(19) United States US 20100079319A1 (12) Patent Application Publication (10) Pub. No.: US 2010/0079319 A1 Berens et al. (43) Pub. Date: Apr. 1, 2010 (54) DATA CONVERSION CIRCUITRY AND METHOD THEREFOR (76) Inventors: Michael T. Berens, Austin, TX (US); James R. Feddeler, Austin, TX (US) Correspondence Address: FREESCALE SEMICONDUCTOR, INC. LAW DEPARTMENT 7700 WEST PARMER LANE MD:TX32/PL02 Publication Classi?cation (51) Int. Cl. H03M 1/10 (2006.01) (52) US. Cl. ...................................................... .. 341/120 (57) ABSTRACT A data converter for converting analog signals to digital sig nals, or for converting digital signals to analog signals is provided. In one embodiment, a production self-test is pro vided. In one embodiment, a high-speed lower-resolution method or mode for a data converter is provided. In one embodiment, a differential data converter With a more stable AUSTIN, TX 78729 (US) comparator common mode voltage is provided. In one embodiment, the input range of a digitally calibrated data _ converter is provided and maintained so that there is no loss in (21) Appl' NO" 12/242’124 input range due to the calibration. In one embodiment, digital post-processing of an uncalibrated result using a previously (22) Filed: Sep. 30, 2008 stored calibration value is provided. f 10 PROCESSOR MEMORY E E . (n 22 D 24 / °°§ < '\ i g; / z “J O: i E z X v-t Lu DATA g9 OTHER- CONVERTER (FIGS. 2. 3. MODULE 14, 18) — ll
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
(19) United States US 20100079319A1
(12) Patent Application Publication (10) Pub. No.: US 2010/0079319 A1 Berens et al. (43) Pub. Date: Apr. 1, 2010
(54) DATA CONVERSION CIRCUITRY AND METHOD THEREFOR
(76) Inventors: Michael T. Berens, Austin, TX (US); James R. Feddeler, Austin, TX (US)
Correspondence Address: FREESCALE SEMICONDUCTOR, INC. LAW DEPARTMENT 7700 WEST PARMER LANE MD:TX32/PL02
A data converter for converting analog signals to digital sig nals, or for converting digital signals to analog signals is provided. In one embodiment, a production self-test is pro vided. In one embodiment, a high-speed lower-resolution method or mode for a data converter is provided. In one embodiment, a differential data converter With a more stable
AUSTIN, TX 78729 (US) comparator common mode voltage is provided. In one embodiment, the input range of a digitally calibrated data
_ converter is provided and maintained so that there is no loss in (21) Appl' NO" 12/242’124 input range due to the calibration. In one embodiment, digital
post-processing of an uncalibrated result using a previously (22) Filed: Sep. 30, 2008 stored calibration value is provided.
CHARGE BOTTOM PLATE OF MINUS DAC 282 CAPACITORS TO VIN 93 CHARGE BOTTOM PLATE
0F PLUS DAC 280 CAPACITORS TO VIN 92 CHARGE BOTTOM PLATE OF DIFFERENTIAL BIAS CAPACITOR 208 TO VREFH 88 WHILE CHARGING
THE COMPARATOR INPUTS TO VCM 94
S 302
WAS DIFFERENTIAL
BIAS CAPACITOR 208 ' SWITCHED?
{303 RESULT ‘286 EQUALS RELEASE COMPARATOR INPUTS AND SWITCH THE DIFFERENTIAL RESULT THE BOTTOM PLATES OF ALL CAPACITORS
T0 VREFL g0 SUBTRACT PREDETERMINE §304 AMOUNT FROM DIFFERENTIAL
SULT TO PR PERFORM A PARTIAL SUCCESSIVE APPROXIMATION RE RESULT 28%DUCE ON THE MINUS DAC 282 UNTIL THE INVERTING ( COMPARATOR INPUT IS SUBSTANTIALLY NEAR 4 310 THE COMPARATOR'S 260, 261 COMMON MODE INPUT VOLTAGE (VCM 94). (NoTE THAT 301
THE MINUS RESULT IS THE SERIAL OUTPUT END OF THE NON-CRITICAL COMPARATOR 261)
. 306 305 FJ
SWITCH THE BOTTOM PLATE OF THE DIFFERENTIAL CAPACITOR 208
TO VREFL 90
THE CRITICAL COMPARATOR 260 OUTPUT HIGH?
