Diagram 2] FIG. 2 is a diagram illustrating an example of the configuration of the correction circuit according to the first embodiment. [Diagram 3] FIG. 3 is a block diagram showing an example of the configuration of a semiconductor circuit including a data converter according to the second embodiment. [Figure 4] FIG. 4 is a circuit diagram showing an example of the comparator in FIG. [Diagram 5] FIG. 5 is a diagram showing a convenient definition of the comparator offset code. [Figure 6] FIG. 6 is a diagram illustrating an example of an output waveform of the comparator. [Figure 7] FIG. 7 is a diagram of example waveforms illustrating the effect of the offset compensation capacitance of a comparator. [Figure 8] FIG. 8 is a diagram showing a convenient redefinition of the comparator offset code including the correction. [Figure 9]FIG. 9 is a diagram showing an example of a first correspondence table between comparator offset correction control codes and offset correction amounts. [Figure 10] FIG. 10 is a diagram showing an example of a second correspondence table between comparator offset correction control codes and offset correction amounts. [Figure 11] FIG. 11 is a diagram showing the relationship between the correction control code and the offset correction amount. [Figure 12] FIG. 12 is a diagram showing the effect of noise on the comparator. [Figure 13] FIG. 13 is a diagram showing the effect of nonlinearity of a comparator. [Figure 14] FIG. 14 shows a circuit diagram of the offset control unit in the second embodiment. [Figure 15A] FIG. 15A is a diagram showing a procedure for checking and diagnosing each weight of an offset correction element of a comparator in the second embodiment. [Figure 15B] FIG. 15B is a continuation of FIG. 15A and shows a procedure for checking and diagnosing each weight of the offset correction element of the comparator. [Figure 16] FIG. 16 is a diagram showing an example of a digital unit for inspection and diagnosis of an offset correction element of a comparator in the second embodiment. [Figure 17] FIG. 17 is a table showing the correction offset amount when the first fixed code is set in the second embodiment. [Figure 18] FIG. 18 is a table showing the correction offset amount when the second fixed code is set in the second embodiment. [Figure 19] FIG. 19 is a schematic diagram showing generation of a first fixed code and execution of a binary search in the second embodiment. [Figure 20] FIG. 20 is a schematic diagram showing a CDF determination in the second embodiment. [Figure 21] FIG. 21 is a schematic diagram showing generation of a second fixed code and weight determination by CDF comparison in the second embodiment. [Figure 22] FIG. 22 is a schematic diagram showing weight determination using check bits in the second embodiment. [Figure 23A] FIG. 23A is a diagram showing a procedure for checking and diagnosing the correction range of the offset correction element of the comparator in the third embodiment. [Figure 23B] FIG. 23B is a continuation of FIG. 23A and shows a procedure for checking and diagnosing the correction range of the offset correction element of the comparator. [Figure 24] FIG. 24 is a diagram showing an example of a digital unit for inspection and diagnosis of an offset correction element of a comparator in the third embodiment. [Diagram 25] FIG. 25 is a schematic diagram showing range determination in the third embodiment. [Figure 26] FIG. 26 is a diagram showing a procedure for inspecting and diagnosing the correction function of the offset correction element of the comparator in the fourth embodiment. [Figure 27] FIG. 27 is a diagram showing an example of a digital unit for inspection and diagnosis of an offset correction element of a comparator in the fourth embodiment. [Figure 28] FIG. 28 is a schematic diagram showing a toggle inspection in the fourth embodiment. [Figure 29] FIG. 29 is a diagram showing an example of the order of examination and diagnosis in the fourth embodiment. [Diagram 30] FIG. 30 is a block diagram showing a correction mechanism of a residue amplifier in the fifth embodiment. [Diagram 31] FIG. 31 is a circuit diagram illustrating an example of a residual amplifier with a gain correction function in the fifth embodiment. [Diagram 32] FIG. 32 is a circuit diagram illustrating an example of an amplification time adjustment unit in the fifth embodiment. [Diagram 33] FIG. 33 is a circuit diagram illustrating an example of an amplification time adjustment unit in the sixth embodiment. [Diagram 34] FIG. 34 is a circuit diagram showing a signal amplifying unit of a residual amplifier in the seventh embodiment. [Diagram 35] FIG. 35 is a circuit diagram illustrating an offset control unit of a residual amplifier in the seventh embodiment. [Diagram 36]FIG. 36 is a block diagram showing a correction mechanism for the MSB capacitance of a DAC in the eighth embodiment. [Figure 37] FIG. 37 is a circuit diagram illustrating an example of a capacitance correction unit in the eighth embodiment. [Figure 38] FIG. 38 is a block diagram showing a correction mechanism for the ADC gain in the ninth embodiment. [Figure 39] FIG. 39 is a circuit diagram illustrating an example of a capacitance correction unit in the ninth embodiment. [Diagram 40] FIG. 40 is a block diagram showing a correction mechanism of a TDC converter (time-to-digital converter) in the tenth embodiment. [Diagram 41] FIG. 41 is a circuit diagram showing an example of a delay circuit in the tenth embodiment. [Diagram 42] FIG. 42 is a circuit diagram showing an example of a reference voltage circuit in the eleventh embodiment. [Figure 43A] FIG. 43A is a diagram showing an example in which testing elements of opposite polarity are controlled in the twelfth embodiment. [Figure 43B] FIG. 43B is a continuation of FIG. 43A, and shows an example in which testing elements of opposite polarity are controlled. [Figure 44A] FIG. 44A is a diagram showing an example in which testing elements of opposite polarity are controlled in the twelfth embodiment. [Figure 44B] FIG. 44B is a continuation of FIG. 44A, and shows an example in which testing elements of opposite polarity are controlled. [Diagram 45] FIG. 45 shows an example of an offset control section of a comparator in the thirteenth embodiment. [Diagram 46] FIG. 46 is a diagram illustrating an example of correction of an AD converter. [Figure 47] FIG. 47 is a diagram illustrating an example of correction of an AD converter. [Figure 48] FIG. 48 shows a circuit diagram of a conventional offset control section. [Figure 49] FIG. 49 is a diagram showing the difference in correction resolution due to weighting of the correction element. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
[0036] (Findings on which the present invention is based) The present inventors have found that the data converter described in the "Background Art" section has the following problems.
[0037] To explain in more detail, if the amount of correction can change during normal use after manufacturing and shipping, an inspection method before manufacturing and shipping and a self-diagnosis function after shipping are required to guarantee all correction states.
[0038] In particular, when used as an in-vehicle component, internal controllability, such as detection of internal analog failures and identification of the location of the failure, is a necessary requirement, and the inventors have found a problem that the correction element itself needs to be inspected.
[0039] When adding analog circuits for inspection and diagnosis, there is an issue that the conversion speed of the AD conversion itself decreases.
[0040] Furthermore, when using a digital method of reading the register value of a memory that indicates the amount of correction for inspection and diagnosis, there is a problem in that it is difficult to identify the location of the failure because it is an indirect method.
[0041] Furthermore, as described above, the conventional configuration is an inspection method that targets the main elements (such as DACs, comparators, and amplifiers) that mainly contribute to the data converter, or the sum or state of the main elements and correction elements, and has the problem that it is not possible to comprehensively inspect and diagnose the correction elements themselves.
[0042] In addition, as described later in this specification, when a correction element is included in both the positive and negative sides of a differential circuit, there is a problem in that it is difficult to identify the location of a failure in the differential circuit, such as which element on either the positive or negative side has failed.
[0043] Furthermore, as will be described later in this specification, when inspecting and diagnosing the offset compensation circuit of a comparator, there is an issue that it is not possible to inspect all of the weightings of the offset correction elements for the comparator because the input / output function of the comparator is nonlinear.
[0044] On the other hand, in order to perform the correction, it is necessary to carry out control different from the normal operation.
[0045] Methods for controlling correction include trimming before shipment and foreground correction during a period when normal operation is stopped, but these methods have the problem of being vulnerable to environmental fluctuations such as temperature, voltage, and deterioration over time.
[0046] In the ADC disclosed in Non-Patent Document 1, a background correction technique is disclosed in which element correction is performed in parallel with normal operation.
[0047] In Non-Patent Document 1, the offset of a comparator and the gain of a residue amplifier of a pipeline ADC are corrected in parallel with normal operation.
[0048] Degradation of characteristics caused by environmental fluctuations (temperature, voltage, aging, etc.), which cannot be addressed by trimming or foreground correction, can be corrected in parallel with normal operation, thereby maintaining the robustness of the system.
[0049] On the other hand, because the selection state of the correction element changes in order to make corrections in response to environmental fluctuations, there are issues with pre-shipment inspections to guarantee all correction states, as well as with diagnosis methods in the event of a malfunction during normal use.
[0050] As described above, it has been shown that a correction element is provided in an analog circuit to perform correction, and that there are methods such as trimming, foreground correction, and background correction depending on the timing of correction.
[0051] In such a correction circuit, special control is performed for correction, errors in the analog circuit are detected, and correction is performed by applying feedback to the components of the data converter.
[0052] There are two main ways to return.
[0053] First, in a conventional AD converter (500) shown in FIG. 46, a quantizer output signal obtained by converting analog information into digital information is output from an analog circuit (501).
[0054] Furthermore, in the digital circuit (502) to which the quantizer output signal is input, an error detection circuit (504) within the digital circuit detects an error in the analog circuit and outputs an error signal, and a correction calculation according to the error signal is performed on the quantizer output signal in the encoding section (502) within the digital circuit to which the error signal is input.
[0055] This method is one of the digital-to-digital feedback correction methods.
[0056] Secondly, in the AD converter (510) shown in FIG. 47, which is a conventional example, a quantizer output signal is output from an analog circuit (511).
[0057] Furthermore, in the digital circuit (512) to which the quantizer output signal is input, an error detection circuit (514) within the digital circuit detects an error in the analog circuit and outputs an error signal, a correction code generation circuit (515) within the digital circuit to which the error signal is input outputs a control code for correcting an element within the analog circuit, the constant of the element is corrected in the analog circuit to which the control code is input and the quantizer output in the corrected state is output, and the usual weighted digital calculation is performed in the encoding section (513) within the digital circuit.
[0058] This method is one of the digital to analog feedback correction methods.
[0059] The digital-to-digital feedback correction method shown in FIG. 46 has the disadvantage that when the elements inside the analog circuit vary excessively, that is, when there is an error equivalent to 1 LSB or more of the AD converter, it cannot be corrected completely by digital calculation alone.
[0060] On the other hand, in the digital-to-analog feedback correction method shown in FIG. 47, even if the elements inside the analog circuit vary excessively, the excessive variation can be corrected by providing a correction element that corrects the excessive variation, and the circuit area can be made small and high accuracy can be maintained.
[0061] On the other hand, since a large number of buses of the control codes fed back from digital to analog are provided, there is a drawback in that the circuit becomes more complicated.
[0062] As described above, in terms of characteristics and robustness, it is best to perform feedback correction from digital to analog and background correction, but this requires inspection and diagnosis before shipment.
[0063] In order to explain the specifications required for the correction element, a supplementary explanation will be given on the relationship between the weighting of the correction element and the resolution.
[0064] FIGS. 49(a) to (c) show the difference in correction resolution due to weighting of the correction element.
[0065] In FIG. 49, the horizontal axis indicates the correction code (the value of the control code), and the vertical axis indicates the offset correction voltage.
[0066] In the example of FIG. 49, the weighting of the offset correction element is changed to calculate the offset correction amount.
[0067] That is, in FIG. 49(a), the weighting between adjacent correction elements is "2", and the weighting from the LSB to the MSB is "1:2:4:8" with the LSB being 1.
[0068] In FIG. 49(b), the weighting between adjacent correction elements is set to "1.8", and the weighting from LSB to MSB is set to "1.0:1.8:(1.8 squared):(1.8 cubed)".
[0069] In FIG. 49(c), the weighting between adjacent correction elements is set to "2.2", and the weighting from LSB to MSB is set to "1:(2.2):(2.2 squared):(2.2 cubed)".
[0070] The correction resolution is determined by the maximum value of the step voltage between adjacent codes.
[0071] In FIG. 49(a), all values are 1 mV, so the correction resolution is 1 mV.
[0072] In the case of FIG. 49(b), the difference between "Code=0" and "Code=1" corresponds to 1 mV, and the difference between "Code=7" and "Code=8" corresponds to 0.208 mV, so the correction resolution is 1 mV.
[0073] In FIG. 49(c), the difference between "Code=7" and "Code=8" is the largest and corresponds to 2.608 mV, so the correction resolution is 2.608 mV.
[0074] As further explained in the example of Figure 49, the test specifications required for correction elements do not require the guarantee of linearity as is done for main elements such as DAC capacitance, but rather require a test of the resolution (maximum value of the step voltage), and it is necessary to test that the weighting requirement between adjacent elements is "2 or less (≦2)."
[0075] In order to solve the above problems, a semiconductor device according to one embodiment of the present disclosure is a semiconductor circuit including a data converter that outputs a digital output signal corresponding to a signal under measurement, and includes an analog circuit that quantizes the input signal under measurement, and a digital circuit that outputs the digital output signal, wherein the analog circuit has a correction element group including correction elements for correcting nonlinearity occurring in the process of converting the signal under measurement to the digital output signal in the analog circuit, and a test element group including test elements for testing the correction elements, and the digital circuit tests the correction element group using the test element group.
[0076] This allows the function and weighting of the correction elements to be inspected and diagnosed, making it possible to inspect the correction element group.
[0077] For example, the digital circuit may include a self-diagnosis code generation circuit that generates a self-diagnosis code for enabling or disabling each correction element included in the correction element group and for enabling or disabling each test element included in the test element group, an error detection circuit that detects the difference between a quantized output signal of the analog circuit output in response to the self-diagnosis code and a predetermined threshold value as an error signal, and a self-diagnosis circuit that generates a self-diagnosis signal indicating a diagnosis result of the correction element group based on the error signal.
[0078] This allows the control code generated by the self-diagnosis code generation circuit to be supplied to the correction element inside the data converter, and the function and weighting of the correction element can be inspected and diagnosed by the self-diagnosis circuit. In other words, this makes it easy to inspect each individual correction element included in the correction element group.
[0079] For example, the self-diagnosis code may include a first fixed code and an optimum value code, the first fixed code being a code for enabling one correction element selected as a test target from the group of correction elements and disabling the other correction elements and test elements, the optimum value code being a code for minimizing the error signal with respect to the first fixed code and generated by a binary search method, and the self-diagnosis circuit may calculate an integrated value of the error signal corresponding to a combination of the first fixed code and the optimum value code, determine whether the integrated value is within a predetermined range, and determine that the group of correction elements is defective if it is not within the predetermined range.
[0080] In this way, if the error cannot be minimized within a predetermined range, it is determined that the correction element group is defective, that is, that one of the correction elements is defective.
[0081] For example, the self-diagnosis code includes a first fixed code, a second fixed code, and an optimum value code, the first fixed code being a code for enabling one correction element selected as a test target from the group of correction elements and disabling the other correction elements and test elements, the optimum value code being a code for minimizing the error signal with respect to the first fixed code and generated by a binary search method, the signal under test being a differential signal including a first polarity signal and a second polarity signal, the group of correction elements including a first element group connected to a transmission path of the first polarity signal and a second element group connected to a transmission path of the second polarity signal, the first fixed code corresponding to the first element group, disabling the second element group, and The fixed code may be a code for disabling one correction element selected as a test target in the first element group, enabling other correction elements having a smaller weight than the selected correction element, and enabling a test element, the optimum value code corresponding to a second element group and disabling the first element group, the self-diagnosis circuit calculating a first integrated value by integrating the error signal corresponding to a combination of the first fixed code and the optimum value code, calculating a second integrated value by integrating the error signal corresponding to a combination of the second fixed code and the optimum value code, determining whether the first integrated value and the second integrated value are larger or smaller, and determining that the test target correction element is defective when the relative magnitude does not satisfy a predetermined condition.
