Classifications

• • H—ELECTRICITY
  • H03—ELECTRONIC CIRCUITRY
  • H03M—CODING; DECODING; CODE CONVERSION IN GENERAL
  • H03M1/00—Analogue/digital conversion; Digital/analogue conversion
  • H03M1/10—Calibration or testing
  • H03M1/1009—Calibration
  • H03M1/1033—Calibration over the full range of the converter, e.g. for correcting differential non-linearity
• • H—ELECTRICITY
  • H03—ELECTRONIC CIRCUITRY
  • H03M—CODING; DECODING; CODE CONVERSION IN GENERAL
  • H03M1/00—Analogue/digital conversion; Digital/analogue conversion
  • H03M1/12—Analogue/digital converters
  • H03M1/1205—Multiplexed conversion systems
  • H03M1/123—Simultaneous, i.e. using one converter per channel but with common control or reference circuits for multiple converters
• • H—ELECTRICITY
  • H03—ELECTRONIC CIRCUITRY
  • H03M—CODING; DECODING; CODE CONVERSION IN GENERAL
  • H03M1/00—Analogue/digital conversion; Digital/analogue conversion
  • H03M1/12—Analogue/digital converters
  • H03M1/14—Conversion in steps with each step involving the same or a different conversion means and delivering more than one bit
  • H03M1/16—Conversion in steps with each step involving the same or a different conversion means and delivering more than one bit with scale factor modification, i.e. by changing the amplification between the steps
  • H03M1/164—Conversion in steps with each step involving the same or a different conversion means and delivering more than one bit with scale factor modification, i.e. by changing the amplification between the steps the steps being performed sequentially in series-connected stages