NO
PERFORM FULL SUCCESSIVE APPRoxINATI0N 3,07 ON THE PLUS DAC 280
(NoTE THAT THE PLUS RESULT IS THE SERIAL OUTPUT OF THE CRITICAL COMPARATOR 260)
SUBTRACT MINUS RESULT 285 FROM PLUS A398 RESULT 284 TO PRODUCE A DIFFERENTIAL
DURING THE SAMPLE PHASE: g 552 — CHARGE ALL UNITS SMALLER THAN
CAPACITOR UNDER TEST TO VREFH 88 EXCEPT Y UNITS THROUGH SWITCHING DURING THE SAMPLE PHASE:
- CHARGE ALL UNITS SMALLER THAN
CAPACITOR UNDER TEST TO vREFH 88 ELZEQLSETRJNEZQSNISEE FgféA'colLoR EXCEPT X UNITS THROUGH SWITCHING - UNDER TEST To VREFH 88 FOR CIRCUITRY 299 (SEE FIG. 10) OFFSET SHIFT
_ ¥gD§EEgLESg0AND - CHARGE CAPACITOR UNDER TEST AND
- REFERENCE SIDE CHARGED TO okkFfTggR LARGER UNITS- To mm 90 DURING ENTIRE SEQUENCE _ REFERENCE SIDE CHARGED To
; 553 vREFL 90 DURING ENTIRE SEQUENCE
DURING THE HOLD PHASE: g 557 - RELEASE COMPARATOR INPUTS USING
SwITCH CIRCUITRY 102 (SEE FIG. 10} DURING THE HOLD PHASE/COMPARE/ RESULT:
‘ SEQEEE8§APACHOR UNDER TEST To - FOLLOw PREVIOUS ROUTINE - RESULT SHOULD BE CENTERED AT
, §554 OFFSET + Y (CAPACITOR UNDER TEST SHOULD NOT SET)
DURING THE COMPARE PHASE: — RUN SUCCESSIVE APPROXIMATION
(RESULT SHOULD BE CENTERED AT X, AND SHOULD BE LESS THAN CAPACITOR UNDER TEST) CAPACITORS UNDER
TEST GREATER THAN Y BEEN TESTED? gsss
DURING RESULT PROCESSING: - COMPARE RESULT TO x VALUE
STORED IN SAMPLE PHASE - SET LIMITS ON COMPARISON BASED
ON REQUIRED MATCHING TOLERANCE NOTES: A AND NOISE ALLOWANCE X REPRESENTS A NUMBER OF IDEALLY SIZED
UNITS EOUAL TO THE MAXIMUM EXPECTED (UNSIGNED) ERROR OF THE CAPACITOR UNDER
558 TEST, PLUS DESIRED OFFSET. FOR EXAMPLE, IF A CAPACITOR IS SUPPOSED TO BE 64LSB +/
CAPACITORS UNDER 4LSB AND THE DESIRED OFFSET IS 12, x TEST GREATER THQAN X SHOULD BE EQUAL TO 16.
BEEN TESTED' WHEN X IS EQUAL/LARGER THAN CAPACITOR UNDER TEST, OFFSET MUST BE INCREASED TO BE GREATER THAN CAPACITOR UNDER TEST BY SETTING ONE UNIT GREATER THAN THE CAPACITOR UNDER TEST FOR OFFSET (TO KEEP RESULT FROM BEING TOO NEAR ZERO) AND
FIG 20 THEN SETTING A NEW vALUE Y EOUAL TO ' THE EXPECTED ERROR ONLY.