[0082] This allows the operating point for self-diagnosis to be set to the operating point that minimizes the above error by the optimum value code. In other words, whichever correction element is the test object, the operating point that minimizes the above error can be used, and the self-diagnosis of the test object correction element can be facilitated. In other words, the test of each individual correction element is facilitated.
[0083] For example, the self-diagnosis circuit may sequentially select one correction element as a test target from the group of correction elements.
[0084] As a result, even if the signal under test is a differential signal, if the error cannot be minimized within a predetermined range, it can be determined that the correction element group is defective, that is, that one of the correction elements is defective.
[0085] For example, one or more of the correction elements in the group of correction elements may also be used as the inspection element.
[0086] This facilitates self-diagnosis of each compensation element even if the signal under test is a differential signal.
[0087] A semiconductor device according to one aspect of the present disclosure includes a data converter that outputs a digital output signal corresponding to a signal under measurement, the data converter comprising an analog circuit (101) and a digital control circuit (102), the analog circuit (101) comprising at least a main element group (114) each having a reference weight for quantizing the signal under measurement, a correction element group (140) each having an adjustment weight for correcting various analog quantities, a test element group (141) each having a test weight for testing the weights of the correction element group, a quantizer (116) that converts information of the signal under measurement into a digital signal, and an analog control unit (117) that controls within the analog circuit, the digital circuit (102) controlling the quantizer output signal of the quantizer (116) to have a predetermined weight. the error detection circuit (104, 119) which compares an integration result obtained by integrating the digital output signal a predetermined number of times with a predetermined threshold value and outputs an error signal in order to detect an error in an adjustment weight of one of the correction element group (140) by referring to one or more test weights of the test element group (141); a self-diagnosis code generation circuit (107, 122) which controls the correction element group (140) and outputs a control code including a plurality of bit signals for diagnosing each correction element; and a self-diagnosis circuit (120) which receives the error signal and outputs a self-diagnosis output signal, and the self-diagnosis circuit performs self-diagnosis of each correction element of the correction element group (140).
[0088] The quantizer (116) according to an embodiment of the present disclosure may be one or more than one.
[0089] Furthermore, the output result of the comparator used for inspection and diagnosis may be one or more than one.
[0090] The size of the test element group (141) according to an embodiment of the present disclosure may be set to be equal to or smaller than the LSB size of the test element (140), thereby enabling weighted testing and diagnosis.
[0091] In one embodiment of the present disclosure, the correction element group (140) and the test element group (141) may be capacitors (400, 401), MOS transistors (372, 373, 374, 375), or resistors (456, 457).
[0092] A self-diagnosis code generation circuit (170) according to an embodiment of the present disclosure includes an optimal correction code output mode for receiving the error output signal and outputting optimal correction codes (292, 288, 289) so as to minimize the quantization error of the quantizer, and a logic inversion code output mode for outputting logic inversion codes (301A, 301B) obtained by logic inverting a first bit signal of the optimal correction signal that controls a correction element to be tested. During testing and self-diagnosis, the circuit includes a first step (261, 262, 263) for setting the circuit to the optimal correction code output mode, supplying the optimal correction code to the analog circuit, and obtaining an output of the error detection circuit as a first error signal, a second step (265, 266) for setting the circuit to the logic inversion code output mode, supplying the logic inversion code to the analog circuit, and obtaining an output of the error detection circuit as a second error signal, and a third step (267, 268, 269) for comparing the magnitude relationship between the first error signal and the second error signal to perform a fault diagnosis of the correction element to be tested.
[0093] In another embodiment, the self-diagnosis code generation circuit outputs a first weight check code output mode (182, 185) in which a first bit signal for controlling a first correction element to be checked is set to 1 or 0, a second correction bit signal group for controlling a correction element having an adjustment weight less than the first correction element is set to the inverted logic of the first bit signal, a check bit signal for controlling the check element is set to the inverted logic of the first bit signal, and other bit signals are set to predetermined fixed signals, the first bit signal being set to the inverted logic of the bit signal when the first weight check code output mode is set, the second bit signal group being set to the inverted logic of the bit signal when the first weight check code output mode is set, and the check bit for controlling the check element is set to the bit signal when the first weight check code output mode is set. and a second weighted check code output mode (184, 187, 188, 189) in which the specified multiple bit signals are output as an inverted logic of the first weighted check code output mode and the other bit signals are output as a predetermined fixed signal. During a test / self-diagnosis, the system includes a first step (152, 153) of setting the first weighted check code output mode, supplying the specified multiple bit signals to the analog circuit, and obtaining an output of the error detection circuit as a first error signal, a second step (156, 157) of setting the double weighted check code output mode, supplying the specified multiple bit signals to the analog circuit, and obtaining an output of the error detection circuit as a second error signal, and a third step (158, 159, 160) of calculating a difference between the first error signal and the second error signal and performing a weighted diagnosis of the correction element group.
[0094] In another embodiment, when inspecting and diagnosing a correction element group of a circuit (114, 130) having a differential configuration (or having two sets), in the first step (152, 153), the correction element group of the polarity not being inspected is used to obtain optimal correction codes (184, 187, 188, 189) that minimize the quantization error, thereby bringing the CDF to approximately 0.5 and making inspection and diagnosis easier.
[0095] In addition, as another embodiment, the method may include a first step (217, 218) of inputting a first reference signal (252) and obtaining the optimum correction code as a first optimum correction code, a second step (219, 220) of inputting a second reference signal and obtaining the optimum correction code as a second optimum correction code, and a third step (221, 222, 224) of calculating the difference between the first optimum correction code and the second optimum correction code to perform range diagnosis of a group of correction elements.
[0096] In another embodiment, when testing a correction element group of a circuit (114, 130) having a differential configuration (or having two sets), the test element group to be used as the reference for the test may be the test element group (355) of the polarity (or the set) on the side being tested, or the test element group (357) of the opposite polarity (or the set not being tested).
[0097] In another embodiment, the data converter may be an analog-to-digital converter (ADC, 110) or a time-to-digital converter (TDC, 440).
[0098] In another embodiment, the test elements (471) may be used as the correction elements (470).
[0099] However, in this embodiment, it is not possible to check the LSB.
[0100] Hereinafter, embodiments will be described with reference to the drawings.
[0101] In order to avoid making the following description unnecessarily redundant and to facilitate understanding by those skilled in the art, repeated descriptions may be omitted.
[0102] Furthermore, the following description and detailed description of the drawings do not limit the subject matter recited in the claims.
[0103] <First embodiment> The circuit configuration and functions of the first embodiment will be described below with reference to the drawings.
[0104] In the first embodiment, the basic functions of a semiconductor circuit having a data converter according to the present disclosure are shown.
[0105] Fig. 1 is a block diagram showing an example of the configuration of a semiconductor circuit 1 including an AD converter 100 as a data converter according to the first embodiment. Fig. 2 is a diagram showing an example of the configuration of a correction circuit 109 according to the first embodiment.
[0106] The semiconductor circuit 1 in Fig. 1 is a circuit formed in a semiconductor device such as an LSI. The semiconductor circuit 1 in the figure includes an AD converter (ADC, analog-to-digital converter) as an example of a data converter. Note that the data converter is not limited to the AD converter 100. The data converter may include a TD converter (TDC, time-to-digital converter) in addition to the AD converter.
[0107] The AD converter 100 includes an analog circuit 101 and a digital circuit 102, and the analog circuit 101 includes a correction circuit 109. The correction circuit 109 includes a correction element group 10, a test element group 20, and drivers 15 to 18 and 20 in the example of FIG.
[0108] The correction element group 10 corrects nonlinearity that occurs in the process of converting a signal under test into a digital output signal in the analog circuit 101. The nonlinearity referred to here includes, for example, offset (amount of deviation occurring in a signal), gain (deviation in magnification), and capacitance value (relative deviation between reference element arrays within). In the following, offset will be described as a specific example of nonlinearity that is the subject of correction by the correction element group 10. Path0 in the figure indicates a path for transmitting the signal under test, which is an external analog signal. It is assumed that an offset, which is a deviation, occurs in the path path0 due to characteristic variations in the analog circuit. The offset is caused by characteristic variations in the circuit elements during manufacturing, and characteristic variations caused by environmental changes and aging. The control code in the figure includes a correction code that controls the correction element group 10 during normal operation, and an automatic correction code that controls the test element group 20 and the correction element group 10 during self-diagnosis.
[0109] The correction element group 10 includes correction elements 11 to 14. The correction elements 11 to 14 are given weights of, for example, 1x, 2x, 4x, and 8x. Each of the correction elements 11 to 14 is controlled to be enabled or disabled by a corresponding driver. A correction element controlled to be enabled adds a correction amount according to the weight to the path path0. A correction element controlled to be disabled does not add a correction amount to the path path0. Here, the correction amount is a physical amount such as a voltage or current. The correction elements 11 to 14 are composed of, for example, a voltage source, a current source, a capacitive element, or a resistive element for generating the correction amount.
[0110] The drivers 15-16 drive, that is, enable or disable, the corresponding correction elements according to the corresponding bits in the control code. The drivers 15-16 are composed of, for example, a gate circuit that outputs a binary signal according to the bit of the control code, or a switch circuit.
[0111] The correction amount is not limited to voltage or current, but may be a physical amount such as delay time, amount of heat, or amount of pressure.
[0112] The test element group 20 includes at least one test element 21. The test element 21 in Fig. 2 is the same element with the same weight as the correction element 11. In other words, the test element 21 has the same configuration as the correction element with the smallest weight among the correction elements included in the correction element group 10.
[0113] The test element group 20 may include other test elements having the same weight as the test element 21, or may include other test elements having a weight different from that of the test element 21. The test element 21 may also be configured to have a weight smaller than the correction element with the smallest weight.
[0114] In FIG. 1, a digital circuit 102 includes at least an encoding unit 103, an error detection circuit 104, a self-diagnosis circuit 105, a correction code generation circuit 106, a self-diagnosis code generation circuit 107, and a selector .
[0115] An external analog signal is input as an input signal to the AD converter 100 .
[0116] The analog circuit 101 has an internal reference voltage source (not shown) that generates a reference voltage, and can perform AD conversion of either an external analog signal or an internal reference voltage.
[0117] During normal operation, an external analog signal is input to analog circuit 101, where the information of the external analog signal is quantized into a digital signal, which is then output as a quantizer output signal.
[0118] The quantizer output signal is input to the encoding unit 103 .
[0119] The encoding unit 103 performs digital calculations on one or more quantizer output signals with predetermined weights, and outputs a digital output signal.
[0120] When a correction element (not shown) in the analog circuit 101 is inspected and diagnosed, the internal reference voltage is converted into an analog signal.
[0121] After being AD converted, the quantizer output signal is input to an error detection circuit 104 .
[0122] The error detection circuit 104 compares this with a predetermined threshold (for example, an expected value) and outputs an error signal.
[0123] The error signal indicates error information for inspecting and diagnosing the correction signal, and may be the sign of the error information (positive error or negative error) or may be quantitative information of the error information.
[0124] Moreover, in order to remove such noise, the signal may be integrated one or more times.
[0125] The error signal is input to a self-diagnosis circuit 105 , a correction code generation circuit 106 , and a self-diagnosis code generation circuit 107 .
[0126] The error signal is input to a correction code generating circuit 106 which outputs a correction code.
[0127] The error signal is input to a self-diagnosis code generating circuit 107, which outputs a self-diagnosis code.
[0128] The correction code is a digital signal having a multi-bit signal, which is used to control the state of a correction element that corrects the analog characteristics inside the analog circuit 101.
[0129] The self-diagnostic code is a digital signal having multiple bits that is used to check and diagnose the control function and weighting of each correction element.
[0130] The self-diagnosis circuit 105 to which the error signal is input holds, compares and calculates the error information for inspecting the state of the correction element, and outputs a self-diagnosis signal.
[0131] The self-diagnosis circuit 105 may perform a diagnostic judgment based on the error signal by setting a threshold value, or may hold error information for two different states and perform a diagnostic judgment of the weighting in order to determine the relative weighting of the correction element.
[0132] In addition, the self-diagnosis signal that is output may output position information indicating which of the correction elements has failed, or may output a signal that suggests the weighting of one of the correction elements that is the subject of inspection and diagnosis.
[0133] The correction code and the self-diagnosis code are input to a selector 108 .
[0134] The selector 108 outputs a control code having a plurality of bits for controlling the state of the correction element in the analog circuit 101 .
[0135] An operation mode signal is input to control the selector 108, and the operation mode signal switches between outputting a correction code as a correction mode for correcting various characteristics of the analog circuit 101, and outputting a self-diagnosis code as a self-diagnosis mode for inspecting and diagnosing each correction element of the correction element group of the analog circuit 101.
[0136] When the control code is a correction code, the analog circuit 101 to which the control code is input outputs a quantizer output signal in a state in which the analog characteristics of the analog circuit 101 have been corrected.
[0137] Also, when the control code is a self-diagnosis code, a quantizer output signal indicating the weighting state of the correction element inside the analog circuit 101 is output.
[0138] Thus, the AD converter 100 in the first embodiment is a semiconductor circuit 1 including a data converter 100 that outputs a digital output signal corresponding to a signal under measurement, and is equipped with an analog circuit 101 that quantizes an input signal under measurement and a digital circuit 102 that outputs a digital output signal. The analog circuit 101 has a correction element group 10 including correction elements for correcting nonlinearity occurring in the process of converting the signal under measurement into a digital output signal in the analog circuit, and a test element group 20 including test elements for testing the correction elements. The digital circuit 102 tests the correction element group 10 using the test element group 20.
[0139] This allows the function and weighting of the correction elements to be inspected and diagnosed, making it possible to inspect the correction element group.
[0140] Here, the digital circuit 102 may include a self-diagnosis code generation circuit 107 that generates a self-diagnosis code for enabling or disabling each correction element included in the correction element group and for enabling or disabling each test element included in the test element group, an error detection circuit 104 that detects the difference between a quantized output signal of the analog circuit output in response to the self-diagnosis code and a predetermined threshold value as an error signal, and a self-diagnosis circuit 105 that generates a self-diagnosis signal indicating a diagnosis result of the correction element group based on the error signal.
[0141] This allows the control code generated by the self-diagnosis code generation circuit to be supplied to the correction element inside the data converter, and the function and weighting of the correction element can be inspected and diagnosed by the self-diagnosis circuit. In other words, this makes it easy to inspect each individual correction element included in the correction element group.
[0142] Therefore, in the case of the first embodiment, in addition to being provided with a correction code generation circuit for correcting the analog circuit 101, a self-diagnosis code generation circuit for inspecting the function and weighting of each correction element is provided, and self-diagnosis codes corresponding to each correction element are generated to comprehensively inspect and diagnose all correction elements, so that it is possible to inspect and diagnose all correction elements.
[0143] In other words, in the conventional technology, linearity and partial failures were inspected in the state of the main element, or in the state in which the main element and the correction element are added together, and normality could only be guaranteed in limited correction states. In contrast, in this embodiment, the normality of each individual correction element is guaranteed, thereby having the effect of making it possible to guarantee the entire correction state.
[0144] As described above, the semiconductor circuit according to the first embodiment is a semiconductor circuit 1 including a data converter 100 that outputs a digital output signal corresponding to a signal under measurement, and is equipped with an analog circuit 101 that quantizes an input signal under measurement, and a digital circuit 102 that outputs a digital output signal. The analog circuit 101 has a correction element group 10 including correction elements for correcting nonlinearity occurring in the process of converting the signal under measurement into a digital output signal in the analog circuit, and a test element group 20 including test elements for testing the correction elements. The digital circuit 102 tests the correction element group 10 using the test element group 20.