DIFFERENTIAL LIN H LIN LIN LIN H LIN H OFFSET VALUE SIDE TESTED A ‘L; A} COMPARATOR SPEED
TESTED IN SUCCESSIVE APPROXIMATION TESTS HP?“
PHASE OF A GUTN N4 TESTS A - + AND HOLD PHASE 2 J- -L J- J FoR GUTNH TESTS -b A
\_,\ / \ / \ / A , \ / GUT AND CORRESPONDING
VREFH SWITCH TESTED INGDNPAREPHASEGF
?LINH LINH LINH LINH LINH .GUTNTEST vIN SNITGH TESTED IN SNALLEST CUT(S) TESTED AGAINST SEQUENCE
SAMPLE PHASE OF GUTNH LARGE CUTS IN SANPLE PHASES TESTS AND A SINGLE, FULL FOR GUTN +1 TESTS (NH > NININUN
RAIL GDNvERSIGN To - NoISE 0R MATCHING TDLERANGE) vREFH FOR LARGEST GUT
FIG. 21 600
( START I {620
CHARGE BOTTOM PLATE 0F MSB GAPAGITDRS 119, 219 £02 (HALF OF THE ToTAL CAPACITANCE) T0 MN 92
wHILE CHARGING THE COMPARATOR INPUTS T0 WM 94
RELEASE COMPARATOR INPUTS AND SWITCH THE BOTTOM AJ PLATE OF MSB CAPACITORS H9, 219 TO VREFL 9O
PERFORM A SUCCESSIVE APPROXIMATION ON THE 13 14 MOST SIGNIFICANT BITS TO PRODUCE A 13 OR
14-BIT CONVERSION RESULT OF VIN/2
OR F304
7
SHIFT THE CONVERSION RESULT LEFT (DOUBLE) AND 305 ROUND (IF DESIRED) TO GET THE 1/2 LSB SHIFT
T0 PRODUCE A 12-BIT CONVERSION RESULT
' 601 END
FIG. 22
US 2010/0079319 A1
DATA CONVERSION CIRCUITRY AND METHOD THEREFOR
CROSS-REFERENCE TO RELATED
APPLICATION(S) [0001] This application is related to US. patent application Ser. No. (Attorney Docket No. AC50034TC), ?led on even date, entitled “DATA CONVERSION CIRCUITRY AND METHOD THEREFOR,” and assigned to the current assignee hereof. [0002] This application is related to US. patent application Ser. No. (Attorney Docket No. AC50035TC), ?led on even date, entitled “DATA CONVERSION CIRCUITRY AND METHOD THEREFOR,” and assigned to the current assignee hereof. [0003] This application is related to US. patent application Ser. No. (Attorney Docket No. AC50036TC), ?led on even date, entitled “DATA CONVERSION CIRCUITRY AND METHOD THEREFOR,” and assigned to the current assignee hereof. [0004] This application is related to US. patent application Ser. No. (Attorney Docket No. AC50037TC), ?led on even date, entitled “DATA CONVERSION CIRCUITRY AND METHOD THEREFOR,” and assigned to the current assignee hereof.
BACKGROUND
[0005] 1. Field [0006] This disclosure relates generally to electrical cir cuitry, and more speci?cally, to electrical circuitry for data conversion. [0007] 2. RelatedArt [0008] Data converters are very useful for converting ana log signals to digital signals, and for converting digital signals to analog signals. Many applications require data converters that have a high resolution, fast conversion time, alloW a broad range of inputs, and yet are cost effective. Other data conversion features may also be important for various appli cations. It is thus important to be able to provide data con ver‘ters that meet a Wide variety of potentially con?icting criteria, While at the same time remain cost effective.
BRIEF DESCRIPTION OF THE DRAWINGS
[0009] The present invention is illustrated by Way of example and is not limited by the accompanying ?gures, in Which like references indicate similar elements. Elements in the ?gures are illustrated for simplicity and clarity and have not necessarily been draWn to scale. [0010] FIG. 1 illustrates, inblock diagram form, a system in accordance With one embodiment. [0011] FIG. 2 illustrates, in partial block diagram form and partial schematic diagram form, a portion of a data converter in accordance With one embodiment. [0012] FIG. 3 illustrates, in partial block diagram form and partial schematic diagram form, a portion of a data converter in accordance With one embodiment.
[0013] FIG. 4 illustrates, in partial block diagram form and partial schematic diagram form, a portion of a DAC in accor dance With one embodiment.
[0014] FIG. 5 illustrates, in How diagram form, a sample calibration method in accordance With one embodiment. [0015] FIG. 6 illustrates, in How diagram form, a sample conversion method in accordance With one embodiment.
Apr. 1,2010
[0016] FIG. 7 illustrates, in graphical diagram form, non linearities due to capacitor mismatch in a binary-Weighted DAC in accordance With the prior art. [0017] FIG. 8 illustrates, in graphical diagram form, non linearities due to capacitor mismatch in a binary-Weighted DAC With oversiZed ?rst scaling capacitor in accordance With one embodiment.
[0018] FIG. 9 illustrates, in graphical diagram form, non linearities due to capacitor mismatch in a binary-Weighted DAC With oversiZed ?rst scaling capacitor after calibration in accordance With one embodiment. [0019] FIG. 10 illustrates, in partial block diagram form and partial schematic diagram form, a portion of a DAC in accordance With one embodiment.
[0020] FIG. 11 illustrates, in How diagram form, a sample conversion method for a 16-bit analog to digital converter (ADC) in accordance With one embodiment. [0021] FIG. 12 illustrates, in graphical diagram form, a transfer function of an ADC With digitally calibrated offset in accordance With one embodiment. [0022] FIG. 13 illustrates, in graphical diagram form, a transfer function of an ADC With digital linearity and gain calibration in accordance With one embodiment. [0023] FIG. 14 illustrates, in partial block diagram form and partial schematic diagram form, a portion of a data con ver‘ter in accordance With one embodiment.
[0024] FIG. 15 illustrates, in How diagram form, a sample differential conversion method in accordance With one embodiment. [0025] FIG. 16 illustrates, in graphical diagram form, an example of a differential conversion in accordance With one embodiment. [0026] FIG. 17 illustrates, in graphical diagram form, another example of a differential conversion in accordance With one embodiment.
[0027] FIG. 18 illustrates, in partial block diagram form and partial schematic diagram form, a portion of a data con ver‘ter in accordance With one embodiment. [0028] FIG. 19 illustrates, in How diagram form, a self-test method for a differential capacitive DAC in accordance With one embodiment.
[0029] FIG. 20 illustrates, in How diagram form, a self-test method for a single-ended capacitive DAC in accordance With one embodiment.
[0030] FIG. 21 illustrates, in schematic diagram form, test coverage of a self-test method in accordance With one embodiment. [0031] FIG. 22 illustrates, in How diagram form, a method for performing a 12-bit conversion in a 16-bit ADC in accor dance With one embodiment.
DETAILED DESCRIPTION
[0032] FIG. 1 illustrates one embodiment of a system 10. In alternate embodiments, system 10 may be implemented as a single integrated circuit, may be implemented as a plurality of integrated circuits, or may be implemented as a combination of integrated circuits and discrete components. Alternate embodiments may implement system 10 in any manner. [0033] In one embodiment, system 10 comprises data con ver‘ter 12, other modules 14, processor 16, memory 1 8, and external bus interface 20, Which are all bi-directionally coupled to each other by Way of a bus 22 or a plurality of electrical signals 22. In one embodiment, system 10 can receive inputs and provide outputs by Way of a bus 24 or a
US 2010/0079319 A1
plurality of electrical signals 24 coupled to external bus inter face 20. In alternate embodiments, system 10 may comprises feWer, more, or different blocks of circuitry than those illus trated in FIG. 1.
[0034] FIG. 2 illustrates one embodiment of a portion of data converter 12 of FIG. 1. In one embodiment, data con verter 12 comprises an ADC Which may be used to convert a differential input voltage VIN 92-VIN 93 into a digital rep resentation stored as a multi-bit binary value in a data register (e.g. ADCRHAzADCRLA or ADCRHBzADCRLB). In one embodiment, this digital representation may be of the value 2AN*(V IN 92-VIN 93)/ (V REFSH 88-VREFSL 90), Where N is the resolution, or number of bits, in data converter 12. In one embodiment, data converter 12 comprises a ?rst input multiplexer Which chooses from among a plurality of positive input voltages (DADP[0:3], AD[4:23], TEMPP) based on softWare con?guration (ADCHN) to create VIN 92, and Which also comprises a second input multiplexer Which chooses from among a plurality of negative input voltages (DADM[0:3], TEMPM) to create VIN 93. Data converter 12 also comprises a reference multiplexer Which chooses from among a plurality of positive reference voltages (V REFH, VALTH, VBGH) to create VREFH 88, and a second reference multiplexer Which chooses from among a plurality of nega tive reference voltages (VREFL, VALTL, VBGL) to create VREFL 90. Note that the terms “positive” and “negative” indicate the polarity of the signal relative to the other, and not to a ?xed reference such as ground. In one embodiment, both positive and negative signals and references are alWays equal to or greater than a ground reference. Alternate embodiments may function in a different manner.