[0145] Here, the digital circuit 102 may include a self-diagnosis code generation circuit 107 that generates a self-diagnosis code for enabling or disabling each correction element included in the correction element group and for enabling or disabling each test element included in the test element group, an error detection circuit 104 that detects the difference between a quantized output signal of the analog circuit output in response to the self-diagnosis code and a predetermined threshold value as an error signal, and a self-diagnosis circuit 105 that generates a self-diagnosis signal indicating a diagnosis result of the correction element group based on the error signal.
[0146] Here, the analog circuit 107 includes a quantizer for quantizing the measured signal or reference signal after correction by the correction element group and converting it into a quantizer output signal, and the digital circuit includes an encoding unit for converting and outputting the quantizer output signal into a digital output signal weighted by a predetermined weighting, an error detection circuit for outputting an accumulation result obtained by accumulating the difference between the quantizer output signal and a predetermined threshold value multiple times as an error signal, a self-diagnosis code generation circuit for enabling or disabling each correction element included in the correction element group and outputting a self-diagnosis code for enabling or disabling each inspection element included in the inspection element group, and a self-diagnosis circuit for receiving the error signal and outputting a self-diagnosis output signal indicating the pass / fail of the correction element group, and self-diagnosis of each correction element in the correction element group may be performed by the self-diagnosis circuit.
[0147] Second Embodiment The circuit configuration and operation of the second embodiment will be described below with reference to the drawings.
[0148] In the second embodiment, in order to show a more specific configuration example, a successive AD converter (FIG. 2) is taken as an example of a data converter, and a configuration example will be described in which each correction element of a group of correction elements that performs offset correction of a comparator is subject to inspection and diagnosis.
[0149] In addition, with regard to components appearing in the second to thirteenth embodiments, components with the same names as those in the first embodiment or other embodiments indicate more specific examples of the same components even if they have different symbols.
[0150] Fig. 3 is a block diagram showing an example of the configuration of a semiconductor circuit including a data converter according to the second embodiment. Fig. 3 shows a successive approximation type successive AD converter 110 as an example of the data converter.
[0151] The successive AD converter 110 in Fig. 3 is a specific example of the AD converter 100 in Fig. 1, and shows an example of a differential circuit configuration. The AD converter 100 of the present disclosure is not limited to this, and can also be applied to a single-ended circuit configuration.
[0152] The components are shown below.
[0153] The successive AD converter 110 includes an analog circuit 111 and a digital circuit 112, to which an external analog signal (VINP, VINM, VINP is the positive side, VINM is the negative side) is input, and after time discretizing and amplitude quantizing the analog signal, it outputs a quantizer output signal, and a predetermined digital calculation is performed by an encoder circuit (not shown), and outputs a digital signal that is the result of the AD conversion.
[0154] The analog circuit 111 includes an analog element section to be corrected and an element control section.
[0155] The analog element section includes a sample and hold circuit 113, a digital-to-analog converter (DAC) 114, a comparator 116, and a driver circuit 115.
[0156] The element control section includes a DAC control section 117.
[0157] The digital circuit includes an element section that controls the DAC and an element section that controls the object of correction.
[0158] The components for controlling the DAC include a SAR control unit 118, a memory unit 124 for DAC direct control mode, and a selector 125.
[0159] The circuit includes an error detection circuit 119, a self-diagnosis circuit 120, a correction code generation circuit 121, a self-diagnosis code generation circuit 122, and a selector 123 as elements for controlling the object of correction.
[0160] The connections of the components are shown below.
[0161] The external analog signals (VINP, VINM) are input to the sample-and-hold circuit.
[0162] During normal operation, the sample-and-hold circuit 113 samples the input signal, time-discretizes it, and outputs an analog information signal, and a voltage is held at the top plate voltage nodes (topPlateP, topPlateN) of the DAC 114.
[0163] Furthermore, the sample and hold circuit 113 is also provided with a mechanism for setting the output to the midpoint reference voltage VCM during correction.
[0164] A driver circuit 115 is connected to the bottom plate side of the DAC 114, and a DAC control section 117 determines whether the high potential side reference voltage VRH or the low potential side reference voltage VRL is used.
[0165] Moreover, the DAC 114 is composed of a capacitance array of a plurality of weighted main elements, and (8C, 4C, 2C, 1C) in FIG. 2 indicate weightings of (8, 4, 2, 1), respectively, based on 1LSB of the DAC.
[0166] The top plate voltage nodes (topPlateP, topPlateN) are input to the comparator 116 that is the object of correction.
[0167] In the comparator 116, the analog information is quantized into digital information, which is then output as a quantizer output signal.
[0168] The quantizer output signal is output to a SAR control circuit 118 and an error detection circuit 119 .
[0169] The SAR control unit 118 outputs a control signal for controlling the DAC 114 to a selector 125 based on the quantizer output signal.
[0170] On the other hand, the selector 125 receives the output signal of the SAR control unit 118 and the DAC direct control mode memory unit 124, and outputs a control signal that is the basis of the DAC control code to the DAC control unit 117 in accordance with the input DAC control mode signal.
[0171] The error detection circuit 119 performs a calculation based on the quantizer output signal and compares it with a predetermined threshold value (such as an expected value, a standard value, or a reference value), and outputs an error signal.
[0172] The error signal is output to a correction code generating circuit 121 and a self-diagnosis code generating circuit 122 .
[0173] The correction code generation circuit 121 determines and outputs a correction code based on the error signal.
[0174] The self-diagnosis code generation circuit 122 determines and outputs a correction code based on the error signal.
[0175] The correction code and the self-diagnosis code are input to a selector 123 .
[0176] Based on the input operation mode signal, the selector 123 outputs either the correction code or the self-diagnosis code as a control code to the correction target.
[0177] The self-diagnosis circuit 120 receives the error signal, the self-diagnosis code and the correction code, and outputs a self-diagnosis output signal.
[0178] The operation of the components will be described below.
[0179] The operating modes are: (1) normal operating mode, (2) correction mode, and (3) inspection / self-diagnosis mode.
[0180] (1) In normal operation mode, external analog signals (VINP, VINM) are sampled (time-discretized) through the sample-and-hold circuit 113 to the DAC top plate voltage nodes (topPlateP, topPlateN).
[0181] The difference between the sampled voltage and a reference voltage generated by charge redistribution of the capacitance of the DAC 114 controlled by the DAC control unit 117 and the driver circuit 115 (called the residual voltage) is input to the comparator 116.
[0182] Comparator 116 compares the positive voltage (topPlateP) and the negative voltage (topPlateN) of the residual voltage, and outputs "1" if the positive voltage is larger, and outputs "0" if the negative voltage is larger.
[0183] In the normal operation mode, the output of the SAR control unit 118 is selected as the output signal of the selector 125 by the DAC control mode signal.
[0184] Based on the quantizer output signal, the SAR control unit 118 outputs a control signal that is the basis of the DAC control code to the DAC control unit 117, and sequentially compares the difference (residual voltage) between the analog signal and the voltage signal generated by the DAC 114, and performs quantization processing by operating so that the residual voltage approaches zero.
[0185] In this embodiment, the number of comparison cycles is 4 bits, but is not limited to this.
[0186] Furthermore, the comparator 116 may be a single comparator that processes a plurality of successive approximation cycles, or a plurality of comparators may be used to process each cycle.
[0187] Also, the quantizer output signal holds a value for each cycle by a circuit (not shown), and in this embodiment, four quantizer output signals are output, each having a 4-bit configuration.
[0188] The quantizer output from the comparator 116 is fed back to both the SAR control unit 118 and the error detection circuit 119. When background correction is performed, the offset state of the comparator is detected by the error detection circuit, the correction code is updated as needed by the correction code generation circuit 121, and is reflected in the offset control code (CCp, CCn), and AD conversion is performed in the offset-corrected state.
[0189] Next, (2) correction mode will be described.
[0190] (2) In the correction mode, there are two types: foreground correction and background correction.
[0191] During foreground correction, the AD conversion operation is stopped, and the internal midpoint reference voltage (VCM) is sampled (time discretized) through the sample-and-hold circuit 113 to the DAC top plate voltage nodes (topPlateP, topPlateN).
[0192] By AD converting the internal midpoint reference voltage (VCM), the input state of the comparator is set to a predetermined state (for example, a zero differential voltage state), and the offset of the comparator 116 is corrected so that the probability of "1" and "0" appearing in the comparator output (quantizer output signal) approaches 50%.
[0193] During foreground correction, the AD conversion operation is stopped, and then the output of the correction code generation circuit is selected as the output of the selector 123 to optimize the correction code, and an output code indicating a specific code held in the DAC direct control mode memory unit 124 is selected as the output of the selector 125 to control the DAC 114.
[0194] During background correction, while continuing the AD conversion operation, the output of the correction code generation circuit is selected as the output of the selector 123, and the correction code changes to an optimal value in response to environmental fluctuations, and the output of the SAR control unit 118 is selected as the output of the selector 125, and the same DAC control operation as in the normal operation mode is performed.
[0195] Since the main focus of this disclosure is inspection and diagnosis of the correction element, details of (2) the correction mode will be omitted.
[0196] Next, (3) Inspection / self-diagnosis mode will be explained.
[0197] Since the subject of inspection and self-diagnosis in this embodiment is the correction element itself for correcting the offset of the comparator, a comparator with an offset correction mechanism will be described before explaining the disclosure.
[0198] Fig. 4 is a circuit diagram showing an example of the comparator in Fig. 3. Comparator 130 in Fig. 4 is an example of comparator 116 in Fig. 3, and includes offset control section 132 as an offset correction mechanism. Offset control section 132 is also an example of correction circuit 109 shown in Fig. 1.
[0199] 4, the comparator 130 includes a differential amplifier section 131, an offset control section 132, and a differential latch circuit 133.
[0200] The offset control section 132 includes variable capacitances CapP and CapN for correction on the positive and negative sides of the differential, and one end of each is connected to the positive and negative output nodes Vgp and Vgn of the differential amplifier section.
[0201] The compensation elements (CapP, CapN) can vary the capacitance value according to the offset compensation amount capacitance control codes (CCp, CCn) supplied from the digital circuit.
[0202] The offset correction function of the comparator and the problems involved will be described below.
[0203] The diagram includes an offset control section 132 .
[0204] Correction control code buses (CC3p, CC2p, CC1p, CC0p, CC3n, CC2n, CC1n, CC0n) supplied from the digital circuit are supplied to the correction element driver circuit 136, which supplies VDD (power supply voltage) or VSS (GND voltage) to the correction element group.
[0205] The correction element group is composed of MOS transistors, and has a capacitance between a node where the drain and source are shorted and a node of the gate driven by the correction element driver circuit 136 .
[0206] In the case of an NMOS transistor, when the driver circuit supplies VDD, a channel is formed in the MOS, increasing the capacitance value.
[0207] On the other hand, when VSS is supplied, no channel is formed, and the capacitance value becomes a parasitic capacitance to GND, resulting in a smaller capacitance value.
[0208] The comparator 130 to be corrected has a random offset Voff mainly due to manufacturing variations in the differential pair of the differential amplifier section 131 .
[0209] In the description of this disclosure, it is necessary to define the sign of the offset for the sake of convenience, so the definition of the offset sign is set in FIG.
[0210] The same explanation can be given even if the symbols are reversed, so this is merely for convenience.
[0211] When the comparator offset 137 is zero, the comparator 130 has a positive input VINCP, a negative input VINCM, a positive output VOUTCP, and a negative output VOUTCM, and when the value of the differential input (VINCP-VINCM) is greater than 0, VOUTCP outputs "1", and in the opposite case, VOUTCP outputs "0", and VOUTCM outputs the inverted logic of VOUTCP.
[0212] When the offset Voff is due to manufacturing variations, the effective differential input of the comparator 130 is (VINCP-Voff-VINCM), and when the value of this effective differential input is greater than 0, VOUTCP outputs "1", and in the opposite case, it outputs "0".
[0213] Figure 6 shows an example of the output waveform of a comparator.
[0214] When the input clock CLKC of the comparator is "Low", the comparator is in a reset state, the internal nodes (Vgp, Vgn) become the power supply potential (VDD), and the output nodes (Voutp, Voutn) become the GND potential (VSS).
[0215] When the input clock CLKC of the comparator transitions to “High”, the comparator is put into an amplification state, the differential input signal (VINCP, VINCM) is amplified by the differential amplifier 131, and the internal nodes (Vgp, Vgn) approach VSS and have a signal gap according to the input.
[0216] In response to the opening of the signals at the internal nodes (Vgp, Vgn), a regeneration operation is performed in the differential latch circuit, transitioning to the comparator output state, and the logic of the output voltages (VOUTCP, VOUTCM) is determined.
[0217] In the example of FIG. 5, it is determined that the voltage of the input signal VINCP is greater than the input signal VINCM, and VOUTCP is output as "1" and VOUTCM is output as "0".
[0218] FIG. 7 shows an example of a waveform illustrating the effect of the offset correction capacitance.
[0219] When the variable correction capacitance CapP is increased by the control signal, the operation of pulling the Vgp node to GND during amplification becomes slower, so that there is a shift in the direction of the gap of Vgp>Vgn.
[0220] Therefore, the output VOUTCP at this time becomes "0".
[0221] On the other hand, when the variable correction capacitance CapN is increased by a control signal, the pulling operation of the Vgn node towards GND during amplification becomes slow, so it shifts in the direction with an opening of Vgp < Vgn.
[0222] Therefore, the output signal VOUTCP at this time becomes "1".
[0223] In this way, the offset is adjusted by adjusting the load balance of the differential circuit.
[0224] In order to consider the signs of the offset correction action by the above-mentioned CapP and the offset correction action by CapN, the definition in FIG. 5 is redefined.
[0225] FIG. 8 shows a convenient redefinition of the comparator offset sign including correction.
[0226] Let the offset correction amount 138 when CapP is increased be Vcalp, the offset correction amount 139 when CapN is increased be Vcaln, the comparator offset 137 be offset Voff, the differential input (VINCP - VINCM) be Vin, and the comparator output be VOUTCP.
[0227] At this time, the substantial differential input VinCP of the comparator 130 is ((VINCP - Voff - Vcalp) - (VINCM - Vcalm)).
[0228] That is, it becomes (Vin - Voff - (Vcalp - Vcalm)).
[0229] Therefore, the actions of CapP and CapN with respect to the offset have a "reverse sign" relationship.
[0230] Furthermore, in order to arrange correction capacitances of 4 bits on the positive side and 4 bits on the negative side, it has a correction action equivalent to 5 bits.
[0231] Specific examples of calculating the offset correction control code and the offset correction amount are shown in FIGS. 9 and 10.
[0232] 9 and 10, the weighting of the correction amount is set to (1.8 cubed, 1.8 squared, 1.8 first power, 1).
[0233] In addition, the offset action amount of each weight is set to (5.832, 3.24, 1.8, 1) [mV].
[0234] As shown in Figs. 9 and 10, since both CapP and CapN act as offset correction capacitances, multiple control codes can be assigned to achieve one offset correction amount, and since there is not a one-to-one relationship, correction can be achieved using any of the control code correspondence tables.
[0235] Ultimately, the sum of (Vcalp-Vcalm) acts as the offset correction amount.
[0236] FIG. 11 shows the relationship between the correction control code and the offset correction amount.
[0237] First Correspondence Table <mapping1>and the second correspondence table <mapping2>The correction amounts based on any of the above can achieve the same correction range.
[0238] As can be seen from the figure, because the weighting is 2 or less (1.8), there comes a point where the amount of correction decreases as the control code increases.
[0239] In other words, because there is redundancy, the resolution is maintained, but in some cases, multiple correction control codes may be applicable to achieve one offset correction amount (there are multiple solutions).
[0240] Therefore, due to redundancy of the correction capacitance, there are cases where the offset correction amount and the control code do not have a one-to-one relationship.
[0241] From the viewpoint of inspection and diagnosis of the correction element itself, as mentioned above, redundancy due to the lack of a one-to-one relationship becomes an issue.
[0242] Assuming that a part of the compensation element fails, or a part of the output of the compensation element's driver circuit fails, it is assumed that a signal indicating the amount of offset compensation of the comparator can be obtained through inspection and diagnosis, either directly or indirectly.
[0243] In this case, if the output indicating the amount of correction is a normal value, it is simply normal due to redundancy, and it cannot be said that all of the correction elements are normal (however, it also has the advantage that even if one fails, the redundancy will allow it to find another solution and recover).
[0244] Furthermore, when the diagnostic output is an abnormal value, as explained in the example of the control code correspondence table for the differential circuit, the correction amount and the control code state do not correspond one-to-one, so there is a problem that it is difficult to identify the location of the faulty element, i.e., whether the fault is on the positive or negative side and which correction element is faulty (hereinafter, this problem will be referred to as the "problem of concealing internal defects due to redundancy").
[0245] In addition, the issue of comparator noise will be discussed.
[0246] As analog circuits such as data converters become more miniaturized, voltages are becoming lower, and comparator noise is becoming an issue in AD conversion, correction, and inspection / diagnosis.
[0247] Figure 12 shows the effect of noise on the comparator.
[0248] In FIG. 12, the horizontal axis indicates the comparator input voltage, and the vertical axis indicates CDF (Cumulative Distribution Function), which is the probability that the comparator outputs "1" depending on the input.
[0249] When the input voltage is zero, the CDF is approximately 0.5, meaning that there is a 50% probability of outputting a "1" and a 50% probability of outputting a "0."
[0250] 12(a) and 12(b) show the simulation results when the comparator output results are repeatedly measured in order to eliminate the effects of noise.
[0251] FIG. 12(a) shows the results of 64 measurements, and FIG. 12(b) shows the results of 4096 measurements.
[0252] As can be seen from the figure, FIG. 12(a) shows that the number of comparison outputs is small, and therefore noise can cause erroneous judgment.
[0253] The effect of noise in the comparator is usually one of the causes of degradation in accuracy in AD conversion results, and can cause problems in the form of erroneous judgments in correction, inspection, and diagnosis (hereinafter, this problem will be referred to as the "problem of erroneous judgments due to noise").
[0254] In addition, the issue of nonlinearity of the comparator is discussed.
[0255] FIG. 13 shows a diagram for explaining the influence of the nonlinearity of the comparator.
[0256] The horizontal axis indicates the effective differential input VinCP of the comparator, and the vertical axis indicates CDF.
[0257] This shows an example in which several compensation control codes are selected and CDF is measured to inspect and diagnose the compensation control codes.
[0258] When selecting two adjacent control codes near the black circle comparator transition point (0.5), the values are measured as "0.4" and "0.6", respectively, making it easy to determine whether the code is normal or abnormal.
[0259] On the other hand, if two control codes are selected at points other than the white circle comparator transition points, both will output "0.0".
[0260] Therefore, it is difficult to judge the functionality, such as whether the control code for correction is functioning normally, or the accuracy, such as whether the magnitude relationship between the individual correction elements is as intended, except when the CDF is around "0.5".
[0261] Thus, due to the nonlinearity, there is a problem that only a part of the correction control code can be accurately inspected and diagnosed (hereinafter, this problem will be referred to as the "problem of state estimation due to comparator nonlinearity").
[0262] Since the object of the present disclosure is to inspect and diagnose a correction element, the analog circuit configuration is not limited thereto, but the comparator of FIG. 3 will be used as an example for explanation.
[0263] In this disclosure, the circuit diagram shown in FIG. 14 is used as the configuration of the comparator offset control unit 132 in FIG.
[0264] A conventional offset control section is shown in FIG. 48, whereas the offset control section of the present disclosure is shown in FIG. 14, but compared to the conventional one, a test element group 141 is added.
[0265] The test element group 141 is used when testing the weights of the correction element group 140, is controlled by control signals CCdp and CCdn, and has the same capacitance value as the LSB (1C) of the correction element group.
[0266] However, the size is not limited to this, and the same effect can be obtained by setting the size of the test element group to a value equal to or smaller than the LSB (1C) of the correction element group.
[0267] The procedure of the weight check according to the present disclosure will be described below based on the flow charts in FIGS. 15A and 15B, with reference to the circuit configurations in FIGS. 2 and 16, and using the tables and schematic diagrams in FIGS. 17 to 21.
[0268] 15A and 15B show a procedure (flowchart) for checking and diagnosing each weight of the offset correction element of the comparator.
[0269] FIG. 16 shows an example of the configuration of the digital section.
[0270] Figures 17 and 18 show the correction offset amount during examination and diagnosis.
[0271] 19, 20, and 21 show examples of calculation of the correction offset amount and CDF during examination and diagnosis.
[0272] FIG. 22 shows a schematic diagram of weight determination using check bits.
[0273] Step 150 begins testing and diagnosing the weights of the comparator correction elements.
[0274] In the circuit configuration of FIG. 2, a predetermined operation mode signal is supplied, the correction code generating circuit 121 is disabled, and the self-diagnosis code generating circuit 122 is enabled, and starts operation.
[0275] In step 151, a correction element to be inspected is selected.
[0276] Since the correction elements are arranged on both the positive side (P side) and the negative side (N side), they are individually selected for inspection and diagnosis.
[0277] Furthermore, which bit of the correction element (which correction element) is to be inspected is selected.
[0278] This may be a programmed sequence, selected sequentially, or selected by information held in a register.
[0279] In this embodiment, the test bit designation signal TestBit output from the state control 173 is received by the test bit selection unit 190, and information that serves as the basis for the control code for the test is output to the first fixed code generation unit and the second fixed code generation unit 183.
[0280] The correction elements are an array with a weighting set to a value of 2 or less, so that the aforementioned "problem of concealment of internal defects due to redundancy" can be solved by performing individual inspection and diagnosis aimed at each correction element.
[0281] In steps 152 and 153, a first fixed code is generated and a binary search is performed.
[0282] In the present disclosure, it is assumed that the same operation is performed for all correction bits and all bits are inspected, but as an example, a case where the MSB of the P-side correction element is inspected will be described.
[0283] The self-diagnosis code output from the first fixed code generation unit 182 is output to (CC3p, CC2p, CC1p, CC0p) (CCdp) of the P-side control code, and the value is set to a fixed value of (1, 0, 0, 0) (0).
[0284] This fixed value is stored and maintained in a first fixed code register 185 .
[0285] On the other hand, for the N-side control code, the binary search control unit 184 searches for an optimal code that minimizes the quantization error (CDF is around 0.5).
[0286] The code during the search is stored in search code register 187, and each time the code where the CDF approaches 0.5 is stored in optimum code register 188, the error from 0.5, which is the center of the CDF, is stored in minimum error register 189 and updated.
[0287] When controlling the binary search, the control signal (CCdn) of the test element on the N side is fixed to (0).
[0288] FIG. 19 shows the state of the binary search, and FIG. 17 shows the correction offset amount when the first fixed code is set.
[0289] In this embodiment, the offset Voff is taken as an example to be 4.05 mV, and the N-side correction amount Vcalm is determined and the N-side control code is determined so that the quantization error is minimized (CDF ≈ 0.5) for the sum of the offset Voff and the correction amount Vcalp for the P-side first fixed code setting (1000) (0).
[0290] The quantizer output signal is input to a subtractor 199 together with an integrated expected value register 191, and the result of the subtraction is accumulated in an accumulator 192 (accumulator).
[0291] Repeated measurements using an integrator can solve the above-mentioned "problem of erroneous judgment due to noise."
[0292] The output of the integrator 192 is supplied as an error signal to the binary search control unit 184 and also to the demultiplexer 181 .
[0293] When the binary search is completed and the optimum code is determined, the first fixed code passes through selector 180C and selector 180A and is supplied to the positive side correction element as a control signal.
[0294] Furthermore, the optimum code register 188 outputs the determined optimum code, which passes through the selectors 180D and 180B and is supplied as a control signal to the N-side correction element.
[0295] In this case, as a result holding, the demultiplexer 181 switches to STEP 2 and stores the error signal in the first accumulation result 193 .
[0296] The first accumulation result is input to CDF determination section 197 and compared with the upper limit value and lower limit value stored in CDF standard value register 195 .
[0297] FIG. 20 shows a schematic diagram of standard value determination using CDF.
[0298] The figure shows an example, but if 0.2≦CDF≦0.8 (step 154), it is determined to be normal and the process moves to the next step (155).
[0299] If it is not, (step 156), it is determined that there is an anomaly due to insufficient resolution.
[0300] Furthermore, if the offset resolution is determined to be abnormal by this CDF judgment, it is difficult to identify the location of the failure, whether it is in the P-side correction element group or the N-side correction element group, but which one is defective is not an issue at this procedural stage.
[0301] In other words, the main purpose is to bring the CDF to the vicinity of 0.5 so that an inspection of a specific bit can be performed in the next step.
[0302] In the next step, the P side is reconstructed, the N side is fixed at the optimal value, and it is seen which way the sign of the CDF deviates from 0.5. It is important in implementing this disclosure that the CDF can be brought close to 0.5 for testing.
[0303] By using this series of procedures, the offset can be moved to a position where the internal state can be estimated, as shown in FIG. 13, thereby solving the above-mentioned "problem of state estimation due to comparator nonlinearity."
[0304] In step 156, the selectors 180C and 180D and the demultiplexer 181 are switched to STEP2.
[0305] The self-diagnosis code output from the second fixed code generation unit 183 is (CC3n, CC2n, CC1n, CC0n) (CCdn), which is set to a fixed value of (0, 1, 1, 1) (1).
[0306] This fixed value is stored and maintained in a second fixed code register 186 .
[0307] On the other hand, the N-side control code is fixed to the optimum code (in this embodiment, (0, 0, 1, 0) (0)) output by the optimum code register 188.
[0308] The second fixed code is supplied to the P-side correction element through selectors 180C and 180A.
[0309] The optimum code is supplied to the N-side correction element through selectors 180D and 180B.
[0310] In this example, it is important that the N-side control code remains intact in this series of operations, and the P-side control code is reconstructed and the amount of change from around 0.5 in the CDF result is determined. Therefore, when a fault is determined, the fault location can be identified as the P-side, and since the inspection is done bit by bit, it is also possible to inspect and diagnose which bit is faulty.
[0311] Steps 157 to 160 will be explained.
[0312] The error signal output from integrator 192 (step 157 ) is provided to a second integration result register 194 .
[0313] The difference between the first and second accumulation results is calculated by a subtractor 200 and supplied to a weight determining unit 198, which determines which CDF is larger (procedure 158), and a self-diagnosis output is made (procedures 159 and 160).
[0314] In the example of this embodiment, since this is the MSB of the P-side correction element, if CDF2≦CDF1, it is determined to be normal, and otherwise, the MSB correction element is determined to have a weight of 2 or more and to be abnormal.
[0315] A schematic diagram of the series of operations is shown in supplementary FIG.
[0316] As a further supplement, FIG. 22 shows a schematic diagram of weight determination using check bits.
[0317] The N-side correction amount Vcalm is controllable, but the offset Voff due to manufacturing variations is uncontrollable (it varies randomly within a certain range).
[0318] Also, in order to check a specific bit on the P side, the bit to be checked must be fixed to "1" or "0."
[0319] Under these constraints, in this embodiment, the P side is fixed (1000 or 0111), the offset Voff is set to the offset amount caused by manufacturing variations, and the CDF is controlled to be around 0.5 by binary search control of the N side that is not inspected, thereby solving the problem.
[0320] This can be considered a preparatory operation for inspection and diagnosis in which the N-side correction amount Vcalm is adjusted by binary search against the sum of the P-side fixed value Vcalp (fixed for inspection) and the manufacturing variation offset Voff, bringing the CDF to around 0.5.
[0321] By performing the procedures based on the flowcharts in Figures 15A and 15B individually on both the P side and the N side, and by performing inspection and diagnosis on all individual bits, it is possible to ensure that the weighting of all correction elements is 2 or less based on the inspection element.
[0322] As described above, by integrating the output result of the quantizer using the integrator in the error detection circuit, the "problem of erroneous determination due to noise" can be solved.
[0323] Furthermore, by using a correction element on the polarity side that is not being tested to move the CDF to near 0.5, it is possible to determine which of the two opposing integration results is larger, which has the effect of solving the "problem of state estimation due to comparator nonlinearity."
[0324] Furthermore, the weighting of the P side and N side can be inspected and diagnosed individually based on the inspection element, and each correction element can be inspected individually to identify defects, which has the effect of solving the "problem of concealing internal defects due to redundancy."
[0325] As described above, in the semiconductor circuit of the second embodiment, the self-diagnosis code includes a first fixed code and an optimum value code, the first fixed code is a code that enables one correction element selected as a test target from the correction element group and disables the other correction elements and the test elements, the optimum value code is a code that minimizes the error signal with respect to the first fixed code and is generated by a binary search method, and the self-diagnosis circuit 105 calculates an integrated value of the error signal corresponding to the combination of the first fixed code and the optimum value code and determines whether the integrated value is within a predetermined range, and if it is not within the predetermined range, determines that the correction element group is defective.
[0326] Here, the self-diagnosis code includes a first fixed code, a second fixed code, and an optimum value code, the first fixed code is a code for enabling one correction element selected as a test target from the correction element group and disabling the other correction elements and the test element, the optimum value code is a code for minimizing an error signal with respect to the first fixed code and is generated by a binary search method, the signal to be measured is a differential signal including a first polarity signal and a second polarity signal, the correction element group includes a first element group connected to a transmission path of the first polarity signal and a second element group connected to a transmission path of the second polarity signal, the first fixed code corresponds to the first element group and disables the second element group, and the second fixed code is a code for minimizing an error signal with respect to the first fixed code and a code for minimizing an error signal with respect to the first fixed code. The optimum value code corresponds to the second element group and disables the first element group. The self-diagnosis circuit 105 calculates a first integrated value by integrating error signals corresponding to a combination of the first fixed code and the optimum value code, calculates a second integrated value by integrating error signals corresponding to a combination of the second fixed code and the optimum value code, determines which is larger than the first integrated value and the second integrated value, and determines that the correction element being tested is defective if the relative magnitude does not satisfy a predetermined condition.
[0327] Here, the self-diagnosis circuit may sequentially select one correction element as a test target from the group of correction elements.
[0328] Here, one or more of the correction elements in the group of correction elements may also be used as the inspection element.
[0329] Here, the quantizer may include one or more comparators, and the error detection circuit may generate an error signal by accumulating the output result of one comparator or the result of digitally calculating the output results of two or more comparators with a predetermined weight as the quantizer output signal.
[0330] Here, the group of correction elements includes a plurality of weighted correction elements, and the inspection element may have a weight equal to or smaller than that of a correction element corresponding to the smallest weight in the group of correction elements.
[0331] Here, the self-diagnosis code generation circuit has a first weight check code output mode and a second weight check code output mode. In the first weight check code output mode, the first bit signal controlling the first correction element to be tested is set to 1 or 0, a second correction bit signal group controlling a correction element having an adjustment weight less than the first correction element is set to an inverted logical value of the first bit signal, a check bit signal controlling the check element is set to an inverted logical value of the first bit signal, and other bit signals are output as predetermined fixed signals. In the second weight check code output mode, the first bit signal is set to an inverted logical value when the first weight check code output mode is set, the second bit signal group is set to an inverted logical value of the first bit signal, a check bit controlling the check element is set to an inverted logical value of the first bit signal, and other bit signals are output as predetermined fixed signals. During testing, the digital circuit may be set to the first weight check code output mode to generate a first error signal based on an error signal from the error detection circuit, and set to the second weight check code output mode to generate a second error signal based on the error detection circuit, and a difference between the first error signal and the second error signal is calculated to perform weighting diagnosis of the correction element group.