[0035] For one embodiment, the SAR (successive approxi mation register) control circuitry 76 begins a conversion by placing the SAR Converter in an initial condition by asserting the INITIALIZE signal. A conversion Will begin When a trigger to convert signal (TRIGGER) is received by the SAR control circuitry 76 from the conversion trigger control cir cuit. Alternate embodiments may provide a trigger signal due to a variety of different circumstances. For example, a trigger may be received When a softWare register bit is Written (ADTRG), or When a hardWare signal ADHWT is received in the right conditions (e. g. these conditions may be determined by softWare con?guration [ADCSC1A-ADCSC1N, ADCSC2, ADCCFG1 and ADCCFG2] and/or hardWare sig nal conditioning [ADHWTSA-ADHWTSN]). When the asserted trigger signal is received, the SAR control circuitry 76 asserts the SAMPLE condition to the SAR converter, Which in turn samples the differential input voltage VIN 92-VIN 93 on the SAR array. The sample value can be modi ?ed by the PG and MG con?gurations stored in the calibration storage circuitry 68. [0036] In one embodiment, the SAR converter samples for a period indicated by softWare con?guration (ADLSMP, ADLSTS) in a multiple of the ADC input clock (ADCK) periods. The ADCK period may be controlled by softWare con?gurations (ADIV, ADICLK, ADACKEN) and hardWare clock sources (ADACK, BUS_CLOCK, and ALTCLK). The SAR control circuitry 76 then places the SAR converter into CONVERT mode. In one embodiment of CONVERT mode, the SAR converter subsequently compares the input voltage (VIN 92-VIN 93) to different fractions of the reference volt age (V REFSH 88-VREFSL 90). During each comparison, the converter successively sets or clears the corresponding digital output bit based on the compare result, and then
Apr. 1,2010
changes either the reference voltage or the input voltage by the appropriate fraction of the reference voltage (eg if com paring the input voltage to the reference voltage divided by tWo, if the comparison is greater, the output bit is set and the next comparison is to 3/4 times the reference voltage; if less, the output bit is cleared and the next comparison is to 1A times the reference voltage; either the reference voltage or the input voltage may be modi?ed during successive approximation). [0037] As the SAR converter approximates, it may modify the result as it proceeds With the values CLPx and CLMx stored in the calibration storage circuitry 68. When the SAR converter has made the appropriate number of successive approximations, SAR control circuitry 76 indicates that it is COMPLETE to the SAR trigger circuitry and instructs the SAR converter to TRANSFER the results to the output cir cuitry. In one embodiment, this output circuitry ?rst adjusts for offset in the OFFSET SUBTRACTOR, then employs averaging if so con?gured in the AVERAGER, and then for mats the data in the appropriate manner in the FORMAT TING circuit. These circuits may be controlled by softWare con?guration (ADCOFS, AVGE and AVGS, and MODE and DIFFn, respectively). The offset value OFS used by the OFF SET SUBTRACTOR, as Well as the con?guration values PG, MG, CLPx, and CLMx, may be created before conversion by the calibration control circuitry 66. Once formatted the result is compared to a compare value (CV1) or range (CV1, CV2) in the COMPARE LOGIC. Based on the softWare con?gura tion to the COMPARE LOGIC (ACFE, ACFGT, ACREN) the comparator Will transfer the result to the result registers (AD CRHAzADCRLA to ADCRHNzADCRLN) and set COMP ARE_TRUE. In one embodiment, the conversion trigger logic and SAR control circuitry 76 Will then determine, based on softWare con?guration (ADCO), Whether to begin another conversion or to ABORT the sequence and turn the SAR converter off. Alternate embodiments of the data converter 12 of FIG. 2 may use more, less, or different circuitry to imple ment circuitry for performing data conversion. [0038] FIG. 3 illustrates one embodiment of a portion of data converter 12 of FIG. 1. Referring to FIG. 3, the illustrated successive approximation (SAR) analog-to-digital converter (ADC or A/D converter) comprises a digital-to-analog con verter (DAC) 62 and a comparator 60 in a feedback loop With logic including a successive-approximation (SAR) register 96. In one embodiment, DAC 62 comprises an array of binary Weighted elements (eg capacitors 110-119 of FIG. 4). Alter nate embodiments may use any type of charge redistribution array for data conversion. In addition, alternate embodiments may use any desired and appropriate binary Weighted ele ments (e. g. resistive elements, capacitive elements, a combi nation thereof, etc.). Note that “N”, “M”, and “P” are being used to represent integers. For example, “bN” is the “nth bit” or “bit N”; similarly “b(N+M+P)” is the “(N+M+P)th bit” or “bit (N+M+P)”. [0039] During a conversion, a voltage input VIN 92 is sampled onto the DAC 62; then during a compare phase the DAC capacitors 110-119 are controlled to successively approximate the input voltage VIN 92 using the comparator 60 output to make decisions on hoW to sWitch the capacitors 110-119. At each step of the approximation, the comparator 60 output is stored in the SAR register 96 and the resulting digital Word (uncalibrated result 84) is the digital representa tion of the analog input voltage VIN 92. [0040] As the resolution of an SAR ADC 12 (see FIG. 2) increases, one of the major limitations is element matching