[0332] Here, the analog circuit is composed of a differential circuit and includes a first element group including a correction element and a test element connected to a positive input path of the differential circuit, and a second element group including a correction element and a test element connected to a negative input path of the differential circuit, the first element group and the second element group constitute a correction element group and a test element group, and the self-diagnosis code generation circuit has a first set test mode for testing the first element group and a second set test mode for testing the second element group, and may perform testing of the first element group and the second element group individually.
[0333] Here, the analog circuit has a redundant configuration including a first element group including a correction element group and a test element group belonging to one node, and a second element group having the same weight and configuration as the first element group, and the self-diagnosis code generation circuit has a first set test mode for testing the first element group and a second set test mode for testing the second element group, and may test the first element group and the second element group individually.
[0334] Here, in the first set test mode, the self-diagnosis code generation circuit outputs, for the first element group, a first bit signal controlling a correction element to be tested as 1 or 0, a second correction bit signal group controlling a correction element having an adjustment weight less than the first correction element as an inverted logical value of the first bit signal, a test bit signal controlling a test element as an inverted logical value of the first bit signal, and other bit signals as predetermined fixed signals, and outputs, for the second element group, an optimal correction code so as to minimize the quantization error of the quantizer, and the digital circuit generates a first error signal based on the error signal of the error detection circuit in the first set test mode, and outputs, for the control code of the first element group, an optimal correction code so as to minimize the quantization error of the quantizer, the second bit signal is set to the inverted logical value of the bit signal at the first step, the second bit signal group is set to the inverted logical value of the bit signal at the first step, the test bit controlling the test element is set to the inverted logical value of the bit signal at the first step, and the other bit signals are output as predetermined fixed signals, while an optimal correction code is maintained and output to the second element group, a second error signal is generated as the output of the error detection circuit, and at the transition point of the quantizer where the quantization error is minimized, the magnitude of the first error signal and the second error signal are judged to be smaller than the specified value to check whether the weight is equal to or smaller than the specified value, and in the second set test mode, the relationship between the first element group and the second element group in the first set test mode may be reversed to perform the same processing.
[0335] Here, the data converter may be an analog-to-digital converter.
[0336] Here, the analog-to-digital converter includes one or more comparators, each of which is configured by a differential circuit, and includes a correction element group and a test element group in each of a positive transmission path and a negative transmission path in the differential circuit in order to correct an offset of the comparator, and the correction elements included in the correction element groups of the positive transmission path and the negative transmission path may be individually tested based on the test element group.
[0337] <Third embodiment> The circuit configuration and operation of the third embodiment will be described below with reference to the drawings.
[0338] This paper describes the inspection and diagnosis of the offset correction element of a comparator.
[0339] The configuration will be described below.
[0340] In the third embodiment, similarly to the second embodiment, a configuration example will be described in which a successive AD converter (FIG. 2) is taken as an example and the correction range of a group of correction elements that perform offset correction of a comparator is inspected and diagnosed.
[0341] The configuration is the same as in the second embodiment, but only the digital section is different.
[0342] One issue with the second embodiment is that the relative accuracy between correction elements can be confirmed by comparing the weights between adjacent elements (for example, between control codes (1000) and (0111)) based on the test element. However, when the test element itself is broken, the range may be reduced or may be too large.
[0343] In the second embodiment, for example, if the test element is greater than 1 LSB of the correction element and passes the test and diagnosis (is determined to be normal), the range becomes large and the resolution becomes coarser than assumed by design.
[0344] Furthermore, if the test element is smaller than 1 LSB of the correction element and passes the test and diagnosis, the resolution is sufficiently fine but the range is smaller than assumed by the design, and the range is narrow and insufficient to correct the offset of the manufacturing variation.
[0345] Since the object of the present disclosure is to inspect and diagnose a correction element, the analog circuit configuration is not limited thereto, but the comparator of FIG. 3 will be used as an example for explanation.
[0346] In this disclosure, the circuit diagram shown in FIG. 14 is used as the configuration of the comparator offset control unit 132 in FIG.
[0347] Hereinafter, the procedure of the range check according to the present disclosure will be described based on the flow charts of FIGS. 23A and 23B, with reference to the circuit configurations of FIGS. 2 and 24, and using the schematic diagram of FIG.
[0348] To start with, the gist is that two types of voltages (±Δ) are generated using the DAC inside the successive approximation ADC, and the correction element inside the comparator is controlled to correct these voltages. Two types of control codes are then stored in a register and compared, making it possible to check and diagnose the absolute amount of correction, i.e., how many codes the generated voltage corresponds to.
[0349] In step 211, a decision is made as to whether to test the P-side or N-side correction elements.
[0350] In this embodiment, as an example, range inspection and diagnosis of the P side will be described.
[0351] In steps 212 and 213, the P-side correction element control code CCp_TEST is output as a fixed code ((1000)(0), (mmmm)(n), where m is the bit signal of each correction element and n is the bit signal for the test element) output from the MSB fixed code register 235 through selectors 234C and 234A.
[0352] On the other hand, for the N-side correction element control code CCn_TEST, the binary search control unit 247 receives the error signal, updates the output of the search code register 238 to minimize the quantization error (CDF is around 0.5), stores the optimal code in the optimal code register 244, and stores the minimum value of the error signal in the minimum error register 241.
[0353] Procedures 214, 215, and 216 are procedures for determining whether the CDF is near 0.5 and determining whether the resolution is normal or abnormal, as in the embodiment, and can be omitted. For ease of explanation, this part is omitted in FIG.
[0354] In steps 217 and 218 , a control code for applying, for example, the equivalent of +3.0 mV (+Δ) to the comparator inputs (topPlateP, topPlateN) is selected from the DAC direct control mode memory section and output through selectors 254 and 255 .
[0355] The P-side control code is controlled using the binary search control unit 245 and the search code register 236 so as to offset this voltage.
[0356] The P-side control code that minimizes the quantization error (CDF is close to 0.5) is held in the optimum code register 242 and also stored in the first code result register 248.
[0357] At this time, for the N-side control code, the output of the optimum code register 244 is outputted through the selectors 234D and 235B.
[0358] In steps 219 and 220 , a control code for applying, for example, the equivalent of −3.0 mV (−Δ) to the comparator inputs (topPlateP, topPlateN) is selected from the DAC direct control mode memory section and output through selectors 254 and 255 .
[0359] The P-side control code is controlled using the binary search control unit 246 and the search code register 237 so as to offset this voltage.
[0360] The P-side control code that minimizes the quantization error (CDF is close to 0.5) is held in the optimum code register 243 and also stored in the first code result register 249.
[0361] At this time, for the N-side control code, the output of the optimum code register 244 is outputted through the selectors 234D and 235B.
[0362] In step 221, the control code difference between the first code result and the second code result is calculated, and the range determination unit 251 determines whether the difference value is between the minimum allowable value RMIN and the maximum allowable value RMAX held in the range standard value register 250.
[0363] If the range is normal, the process moves to steps 222 and 223.
[0364] If the range is not good, the process moves to step 224, where a range error signal is output as a self-diagnosis output signal.
[0365] FIG. 25 shows a schematic diagram of this embodiment.
[0366] When a test voltage +Δ is applied, the P-side control code becomes ((1011)(0)).
[0367] On the other hand, when a test voltage -Δ is applied, the P-side control code becomes ((0011)(0)).
[0368] At this time, the code difference is equivalent to (1000) and is determined to be normal.
[0369] In this example, a voltage equivalent to half the MSB of the correction range is generated by the DAC, and the control codes equivalent to MSB ±(MSB / 2) are determined by a binary search. By taking the code difference, the correspondence between the applied voltage and the control code can be determined, making it possible to inspect and diagnose the absolute quantity of the control code.
[0370] Therefore, by combining the inspection and diagnosis of the relative accuracy in the second embodiment with the inspection and diagnosis of the absolute value in the third embodiment, it is possible to have the effect of being able to inspect and diagnose both the resolution and range of the correction element.
[0371] As described above, in the semiconductor circuit according to the third embodiment, the data converter includes an input switching unit which switches whether a signal input to the data converter is a measured signal or a reference signal, the self-diagnosis code generation circuit has an optimum correction code output mode which outputs an optimum correction code which minimizes a quantization error, and the digital circuit is set to the optimum correction code output mode during testing, selects a first reference signal by the input switching unit, obtains an optimum correction code as the first optimum correction code, is set to the optimum correction code output mode, selects a second reference signal by the input switching unit, obtains an optimum correction code as the second optimum correction code, calculates a difference between the first optimum correction code and the second optimum correction code, and performs correction range diagnosis of the correction element group.
[0372] <Fourth embodiment> The circuit configuration and operation of the fourth embodiment will be described below with reference to the drawings.
[0373] This paper describes the inspection and diagnosis of the offset correction element of a comparator.
[0374] The configuration will be described below.
[0375] In the fourth embodiment, as in the second and third embodiments, a configuration example will be described in which a successive AD converter (FIG. 2) is taken as an example and the functions of a group of correction elements that perform offset correction of a comparator are inspected and diagnosed, but only the digital section is different.
[0376] The method of inspecting the weight of the correction element shown in the second embodiment has a problem of time since it is necessary to increase the number of repeated measurements in order to remove noise in order to check the accuracy of the correction element.
[0377] In contrast to this, the third embodiment provides a simpler method for removing obvious initial defects.
[0378] Since the object of the present disclosure is to inspect and diagnose correction elements, the analog circuit configuration is not limited thereto, but the comparator in FIG. 3 will be used as an example for explanation.
[0379] In this disclosure, the circuit diagram shown in FIG. 14 is used as the configuration of the comparator offset control unit 132 in FIG.
[0380] Hereinafter, the procedure of the functional inspection of the present disclosure will be described based on the flow chart of FIG. 26, with reference to the circuit configurations of FIG. 2 and FIG. 27, and using the schematic diagram of FIG.
[0381] To start with, to give an overview, the comparator decision point is brought to near the minimum quantization error (CDF near 0.5) where the internal state can be estimated by binary search, and each bit of the control code (correction code) at that point is inverted. The CDF will deviate from 0.5 by a certain amount, but because the direction of the deviation is known, it is possible to detect failures in the driver circuit inside the analog circuit or failures in the device (no reaction, offset correction not working for the intended code, etc.) (a schematic diagram is shown in Figure 28).
[0382] In steps 261, 262, and 263, the selectors 284C and 284D and the demultiplexer 295 are switched to STEP1.
[0383] In this embodiment, a code mapping table that combines the P side and the N side is used.
[0384] As an example, the mapping table in FIG. 10 is used.
[0385] Since there are 4 bits on the P side and 4 bits on the N side, it can be managed with a control signal equivalent to 5 bits.
[0386] The device is provided with a code mapping unit (290A, 290B) that converts this 5-bit equivalent signal into a positive / negative 4-bit equivalent signal.
[0387] The binary search control unit 293 and search code register 291 that receive the error signal, the code mapping units 290A and 290B, and the positive side search code register 286 and the negative side search code register 287 that hold the positive and negative search codes execute the binary search and store the results, and a control code is determined to correct the manufacturing variation Voff (this procedure is the same as the correction execution block in the correction code generation unit, so functional blocks can be shared).
[0388] Moreover, the error signal is stored in the first accumulation result register 297 and is set as the first CDF result (CDF1).
[0389] In step 264, one bit (MSB to LSB and test element) of which polarity (P side, N side) is to be tested and diagnosed is selected.
[0390] The check bit selection unit 285 outputs the selected bit information to be checked as a bit inversion signal to the bit inversion units 301A and 301B.
[0391] In steps 265 and 266, the selectors 284C and 285B and the demultiplexer 295 are switched to STEP 2, and the CDF is measured when the bit to be inspected is inverted.
[0392] At this time, the output from the optimum code register 292, which holds the optimum code obtained by the binary search, is supplied to the positive optimum code register 288 and the negative optimum code register 289 through the code mapping unit 290B, and is supplied to the control code of the comparator through the selectors 284C and 284D.
[0393] The CDF result passes through a demultiplexer 295 and is stored in a second accumulation result register 298 as a second CDF result (CDF2).
[0394] Self-diagnosis is performed in steps 267, 268, and 269.
[0395] The output of the first accumulation result register 297 and the second accumulation result register 298 are output to a subtractor 296 , and the subtraction result is supplied to a toggle determination section 300 .
[0396] A toggle determination unit 300 receives the output of the toggle standard value register 299 and determines whether the sign relationship between CDF1 and CDF2 is in the correct direction.
[0397] We provide additional explanation on the sign relationship between CDF1 and CDF2.
[0398] In this embodiment, Vin is 0, Vcalp and Vcalm are correction amounts, and Voff is a manufacturing offset, so that the substantial comparator input VinCP is (-Voff-Vcalp+Vcalm).
[0399] Here, Vcalp and Vcalm have an "opposite sign relationship" as an effect.
[0400] In other words, when any of the P-side correction bits transitions from 0 to 1, VinCP decreases, so CDF2 should decrease by a certain amount (go below 0.5), and when it transitions from 1 to 0, VinCP increases, so CDF2 should increase by a certain amount (go above 0.5).
[0401] In contrast, when any of the N-side correction bits transitions from 0 to 1, VinCP increases, so CDF2 should rise by a certain amount (go above 0.5), and when it transitions from 1 to 0, VinCP decreases, so CDF2 should fall by a certain amount (go below 0.5).
[0402] As can be seen from this, by inputting information indicating which polarity of the comparator is being checked (PosNegFlag) to the toggle determination section, it is possible to determine whether an abnormality has occurred.
[0403] The toggle standard value is set with a predetermined offset (threshold) in the expected sign direction from 0.5, and by observing the change in CDF2, it is possible to determine whether the driver circuit that drives the correction capacitance and the correction capacitance element are functioning normally.
[0404] In this embodiment, two values separated by a predetermined amount are compared and judged, so instead of an accuracy judgement as in the second embodiment, the number of repeated measurements can be reduced, and by performing this test first, the test time can be shortened.
[0405] Figure 29 shows an example of an examination and diagnosis sequence.
[0406] In step 311, the toggle check and diagnosis (FIGS. 26 and 27) described in the fourth embodiment is executed.
[0407] The measurement time can be shortened compared to weight inspection, and any defective products in step 311 can be removed as defective products.
[0408] Procedure 312 (FIGS. 15A and 15B, FIG. 16) is a weighting test of the individual correction elements described in the second embodiment, and procedure 313 (FIGS. 23A and 23B, 24) is a test of the correction range described in the fourth embodiment.
[0409] Through this series of inspections, we can guarantee both the functionality and accuracy of each correction element.
[0410] As described above, in the semiconductor circuit according to the fourth embodiment, the self-diagnosis code generation circuit has an optimal correction code output mode in which, upon receiving an error signal, it outputs an optimal correction code so as to minimize the quantization error of the quantizer, and a logic inversion code output mode in which it outputs a logic inversion code obtained by logically inverting a first bit signal of the optimal correction code that controls the correction element being tested, and during testing, the digital circuit is set to the optimal correction code output mode, supplies the optimal correction code to the analog circuit, generates a first error signal based on the error signal from the error detection circuit, and is set to the logic inversion code output mode, supplies the logic inversion code to the analog circuit, generates a second error signal based on the error signal from the error detection circuit, compares the magnitude relationship between the first error signal and the second error signal, and performs fault diagnosis of the correction element being tested.
[0411] <Fifth embodiment> The circuit configuration and operation of the fifth embodiment will be described below with reference to the drawings.
[0412] This paper describes the inspection and diagnosis of the gain correction element of the residual amplifier.
[0413] In the fifth embodiment, a configuration example in which a gain correction element of a residual amplifier is to be inspected and diagnosed will be described by taking a pipeline AD converter 320 as an example.
[0414] The configuration will be described below.
[0415] 30, the analog section is made up of a sample-and-hold circuit 323, a first sub-ADC 324, a residual amplifier 325, and a second sub-ADC 326.
[0416] The sample-and-hold circuit 323 samples the differential signals (VINP, VINM) during normal operation, and samples the midpoint reference voltage VCM (corresponding to a zero differential signal) during correction, inspection, and diagnosis.
[0417] The first sub-ADC 324 is configured, for example, by a successive approximation ADC as shown in FIG. 2. During normal operation, the first sub-ADC 324 quantizes the sampled voltages (VINP, VINM) and holds the residual voltage, and during correction, inspection, and diagnosis, for a differential signal of zero, the first sub-ADC 324 controls the internal DAC voltage in accordance with the code stored in the DAC direct control mode memory unit 332 to generate a predetermined voltage and hold it as a residual.
[0418] The residual amplifier 325 amplifies the residual voltage held by the first sub-ADC 324 by a predetermined factor after the first sub-ADC completes AD conversion.
[0419] The second sub-ADC 326 is, for example, a successive approximation ADC, and quantizes the voltage obtained by the residue amplifier output.
[0420] In a pipeline ADC, a high-power amplifier is required to maintain the accuracy of the magnification (gain) of the residual amplifier. However, with the progress of low-voltage and correction technologies, an open-loop integrating amplifier as shown in Non-Patent Document 3 is being used, and progress is being made in reducing power consumption.
[0421] However, the open-loop integrating amplifier has a large variation in gain due to manufacturing variations, but has low power consumption.
[0422] Therefore, the gain variation of the residual amplifier is corrected using digital circuitry, achieving both low power consumption and high accuracy.
[0423] A representative example of the correction technique is the method of gain estimation using dither described in Non-Patent Document 1.
[0424] The internal DAC of the first sub-ADC is directly controlled to add a predetermined random voltage to the residual voltage, and the internal DAC of the second sub-ADC is used to subtract an amount equivalent to the predetermined random voltage from the voltage amplified by the residual amplifier.
[0425] This operation causes the gain error to remain in the residual voltage held by the second sub-ADC, and the gain error information mixed into the conversion result of the second sub-ADC is processed to estimate and correct the gain error of the residue amplifier.
[0426] FIG. 31 is a circuit diagram showing an example of the residual amplifier 325 with a gain correction function.
[0427] The residual amplifier 325 includes a signal amplifier unit 350 and an amplification time adjustment unit 351 .
[0428] In the case of an open-loop integrating amplifier, the longer the amplification time, the greater the gain can be, so the gain is adjusted by adjusting the amplification time.
[0429] FIG. 32 shows an example of the amplification time adjustment unit 351 used in the present disclosure.
[0430] The amplification time adjustment unit 351 includes a correction element group 354 and a test element group 355 .
[0431] The correction element group 354 is composed of MOS capacitance elements having a weighting of 2 or less (≦2), and the test element group 355 is composed of MOS capacitance elements having a weighting equal to the LSB of the correction element group, for example.
[0432] In the present disclosure, the size of the test element group is not limited to the above-mentioned size, and the same effect can be obtained as long as it is 1 LSB or less of the correction element group.
[0433] The correction elements are controlled by control codes (CG3p, CG2p, CG1p, CG0p), and the inspection elements are controlled by a control code (CGdp).
[0434] In the following, when the bit signal of the control code is "1", the capacitive element is enabled and the gain increases, and when the bit signal of the control code is "0", the capacitive element is disabled and the gain decreases.
[0435] The control codes are written as follows: (CG3p, CG2p, CG1p, CG0p) (CGdp) should be written as (xxxx) (y).
[0436] The operation will be explained below.
[0437] The weights of the individual correction elements in the correction element group are inspected and diagnosed using a procedure similar to the inspection of the weights of the correction elements of the comparator shown in FIGS. 15A and 15B.
[0438] That is, the sampling voltage is VCM, and in the first procedure as an example of an MSB correction element test, (CG3p, CG2p, CG1p, CG0p) (CGdp) is set to (1, 0, 0, 0) (0) by the self-diagnosis code generation circuit 336, and the AD conversion result (multiple-bit quantizer output result) of the second sub-ADC 326 is obtained and accumulated using the same gain error measurement method as the dither method, and is determined as the first accumulation result.
[0439] Furthermore, in a second procedure, the self-diagnosis code generation circuit 336 sets (CG3p, CG2p, CG1p, CG0p) (CGdp) to (0, 1, 1, 1) (1), and the AD conversion result of the second sub-ADC 326 is obtained using the same gain measurement method as the dither method, and is accumulated to obtain a second accumulation result.
[0440] The first integration result is compared with the second integration result to measure the weight error of the gain correction element and make a judgment by the self-diagnosis circuit.
[0441] The weightings below the MSB can be diagnosed and inspected in the same way.
[0442] It should be noted here that because the inspection and diagnosis are not performed near the optimal gain setting (the gain setting at which the first and second sub-ADCs are connected while maintaining linearity), the CDF of the downstream comparator is not necessarily near 0.5, and AD conversion is required to measure the error for inspection and diagnosis.
[0443] That is, the residual voltage mixed with the gain error may be larger than 1 LSB of the second sub-ADC in some cases, and AD conversion (quantization) using multiple comparators is required for weighted measurement.
[0444] Whereas in the case of comparator offset correction element weight inspection, only one quantizer output result (the circuit of the comparator itself to be corrected) is used, in this embodiment of the amplifier gain correction element weight inspection, inspection and diagnosis are performed based on the accumulation result of multiple quantizer output results of the second sub-ADC.
[0445] In the case of residual amplifier gain error measurement, the comparator weight measurement procedure (procedure 153 in FIG. 15A and procedure 157 in FIG. 15B) is replaced with storing the integrated value of the AD conversion results (output results of multiple quantizers) of the second sub-ADC in a register, thereby enabling inspection and diagnosis of the residual amplifier gain correction element by a procedure similar to that of FIG. 15A and FIG. 15B.
[0446] As described above, by providing a test element in addition to the gain correction element of the residual amplifier, generating an input code for test and diagnosis by the self-diagnosis code generation circuit, and utilizing the integrated value of the AD conversion result of the sub-ADC in the subsequent stage of the pipeline, it is possible to have the effect of testing and diagnosing the weighting of each gain correction element.
[0447] As described above, in the semiconductor circuit according to the fifth embodiment, the self-diagnosis code generation circuit has a first weight check code output mode and a second weight check code output mode. In the first weight check code output mode, the first bit signal for controlling the first correction element to be tested is set to 1 or 0, a second correction bit signal group for controlling correction elements having adjustment weights less than the first correction element is set to the inverted logical value of the first bit signal, a check bit signal for controlling the check element is set to the inverted logical value of the first bit signal, and other bit signals are output as predetermined fixed signals. In the second weight check code output mode, The bit signal is set to the inverted logical value when the first weighted check code output mode is set, the second bit signal group is set to the inverted logical value of the first bit signal, the check bit controlling the check element is set to the inverted logical value of the first bit signal, and the other bit signals are output as predetermined fixed signals, and the digital circuit is set to the first weighted check code output mode during testing to generate a first error signal based on the error signal from the error detection circuit, and is set to the second weighted check code output mode to generate a second error signal based on the error signal from the error detection circuit, calculates the difference between the first error signal and the second error signal, and performs weighting diagnosis of the correction element group.
[0448] Sixth embodiment The circuit configuration and operation of the sixth embodiment will be described below with reference to the drawings.
[0449] This paper describes the inspection and diagnosis of the gain correction element of the residual amplifier.
[0450] The configuration will be described below.
[0451] As the circuit configuration, the block diagram (FIG. 30) and the residual amplifier configuration (FIG. 31) of the fifth embodiment are used, and as the amplification time adjustment section, the amplification time adjustment section of FIG. 33 is used.
[0452] The amplification time adjustment section 351 is composed of a first set element group 358 and a second set element group 359, each of which includes a correction element group (354, 356) and a test element group (355, 357).
[0453] The first set group of elements 358 is controlled by a control code (CG3p, CG2p, CG1p, CG0p) (CGdp), and the second set group of elements 359 is controlled by a control code (CG3n, CG2n, CG1n, CG0n) (CGdn).
[0454] The operation will be explained below.
[0455] The weights of the individual correction elements in the correction element group are inspected and diagnosed using a procedure similar to the inspection of the weights of the correction elements of the comparator shown in FIGS. 15A and 15B.
[0456] In this embodiment, since there are two sets, the first set element group 358 and the second set element group 359, inspection and diagnosis can be performed with optimal gain settings.
[0457] For example, when inspecting the MSB of the first set of elements, as a first procedure, the control code (CG3p, CG2p, CG1p, CG0p) (CGdp) is set to (1, 0, 0, 0) (0), and the control code (CG3n, CG2n, CG1n, CG0n) (CGdn) is set to the code (y, y, y, y) (0) obtained by searching for the optimal value (optimum gain setting) through a binary search.
[0458] As a second procedure, the control code (CG3p, CG2p, CG1p, CG0p) (CGdp) is set to (1, 0, 0, 0) (0), the control code (CG3n, CG2n, CG1n, CG0n) (CGdn) is set to (y, y, y, y) (0), and the initial quantizer output result of the second sub-ADC326 is accumulated to obtain the first accumulation result.
[0459] As a third procedure, the control code (CG3p, CG2p, CG1p, CG0p) (CGdp) is set to (0, 1, 1, 1) (1), the control code (CG3n, CG2n, CG1n, CG0n) (CGdn) is set to (y, y, y, y) (0), and the initial quantizer output result of the second sub-ADC326 is accumulated to obtain the second accumulation result.
[0460] By comparing the first integration result with the second integration result, the weighting of the correction element of the amplifier gain element can be checked.
[0461] Compared to the fifth embodiment, the sixth embodiment is capable of performing weighted measurements near the optimal gain, and therefore requires only one quantizer output (1-bit processing), making it possible to reduce the scale of the inspection and diagnosis circuitry.
[0462] As described above, in the semiconductor circuit according to the sixth embodiment, the self-diagnosis code generation circuit has a first weight check code output mode and a second weight check code output mode. In the first weight check code output mode, the first bit signal for controlling the first correction element to be tested is set to 1 or 0, a second correction bit signal group for controlling correction elements having adjustment weights less than the first correction element is set to the inverted logical value of the first bit signal, a check bit signal for controlling the check element is set to the inverted logical value of the first bit signal, and other bit signals are set to predetermined fixed signals. In the second weight check code output mode, The bit signal is set to the inverted logical value when the first weighted check code output mode is set, the second bit signal group is set to the inverted logical value of the first bit signal, the check bit controlling the check element is set to the inverted logical value of the first bit signal, and the other bit signals are output as predetermined fixed signals, and the digital circuit is set to the first weighted check code output mode during testing to generate a first error signal based on the error signal from the error detection circuit, and is set to the second weighted check code output mode to generate a second error signal based on the error signal from the error detection circuit, calculates the difference between the first error signal and the second error signal, and performs weighting diagnosis of the correction element group.
[0463] Here, the analog circuit is composed of a differential circuit and includes a first element group including a correction element and a test element connected to a positive input path of the differential circuit, and a second element group including a correction element and a test element connected to a negative input path of the differential circuit, the first element group and the second element group constitute a correction element group and a test element group, and the self-diagnosis code generation circuit has a first set test mode for testing the first element group and a second set test mode for testing the second element group, and may perform testing of the first element group and the second element group individually.
[0464] <Seventh embodiment> The circuit configuration and operation of the seventh embodiment will be described below with reference to the drawings.
[0465] This paper describes the inspection and diagnosis of the offset correction element of the residue amplifier.
[0466] As the circuit configuration, the signal amplifier shown in FIG. 34 obtained by modifying the signal amplifier of the block diagram (FIG. 30) and the residual amplifier configuration (FIG. 31), and the offset control section of the residual amplifier shown in FIG. 35 are used.
[0467] In the signal amplifier shown in FIG. 34, the NMOS transistor is variably controlled for offset correction.
[0468] That is, it has a differential pair 371 with an offset adjustment function.
[0469] In the offset control section shown in FIG. 35, the switches of the test element groups 374 and 375 and the correction element groups 372 and 373 are selected by a driver signal (not shown).
[0470] The offset is corrected and controlled by adjusting the positive / negative balance of the size of the differential pair.
[0471] In this circuit, since it is provided with test element groups 374 and 375, the same procedure as in the second embodiment is used, and the weighting of each correction element can be tested and diagnosed by using the integrated value of the conversion result of the second sub-ADC 326.
[0472] In order to adjust the offset, other than the method of changing the balance of the MOS of the differential pair used as shown in FIGS. 34 and 35, the offset may be adjusted by adding a variable capacitance to the differential pair output of FIG. 34, or the back gate of the MOS may be adjusted. Regardless of which offset adjustment method is used, the inspection and diagnosis circuit of this embodiment can inspect the weight of the correction element.
[0473] As described above, in the semiconductor circuit according to the seventh embodiment, the analog-to-digital converter is a pipeline analog-to-digital converter and includes a residue amplifier and one or more comparators, the residue amplifier is configured as an open-loop integrating amplifier, and as a circuit configuration for adjusting the offset of the residue amplifier, a group of correction elements is provided on each of the positive and negative sides of the internal output of the differential circuit, to adjust the offset by positive / negative balance, and the group of correction elements is inspected using the integration result of the comparator.
[0474] <Eighth embodiment> The circuit configuration and operation of the eighth embodiment will be described below with reference to the drawings.
[0475] This paper describes the inspection and diagnosis of the DAC capacitance correction element for correcting the linearity of the DAC.
[0476] The circuit configuration is shown in the block diagram of FIG. 36, and the correction section is shown in the circuit of FIG.
[0477] The analog circuit block configuration is the same as in the second embodiment, but the DAC portion is different.
[0478] The MSB capacitance (397A, 397B) of the DAC is set to a value (8C-Δ) obtained by subtracting Δ from the expected capacitance value, and correction capacitance (398A, 398B) is provided that can perform switching in conjunction with the MSB capacitance.
[0479] The correction capacitances (398A, 398B) have the configuration shown in FIG. 37, and by taking the NAND logic of the signal supplied from the DAC control unit and the control codes (CD3p, CD2p, CD1p, CD0p) (CDdp), (CD3n, CD2n, CD1n, CD0n) (CDdn), the correction capacitances are switched in conjunction with each other in normal DAC operation.
[0480] By selecting a large correction capacitance (398A, 398B), the apparent MSB capacitance increases, the step voltage of the MSB increases, and the linearity can be adjusted.
[0481] This correction method can also be applied to the MSB and below of the DAC.
[0482] As for the inspection and diagnosis method, the weighting inspection and diagnosis can be performed in a procedure similar to that of the comparator weighting inspection and diagnosis in the second embodiment and the amplifier gain weighting inspection and diagnosis method in the fifth embodiment.
[0483] Specifically, when inspecting the MSB of the P-side correction element, a first accumulation result is obtained by setting (CD3p, CD2p, CD1p, CD0p) (CDdp) to (1, 0, 0, 0) (0) and (CD3n, CD2n, CD1n, CD0n) (CDdn) to the optimum value (the transition point where the quantization error of the comparator is minimized) determined by a binary search, and a second accumulation result is obtained by setting CD3p, CD2p, CD1p, CD0p) (CDdp) to (0, 1, 1, 1) (1) and (CD3n, CD2n, CD1n, CD0n) (CDdn) to the optimum value, and by comparing the first accumulation result and the second accumulation result, it is possible to perform relative weighting inspection and diagnosis between the correction elements based on the inspection element.
[0484] As described above, in the semiconductor circuit according to the eighth embodiment, the analog / digital converter is a successive approximation type analog / digital converter, and includes an internal digital / analog converter having a differential circuit configuration including a plurality of capacitors (M pieces) provided in each of a positive-side transmission path and a negative-side transmission path, and one or more comparators. The correction element group and the inspection element group for correcting the linearity of N (N<M) of the plurality of capacitors are provided, and the inspection of the correction elements for the internal digital / analog converter is individually performed for the positive-side correction element group and the negative-side correction element group using the integration result of the comparator.
[0485] <Ninth Embodiment> Hereinafter, the circuit configuration and operation of the ninth embodiment will be described with reference to the drawings.
[0486] The inspection and diagnosis of the DAC gain correction element for correcting the ADC gain will be described.
[0487] As the circuit configuration, the block diagram of FIG. 38 and the circuit of FIG. 39 are used as the correction unit.
[0488] The analog circuit block configuration is the same as that of the second embodiment, but the DAC part is different.
[0489] Correction capacitors (430A, 430B) for adjusting the DAC gain are provided for the binary capacitor arrays (427A, 428A, 429A, 427B, 428B, 429B) of the DAC.
[0490] The correction capacitors (430A, 430B) have the configuration shown in FIG. 39 and are controlled by signals supplied from the DAC control unit and control codes (CE3p, CE2p, CE1p, CE0p) (CEdp), (CE3n, CE2n, CE1n, CE0n) (CEdn).
[0491] In the normal DAC operation, since this correction capacitor is a capacitor that is constantly connected to the VCM potential, the output voltage of the DAC decreases as the connected correction capacitor increases, and thus the ADC gain increases.
[0492] As inspection and diagnosis methods, weight-based inspection and diagnosis of the comparator in the second embodiment, and weight-based inspection and diagnosis methods for amplifier gain in the fifth embodiment can be performed in the same procedure for weight-based inspection and diagnosis.
[0493] As described above, in the semiconductor circuit according to the ninth embodiment, the analog-to-digital converter is a successive approximation type analog-to-digital converter, and includes an internal digital-to-analog converter having a differential circuit configuration including a plurality of capacitors (M pieces) in each of a positive-side transmission path and a negative-side transmission path, and one or more comparators. The correction element group is for gain correction for N (N < M) of the plurality of capacitors, and the inspection of the correction elements for the internal digital-to-analog converter is individually performed for the positive-side transmission path and the negative-side transmission path using the integration result of the comparator.
[0494] <Tenth Embodiment> Hereinafter, the circuit configuration and operation of the tenth embodiment will be described with reference to the drawings.
[0495] Inspection and diagnosis of correction elements for delay control in a TD converter (time-to-digital converter) will be described.
[0496] This disclosure states that it can be applied not only to an AD converter that compares an analog with a predetermined voltage and quantizes it, but also to inspection and diagnosis of correction elements of a TD converter that compares an input clock with a reference clock and performs time quantization.
[0497] As a circuit configuration, an example in which the TD converter shown in Patent Document 6 is modified is shown.
[0498] The basic operations and functions are the same as those in Patent Document 6, but the functional blocks for performing inspection and diagnosis are different.
[0499] The circuit block of this disclosure is shown in FIG. 40, and the variable delay control unit is shown in FIG. 41.
[0500] In this disclosure, a group of test elements 451 is added to the variable delay control units (DL0, DL1, DL2, DL3), and a block for correction, test, and diagnosis is added, which has a selector 441, a correction code generation circuit 442, a self-diagnosis code generation circuit 443, an error detection circuit 444, and a self-diagnosis circuit 445.
[0501] As shown in FIG. 41, the variable delay control section includes a correction element group 450 and a test element group 451, and adjusts the amount of bias current with a control code (CT3, CT2, CT1, CT0) (CTd) to control the amount of delay.
[0502] The bias current source is provided by a MOS transistor.
[0503] As for the inspection and diagnosis method, the weighting inspection and diagnosis can be performed in a procedure similar to that of the comparator weighting inspection and diagnosis in the second embodiment and the amplifier gain weighting inspection and diagnosis method in the fifth embodiment.
[0504] Specifically, when inspecting the MSB of the correction element, (CT3, CT2, CT1, CT0) (CTd) is set to (1, 0, 0, 0) (0), the quantized information obtained from the flip-flops (FF0, FF1, FF2, FF3) is accumulated to obtain a first accumulation result, (CT3, CT2, CT1, CT0) (CTd) is set to (0, 1, 1, 1) (1), the quantized information obtained from the flip-flops (FF0, FF1, FF2, FF3) is accumulated to obtain a second accumulation result, and by comparing the first accumulation result and the second accumulation result, relative weighting inspection and diagnosis between the correction elements based on the inspection element can be performed.
[0505] As described above, in the semiconductor circuit according to the tenth embodiment, the signal under test is a pulse signal that takes the value of 1 or 0, the data converter is a time-to-digital converter that compares the phase difference between the pulse signal and a reference clock signal and converts it into a digital signal, and the group of correction elements corrects the transition timing of the pulse signal between 1 and 0.
[0506] Here, the time-to-digital converter includes a multi-stage delay circuit with a controllable delay amount, a group of correction elements that correct the delay amount of the delay circuit, a group of test elements that test the group of correction elements, a group of flip-flops that compare the pulse signal with the delay stage output by capturing the logic of each output of the delay circuit in accordance with the edge timing of the pulse signal, a counter that obtains an accumulation result by accumulating the digital output signals of the group of flip-flops, and an encoder that captures and encodes the digital output signals of the group of flip-flops, and may perform self-diagnosis of the group of correction elements based on the accumulation result.
[0507] <Eleventh embodiment> In the second embodiment, a MOS varactor capacitance is used as the correction element (FIG. 14), in the seventh embodiment, a MOS differential pair array is used as the correction element (FIG. 35), in the eighth embodiment, a capacitance array is used as the correction element (FIG. 37), and in the tenth embodiment, a current source array (actually an array of MOS transistors) is used (FIG. 41).
[0508] Thus, various devices are applied as the correction element.
[0509] An example of a correction element using a resistor will be described below as an eleventh embodiment.
[0510] FIG. 42 is a circuit diagram showing an example of a reference voltage circuit according to the eleventh embodiment of the present disclosure.
[0511] The configuration is described below.
[0512] An input resistor 458 (RI-Δ) and a feedback resistor 459 (RF) are connected in series, and their common connection point is connected to an inverting input terminal of an amplifier 455. A reference voltage VCM is connected to a non-inverting input terminal of the amplifier 455. The other end of the feedback resistor 459 is connected to the output of the amplifier 455.
[0513] A reference voltage VR is input to the input of the input resistor 458 .
[0514] In this way, an inverting amplifier configuration is adopted and the voltage level is amplified.
[0515] A group of correction elements 456 for fine adjustment and a group of correction elements 457 for testing the group of correction elements are connected in parallel to input resistor 458 (RI-Δ), and the inputs of the group of correction elements and the group of test elements are controlled by a control signal (not shown) to be either VR or VCM, and the voltage levels are corrected.
[0516] Only during testing, the input of 457 is made switchable between VR and VCM, and during times other than testing, the input of 457 is set to VCM, thereby performing a weighting test of the resistance of the correction element group 456.
[0517] The output voltage VREFOUT can be used to correct the characteristics of each analog block, for example, as a reference voltage for an AD converter, to adjust the DAC gain, or as a voltage for adjusting the bias current.
[0518] As for the inspection and diagnosis method, the weighting inspection and diagnosis can be performed in a procedure similar to that of the comparator weighting inspection and diagnosis in the second embodiment and the amplifier gain weighting inspection and diagnosis method in the fifth embodiment.
[0519] The correction elements are inspected and diagnosed using the comparator inside the AD converter that receives VREFOUT as a reference voltage and the sub-ADC conversion results.
[0520] In this way, the weighting of resistors used as correction elements can also be inspected and diagnosed by the method of the present disclosure.
[0521] As described above, in the semiconductor circuit of the eleventh embodiment, the correction element and the test element are composed of at least one of a capacitance element, a resistance element, and a MOS transistor, and are used for correcting an analog voltage or correcting time.
[0522] <Twelfth embodiment> In the second embodiment, which targets the offset of the comparator, when testing the P-side correction element (FIG. 14), a test element on the N-side is used, and similar testing and diagnosis can be performed using the test element on the opposite polarity side as the reference.
[0523] 43A and 43B show an example in which test elements of opposite polarity are controlled.
[0524] 43A and 43B show an example in which the MSB of the P side is checked, but codes can be generated using the same principle for other bits.
[0525] FIG. 43A shows a second embodiment, in which a P-side inspection element is used as a reference for inspecting a P-side correction element.
[0526] On the other hand, FIG. 43B shows a twelfth embodiment, which uses an N-side correction element.
[0527] In the comparator offset, the P-side correction amount Vcalp and the N-side correction amount Vcalm "act with opposite signs," so in the first fixed code, the N-side check bit is set to "1."
[0528] This is equivalent to the P side code subtracting the decimal number 1 from the decimal number 8.
[0529] Next, in the second fixed code, the check bit on the N side is set to "0".
[0530] This is equivalent to the P side code subtracting decimal 0 from decimal 7 (doing nothing).
[0531] Therefore, by comparing the first and second integration results obtained using the first and second fixed codes, the weighting of the MSB of the P-side correction element can be inspected and diagnosed.
[0532] In a circuit (FIG. 39) showing an example of a capacitance correction unit in the ninth embodiment, which targets a DAC gain correction element, a similar inspection and diagnosis can be performed using the test element on the opposite polarity side as a reference.
[0533] 44A and 44B show an example in which test elements of opposite polarity are controlled.
[0534] Although Fig. 44A and Fig. 44B show an example in which the MSB on the P side is inspected, codes can be generated using the same principle for other bits.
[0535] FIG. 44A shows a ninth embodiment, in which a P-side inspection element is used as a reference for inspecting a P-side correction element.
[0536] On the other hand, FIG. 44B shows a twelfth embodiment, in which an N-side correction element is used.
[0537] When a compensation capacitance is provided on the positive and negative sides of a differential, such as a DAC gain compensation element, the compensation amounts "act with the same sign."
[0538] In other words, if the correction capacitance of either polarity is increased, the DAC gain decreases.
[0539] Therefore, when the method of controlling the test element on the opposite polarity side is adopted, the test bit on the N side is set to "0" in the first fixed code.
[0540] This is equivalent to the P side code adding decimal 0 to decimal 8 (doing nothing).
[0541] Next, in the second fixed code, the check bit on the N side is set to "1".
[0542] This is equivalent to adding decimal 1 to the P side code, which is decimal 7.
[0543] Therefore, by comparing the first and second integration results obtained using the first and second fixed codes, the weighting of the MSB of the P-side correction element can be inspected and diagnosed.
[0544] As mentioned above, if there are correction elements and test elements in both differential circuits, it is clear that either polarity or a combination of polarities may be used as the test elements.
[0545] As described above, in the semiconductor circuit according to the twelfth embodiment, the analog-to-digital converter is a pipeline analog-to-digital converter and includes a residual amplifier and one or more comparators, the residual amplifier is configured as an open-loop integrating amplifier, and as a circuit configuration for adjusting the gain of the residual amplifier, a correction element group is provided in each of a positive-side transmission path and a negative-side transmission path inside the comparator, or a correction element group is provided for adjusting a delay amount for adjusting an amplification time of the residual amplifier, and the digital circuit inspects the correction element group using an integration result of the comparator.
[0546] <Thirteenth embodiment> The circuit configuration and operation of the thirteenth embodiment will be described below with reference to the drawings.
[0547] The offset correction element of the comparator according to the second embodiment is the object of inspection and diagnosis, and an example of an offset control section of the comparator shown in FIG. 45 is used.
[0548] In FIG. 45, 1LSB of the correction element group is also used as the test element group.
[0549] In the case shown in the figure, there is no way to inspect 2C at (LSB+1), but it is possible to inspect and diagnose the capacitance weightings at (LSB+2) and above (that is, at (MSB-1) and above).
[0550] However, in other words, (6C, 3C, 2C, 1C) in the figure has an equivalent ratio to (3C, 1.5C, 1C, 0.5C), so it can be seen as a correction capacitance array with a weighting of 1.5 or less, with the test element being 0.5C.
[0551] On the other hand, since the test element is equivalent to 0.5C, there is a drawback in that the correction range for the number of bits is reduced.
[0552] The embodiments of the present disclosure have been described above.
[0553] It should be noted that the present disclosure is not limited to the configurations of the above-described embodiments, and naturally includes various modifications and alterations that a person skilled in the art could make within the scope of the present disclosure.
[0554] As described above, in the semiconductor circuit according to the thirteenth embodiment, one or more correction elements in the group of correction elements also serve as test elements.
[0555] Although the semiconductor circuit according to one or more aspects of the present invention has been described based on the embodiment, the present invention is not limited to this embodiment. As long as it does not deviate from the spirit of the present invention, various modifications conceived by a person skilled in the art to this embodiment and forms constructed by combining components in different embodiments may also be included within the scope of one or more aspects of the present invention. [Industrial Applicability]
[0556] The semiconductor circuit of the present disclosure can be used as a data converter, such as an analog-to-digital converter or a time-to-digital converter. [Explanation of symbols]
[0557] 100, 500, 510 AD converter 101, 111, 321, 381, 411, 501, 511 Analog circuits 102, 112, 322, 382, 412, 502, 512 Digital circuits 103, 447, 503, 513 Encoder section 104, 119, 172, 231, 283, 334, 395, 425, 444, 504, 514 Error detection circuit 105, 120, 171, 233, 282, 335, 396, 426, 445 Self-diagnosis circuit 106, 121, 336, 393, 423, 442, 515 Correction code generation circuit 107, 122, 170, 232, 281, 337, 394, 424, 443 Self-diagnostic code generator 108, 123, 125, 180A, 180B, 180C, 180D, 234A, 234B, 234C, 234D, 254, 255, 284C, 284D, 333, 338, 389, 390, 419, 420, 441, 472 Selector 110, 380, 410 Sequential AD converter 113, 323, 383, 413 Sample and hold circuit 114, 328, 385, 415 DAC (Digital-to-Analog Converter) 115, 329, 386, 416 Driver circuit 116, 130, 327, 384, 414 Comparators 117, 340, 387, 417 DAC control section 118, 253, 331, 339, 391, 421 SAR Control Unit 124, 252, 332, 392, 422 Memory section for DAC direct control mode 131 Differential amplifier section 132 Offset control section 133 Differential Latch Circuit 134 Correction capacity control section 135, 140, 352, 354, 356, 372, 373, 400, 402, 435, 437, 456, 470 Correction element group 136, 353 Correction element driver circuit 137 Comparator Offset 138, 139 Correction amount 141, 355, 357, 374, 375, 401, 403, 436, 438, 457, 471 Inspection element group 150~163, 210~226, 260~271, 310~314 Examination and diagnostic procedures 173, 230, 280 State control section 181, 295 Demultiplexer 182 First fixed code generator 183 Second fixed code generator 184, 245, 246, 247, 293 Binary search control section 185 First Fixed Code Register 186 Second Fixed Code Register 187, 236, 237, 238, 291 Search Code Register 188, 242, 243, 244, 292 Optimal Code Register 189, 239, 240, 241, 294 Minimum Error Register 190, 285 Check bit selection section 191 Accumulation Expected Value Register 192 Accumulator 193 First accumulation result register 194 Second accumulation result register 195 CDF Standard Value Register 196 Weight Standard Value Register 197 CDF judgment section 198 Weight Judgment Unit 199, 200, 256, 296 Subtractor 235 MSB fixed code register 248 First Code Result Register 249 Second Code Result Register 250 Range Standard Value Register 251 Range Judgment Unit 286 Positive side search code register 287 Negative side search code register 288 Positive Side Optimal Code Register 289 Negative Optimal Code Register 290A, 290B Code Mapping Section 297 First accumulation result register 298 Second accumulation result register 299 Toggle Standard Value Register 300 Toggle judgment unit 301A, 301B Bit inversion section 320 Pipeline AD Converter 324 1st Sub ADC 325 Residual Amplifier 326 Second Sub ADC 330 Pipeline Control Section 350 Signal Amplification Unit 351 Amplification time adjustment unit 358 First set of elements 359 Second Set of Elements 370 Signal Amplification Unit 371 Differential Pair with Offset Adjustment 397A, 397B DAC MSB Capacitance 398A, 398B, 430A, 430B DAC compensation capacitance Binary Capacitance for 427A, 428A, 429A, 427B, 428B, 429B DACs 440 TD Converter (Time-to-Digital Converter) 446 Edge detection and counter circuit 455 Amplifier 458 Input Resistance 459 Feedback resistor topPlateP Positive comparator input voltage topPlateN Negative comparator input voltage VRH High voltage reference VRL Low voltage reference VCM Midpoint Reference Voltage CCp Control code for the positive comparator offset correction capacitance CCn Control code for negative comparator offset correction capacitance Vgp Positive output of the differential amplifier of the comparator Vgn Negative output of the differential amplifier of the comparator CapP Compensation capacitance on the positive side of the comparator CapN Compensation capacitance on the negative side of the comparator
Claims
1. A semiconductor circuit including a data converter that outputs a digital output signal corresponding to a signal under test, an analog circuit for quantizing the input signal under test; a digital circuit that outputs the digital output signal; the analog circuit includes a correction element group including a correction element for correcting nonlinearity occurring in a process of converting the signal under test into the digital output signal in the analog circuit; a test element group including a test element for testing a function of the correction element or a weighting of the correction element; the digital circuit tests the correction elements using the test elements; The digital circuit includes: a self-diagnosis code generating circuit that generates a self-diagnosis code for enabling or disabling each correction element included in the correction element group and for enabling or disabling each test element included in the test element group; an error detection circuit that detects, as an error signal, a difference between a quantized output signal of the analog circuit output in response to the self-diagnosis code and a predetermined threshold value; a self-diagnosis circuit for generating a self-diagnosis signal indicative of a diagnosis result of the correction element group based on the error signal; Equipped with the self-diagnosis code includes a first fixed code and an optimum value code, the first fixed code is a code for enabling one correction element selected as a test target from the group of correction elements and disabling other correction elements and test elements; the optimum value code is a code that minimizes the error signal with respect to the first fixed code, and is generated by a binary search method; The self-diagnosis circuit includes: calculating an integrated value of the error signal corresponding to a combination of the first fixed code and the optimum value code; It is judged whether the integrated value is within a predetermined range, and if it is not within the predetermined range, it is judged that the correction element group is defective. Semiconductor circuit.
2. A semiconductor circuit including a data converter that outputs a digital output signal corresponding to a signal under test, an analog circuit for quantizing the input signal under test; a digital circuit that outputs the digital output signal; the analog circuit includes a correction element group including a correction element for correcting nonlinearity occurring in a process of converting the signal under test into the digital output signal in the analog circuit; a test element group including a test element for testing a function of the correction element or a weighting of the correction element; the digital circuit tests the correction elements using the test elements; The digital circuit includes: a self-diagnosis code generating circuit that generates a self-diagnosis code for enabling or disabling each correction element included in the correction element group and for enabling or disabling each test element included in the test element group; an error detection circuit that detects, as an error signal, a difference between a quantized output signal of the analog circuit output in response to the self-diagnosis code and a predetermined threshold value; a self-diagnosis circuit for generating a self-diagnosis signal indicative of a diagnosis result of the correction element group based on the error signal; Equipped with the self-diagnosis code includes a first fixed code, a second fixed code, and an optimum value code, the first fixed code is a code for enabling one correction element selected as a test target from the group of correction elements and disabling other correction elements and test elements; the optimum value code is a code that minimizes the error signal with respect to the first fixed code, and is generated by a binary search method; the signal under measurement is a differential signal including a first polarity signal and a second polarity signal, The correction element group includes a first element group connected to a transmission path of the first polarity signal; a second element group connected to a transmission path of the second polarity signal, the first fixed code corresponds to a first group of elements and disables a second group of elements; the second fixed code is a code for disabling one correction element selected as a test target in the first element group, enabling other correction elements having a smaller weight than the correction element, and enabling a test element; the optimum value code corresponds to a second group of elements and disables a first group of elements; The self-diagnosis circuit includes: calculating a first integrated value by integrating the error signal corresponding to a combination of the first fixed code and the optimum value code; calculating a second integrated value by integrating the error signal corresponding to a combination of the second fixed code and the optimum value code; A magnitude relationship between the first integrated value and the second integrated value is determined, and if the magnitude relationship does not satisfy a predetermined condition, the correction element to be inspected is determined to be defective. Semiconductor circuit.
3. The self-diagnosis circuit sequentially selects one correction element as a test target from the group of correction elements.
3. The semiconductor circuit according to claim 1.
4. Among the group of correction elements, one or more correction elements are also used as the inspection element. The semiconductor circuit according to claim 1 .
5. A semiconductor circuit including a data converter that outputs a digital output signal corresponding to a signal under test, an analog circuit for quantizing the input signal under test; a digital circuit that outputs the digital output signal; the analog circuit includes a correction element group including a correction element for correcting nonlinearity occurring in a process of converting the signal under test into the digital output signal in the analog circuit; a test element group including a test element for testing a function of the correction element or a weighting of the correction element; The digital circuit uses the test elements to test the correction elements. The analog circuit includes: a quantizer for quantizing the signal under test or the reference signal after the correction by the group of correction elements and converting the signal under test or the reference signal into a quantizer output signal; The digital circuit includes: an encoding unit that converts the quantizer output signal into a predetermined weighted digital output signal and outputs the converted signal; an error detection circuit that outputs an integration result obtained by integrating a difference between the quantizer output signal and a predetermined threshold value multiple times as an error signal; a self-diagnosis code generating circuit that outputs a self-diagnosis code for enabling or disabling each correction element included in the correction element group and for enabling or disabling each test element included in the test element group; a self-diagnosis circuit that receives the error signal and outputs a self-diagnosis output signal indicating whether the correction element group is good or bad; Equipped with performing a self-diagnosis of each of the correction elements of the correction element group by the self-diagnosis circuit; The self-diagnosis code generation circuit includes: an optimum correction code output mode in which an optimum correction code is output so as to minimize the quantization error of the quantizer in response to the error signal; a logic inversion code output mode for outputting a logic inversion code obtained by logically inverting a first bit signal of the optimum correction code that controls the correction element to be tested; having The digital circuit includes: during testing, setting the optimum correction code output mode, supplying the optimum correction code to the analog circuit, and generating a first error signal based on an error signal from the error detection circuit; setting the logic inversion code output mode, supplying the logic inversion code to the analog circuit, and generating a second error signal based on the error signal from the error detection circuit; The magnitude relationship between the first error signal and the second error signal is compared to perform a fault diagnosis on the correction element to be inspected. Semiconductor circuit.
6. A semiconductor circuit including a data converter that outputs a digital output signal corresponding to a signal under test, an analog circuit for quantizing the input signal under test; a digital circuit that outputs the digital output signal; the analog circuit includes a correction element group including a correction element for correcting nonlinearity occurring in a process of converting the signal under test into the digital output signal in the analog circuit; a test element group including a test element for testing a function of the correction element or a weighting of the correction element; The digital circuit uses the test elements to test the correction elements. The analog circuit includes: a quantizer for quantizing the signal under test or the reference signal after the correction by the group of correction elements and converting the signal under test or the reference signal into a quantizer output signal; The digital circuit includes: an encoding unit that converts the quantizer output signal into a predetermined weighted digital output signal and outputs the converted signal; an error detection circuit that outputs an integration result obtained by integrating a difference between the quantizer output signal and a predetermined threshold value multiple times as an error signal; a self-diagnosis code generating circuit that outputs a self-diagnosis code for enabling or disabling each correction element included in the correction element group and for enabling or disabling each test element included in the test element group; a self-diagnosis circuit that receives the error signal and outputs a self-diagnosis output signal indicating whether the correction element group is good or bad; Equipped with performing a self-diagnosis of each of the correction elements of the correction element group by the self-diagnosis circuit; the self-diagnosis code generation circuit has a first weight check code output mode and a second weight check code output mode; In the first weight check code output mode, A first bit signal that controls a first correction element to be tested is set to 1 or 0; a second correction bit signal group for controlling a correction element having an adjustment weight less than the first correction element is set to an inverted logical value of the first bit signal; a test bit signal for controlling the test element is an inverted logical value of a first bit signal; The other bit signals are fixed as predetermined signals. Output, In the second weight check code output mode, the first bit signal is set to an inverted logical value when the first weight check code output mode is set; the second correction bit signal group is an inverted logical value of the first bit signal; a test bit for controlling the test element is an inverted logical value of a first bit signal; The other bit signals are fixed as predetermined signals. Output, The digital circuit includes: during a check, a first weight check code output mode is set to the first weight check code output mode, and a first error signal is generated based on the error signal from the error detection circuit; setting the second weight check code output mode to generate a second error signal based on the error signal from the error detection circuit; A difference between the first error signal and the second error signal is calculated, and a weighting diagnosis of the correction element group is performed. Semiconductor circuit.
7. A semiconductor circuit including a data converter that outputs a digital output signal corresponding to a signal under test, comprising: an analog circuit for quantizing the input signal under test; a digital circuit that outputs the digital output signal; the analog circuit includes a correction element group including a correction element for correcting nonlinearity occurring in a process of converting the signal under test into the digital output signal in the analog circuit; a test element group including a test element for testing a function of the correction element or a weighting of the correction element; The digital circuit uses the test elements to test the correction elements. The analog circuit includes: a quantizer for quantizing the signal under test or the reference signal after the correction by the group of correction elements and converting the signal under test or the reference signal into a quantizer output signal; The digital circuit includes: an encoding unit that converts the quantizer output signal into a predetermined weighted digital output signal and outputs the converted signal; an error detection circuit that outputs an integration result obtained by integrating a difference between the quantizer output signal and a predetermined threshold value multiple times as an error signal; a self-diagnosis code generating circuit that outputs a self-diagnosis code for enabling or disabling each correction element included in the correction element group and for enabling or disabling each test element included in the test element group; a self-diagnosis circuit that receives the error signal and outputs a self-diagnosis output signal indicating whether the correction element group is good or bad; Equipped with performing a self-diagnosis of each of the correction elements of the correction element group by the self-diagnosis circuit; The data converter includes: an input switching unit that switches between a signal to be input to the data converter as a signal under test and the reference signal; The self-diagnosis code generation circuit includes: an optimum correction code output mode for outputting an optimum correction code that minimizes the quantization error of the quantizer; The digital circuit includes: During the inspection, an optimum correction code output mode is set, a first reference signal is selected by the input switching unit, and the optimum correction code is obtained as a first optimum correction code; setting an optimum correction code output mode, selecting a second reference signal by the input switching unit, and obtaining the optimum correction code as a second optimum correction code; A difference between the first optimum correction code and the second optimum correction code is calculated, and a correction range diagnosis of the correction element group is performed. Semiconductor circuit.
8. the quantizer comprises one or more comparators; The error detection circuit generates the error signal by integrating the output result of one comparator or the output results of two or more comparators, which are digitally calculated with a predetermined weighting, as the quantizer output signal.
8. The semiconductor circuit according to claim 5.
9. Among the group of correction elements, one or more correction elements are also used as the inspection element.
8. The semiconductor circuit according to claim 5.
10. the correction element group includes a plurality of weighted correction elements; The test element has a weight equal to or smaller than a correction element corresponding to the smallest weight among the correction elements.
8. The semiconductor circuit according to claim 5.
11. The correction element and the inspection element are composed of at least one of a capacitance element, a resistance element, and a MOS transistor, and are used for correcting an analog voltage or correcting time.
8. The semiconductor circuit according to claim 5.
12. the analog circuit is composed of a differential circuit, a first element group including a correction element and a test element connected to a positive input path of the differential circuit; a second element group including a correction element and a test element connected to a negative input path of the differential circuit; Equipped with the first element group and the second element group constitute the correction element group and the test element group, The self-diagnosis code generation circuit includes: a first set inspection mode for inspecting the first group of elements; a second set inspection mode for inspecting the second group of elements; having The first element group and the second element group are individually inspected.
8. The semiconductor circuit according to claim 5.
13. The analog circuit includes: a first element group including the correction element group and the test element group, the first element group belonging to one node; a second group of elements having the same weights and configurations as the first group of elements; A redundant configuration including: The self-diagnosis code generation circuit includes: a first set inspection mode for inspecting the first group of elements; a second set inspection mode for inspecting the second group of elements; having The first element group and the second element group are individually inspected.
8. The semiconductor circuit according to claim 5.
14. The self-diagnosis code generation circuit includes: In the first set inspection mode, For the first element group, A first bit signal that controls the correction element to be inspected is set to 1 or 0; a second correction bit signal group for controlling a correction element having an adjustment weight less than the first correction element is set to an inverted logical value of the first bit signal; a test bit signal for controlling the test element is an inverted logical value of a first bit signal; The other bit signals are output as predetermined fixed signals. With respect to the second element group, outputting an optimal correction code so that the quantization error of the quantizer is minimized; The digital circuit includes: In the first set test mode, a first error signal is generated based on the error signal of the error detection circuit; Regarding the control code of the first group of elements, The first bit signal is an inverted logical value of a bit signal at a first step, the second correction bit signal group is an inverted logical value of the bit signal at the first step; A test bit for controlling the test element is set to an inverted logical value of a bit signal at the first step; The other bit signals are output as predetermined fixed signals, With respect to the second element group, maintain and output the optimal correction code; generating a second error signal as an output of the error detection circuit; At a transition point of the quantizer where the quantization error is minimized, a comparison is made between the first error signal and the second error signal to check whether the weight is equal to or smaller than a specified value; In the second set inspection mode, The relationship between the first element group and the second element group in the first set inspection mode is exchanged, and the same process is performed.
14. The semiconductor circuit according to claim 12 or 13.
15. When inspecting a correction element included in the first element group, Using the test element included in the second element group, When inspecting a correction element included in the second element group, Using the test element included in the first element group 14. The semiconductor circuit according to claim 12 or 13.
16. the data converter is an analog-to-digital converter; 16. The semiconductor circuit according to claim 5.
17. the analog-to-digital converter comprises one or more comparators; the comparator is configured by a differential circuit, a correction element group and a test element group in each of a positive side transmission path and a negative side transmission path in the differential circuit for correcting an offset of the comparator; Equipped with The correction elements included in the correction element group of the positive side transmission path and the negative side transmission path are individually inspected based on the inspection element group.
17. The semiconductor circuit of claim 16.
Citation Information
Patent Citations
Preparation of thienylacetic acid
JP1979012369A
Da conversion device and ad conversion device
JP2009118488A
Time digital converter
JP2010273118A
A / d converter and a / d conversion method
JP2011188097A
A / d conversion circuit and method of testing the same
JP2013005185A