Definitions

• ADC analog-to-digital converter
• DAC digital-to-analog converter
• Pipelined ADCs have been used extensively (for example) in high performance digital communication systems, waveform acquisitions, and instrumentations. While the speed of state-of-the-art pipelined ADC has exceeded 100 MSPS, the resolution is generally limited by the inter-stage gain error and/or DAC gain error resulting from circuit non-idealities (i.e., capacitor mismatch and finite operational amplifier (opamp) gain, and so forth). Thus, most pipelined ADCs with more than 12-bit resolution usually require some linearity enhancement techniques.
• ADC 100 generally comprises channels or ADCs 102-1 and 102-2, adders 104-1 and 104-2, and a divider 106.
• ADCs 102-1 and 102-2 have the same general structure, and, in operation, receive the same analog input signal AIN so as to perform a data conversion at approximately the same time (generating digital output signals DA and DB, respectively).
• An example embodiment accordingly, provides an apparatus.
• the apparatus comprises a plurality of pipelined analog-to-digital converters (ADCs), wherein each pipelined ADC is adapted to receive an analog input signal, and wherein each pipelined ADC has a transfer function that is adjustable, and wherein each pipelined ADC includes a compensator; and an adjustment circuit that is coupled to each pipelined ADC, wherein the adjustment circuit adjusts the transfer function for each pipelined ADC so as to generally eliminate an estimation ambiguity, and wherein the adjustment circuit estimates an inter-stage error that includes at least one of an inter-stage gain error and a digital-to-analog converter (DAC) gain error and adjusts the compensator for each pipelined ADC to compensate for the inter-stage error.
• ADCs analog-to-digital converters
• each pipelined ADC further comprises: a plurality of stages that are coupled to one another in a sequence; and a backend sub-ADC that is coupled to a last stage of the sequence.
• each compensator further comprises: a digital adder that is coupled to each stage of the sequence of its pipelined ADC; and a digital multiplier that is coupled between to the backend sub-ADC and the digital adder and that is coupled to the adjustment circuit, wherein the adjustment circuit adjusts the gain for the digital multiplier to compensate for the inter-stage error.
• each stage for each pipelined ADC further comprises: an input terminal; a sub-ADC that is coupled to the input terminal; a DAC that is coupled to the sub-ADC; an subtractor that is coupled to input terminal and the DAC; and a residue amplifier that is coupled to the subtractor.
• each sub- ADC further comprises a plurality of comparators that are adapted to be shifted so as to adjust the transfer function of its pipelined ADC.
• the adjustment circuit shifts the sub- ADC of the first stage of at least one of the pipelined ADCs by 1 ⁇ 4 of a least significant bit (LSB).
• LSB least significant bit
• each stage for the pipelined ADC further comprises an analog multiplier that is coupled between the input terminal and the sub-ADC, wherein gain of the analog multiplier is adjusted by the adjustment circuit.
• a method for calibrating a ADC having a first pipelined ADC and a second pipelined ADC comprising shifting a first set of comparators of a first sub-ADC of a first stage of the first pipelined ADC by a first amount to adjust a first transfer function of the first pipelined ADC; shifting a second set of comparators of a first sub-ADC of a first stage of the second pipelined ADC by a second amount to adjust a second transfer function of the second pipelined ADC; estimating an inter-stage error for the ADC once first set of comparators and the second set of comparators have been shifted, wherein the inter-stage error includes at least one of an inter-stage gain error and a DAC gain error; and adjusting a first compensator of the first pipelined ADC and a second compensator of the second pipelined ADC to compensate for the inter-stage error.
• the step of adjusting further comprises: adjusting a first gain of a first digital multiplier of the first pipelined ADC; multiplying a digital output from a first back-end sub-ADC of the first pipelined ADC by the first gain; adding a digital output for each stage of the second pipelined ADC and for the first digital multiplier together; adjusting a second gain of a second digital multiplier of the second pipelined ADC; multiplying a digital output from a second back-end sub-ADC of the second pipelined ADC by the second gain; and adding a digital output for each stage of the second pipelined ADC and for the second digital multiplier together.
• the method further comprises: estimating a gain mismatch between the first and second pipelined ADCs; estimating an offset mismatch between the first and second pipelined ADCs; and compensating for the gain and offset mismatches.
• the first and second amounts are 1 ⁇ 4 of an LSB.
• an apparatus comprising a first pipelined ADC having a first transfer function and having: a first track-and-hold (T/H) circuit that is adapted to receive an analog input signal; a first set of stages that are coupled to one another in a first sequence, wherein a first stage of the first sequence is coupled to the first T/H circuit, and wherein at least one of the stages from the first set of stages is adjustable so as to adjust a first transfer function; a first backend sub-ADC that is coupled to a last stage of the first sequence; and a first compensator that is coupled to each stage from the first set of stages and the first backend sub-ADC; a second pipelined ADC having a second transfer function and having: a second T/H circuit that is adapted to receive the analog input signal; a second set of stages that are coupled to one another in a second sequence, wherein a first stage of the second sequence is coupled to the second T/H circuit, and wherein at least one of the stages from
• the first compensator further comprises a first digital adder that is coupled to each stage from the first set of stages; and a first digital multiplier that is coupled between to the first backend sub-ADC and the first digital adder and that is coupled to the adjustment circuit.
• the first compensator further comprises: a second digital adder that is coupled to each stage from the second set of stages; and a first digital multiplier that is coupled between to the second backend sub-ADC and the second digital adder and that is coupled to the adjustment circuit.
• each stage from the first and second sets of stages further comprises: an input terminal; a sub- ADC that is coupled to the input terminal; a DAC that is coupled to the sub-ADC; an subtractor that is coupled to input terminal and the DAC; and a residue amplifier that is coupled to the subtractor.
• each sub- ADC further comprises a flash ADC having a plurality of comparators, wherein each of the comparators are adapted to be shifted.
• the adjustment circuit shifts the sub-ADC of the first stage of at least one of the pipelined ADCs by 1 ⁇ 4 of an LSB.
• the apparatus further comprises an output circuit that is coupled to the first and second digital adders.
• the first pipelined ADC further comprises a first mismatch compensator that is coupled between the first T/H circuit and the first stage of the first set of stages
• second pipelined ADC further comprises a second mismatch compensator that is coupled between the second T/H circuit and the first stage of the second set of stages
• the adjustment circuit estimate gain and offset mismatches between the first and second pipelined ADC and adjusts the first and second mismatch circuit.
• the adjustment circuit estimates the inter-stage error, the gain mismatch, and the offset mismatch using a least mean square (LMS) algorithm.
• LMS least mean square
• FIG. 1 shows an example of a conventional ADC
• FIG. 2 shows an example of an ADC in accordance with an example embodiment
• FIG. 3 shows an example of a pipelined ADC of FIG. 2
• FIGS. 4 and 5 show examples of a stage of FIG. 3
• FIG. 6 shows an example of the sub-ADC of FIGS. 4 and 5;
• FIG. 7 A and 7B depict adjustments for the transfer functions of the pipelined
• FIGS. 8A and 8B depict the spurious-free dynamic range (SFDR) of the ADC of
• FIG. 2 with and without calibration
• FIG. 9 depicts the convergence of a calibration method used by the ADC of FIG. DETAILED DESCRIPTION OF EXAMPLE EMBODIMENTS
• FIG. 2 illustrates an example ADC 200 that generally comprises pipelined ADCs
• each of ADCs 202-1 and 202-2 have generally the same structure and receive the analog input signal AIN so as to generate output signals D a and D b for the output circuit 206 (which can average these signals and can perform digital correction).
• the adjustment circuit 204 provides adjustments to the ADCs 202-1 and 202-2 to compensate for inter-stage gain errors and/or DAC gain errors (within ADCs 202-1 and 202-2) and gain/offset mismatches (between ADCs 202-1 and 202-2).
• pipeline ADC 202 generally comprises a compensator 314 (which generally includes a digital multiplier or digital gain block 312 and an adder or combiner 310) and a pipeline 301 (which includes a track-and-hold (T/H) circuit 302, a mismatch compensator 308, a set of stages 304-1 to 304-N coupled together in a sequence, and a backend sub-ADC 306).
• compensator 314 which generally includes a digital multiplier or digital gain block 312 and an adder or combiner 310
• pipeline 301 which includes a track-and-hold (T/H) circuit 302, a mismatch compensator 308, a set of stages 304-1 to 304-N coupled together in a sequence, and a backend sub-ADC 306).
• T/H track-and-hold
• the adjustment circuit 206 receives an output from the adder 310 (which combines the digital outputs from stages 304-1 to 304-N and backend sub- ADC 306) and should be able to compensate for inter-stage gain error and/or DAC gain error by perform adjustments (namely, adjusting the gain) to the digital multiplier 312 (which is coupled between the adder 310 and backend sub-ADC 306).
• the adder 310 which combines the digital outputs from stages 304-1 to 304-N and backend sub- ADC 306
• the digital multiplier 312 which is coupled between the adder 310 and backend sub-ADC 306.
• Di, a and Di, b are the output from the stages (i.e., 304-1), D 2 , a and D 2 , b are the output from the backend sub- ADCs (i.e., 306), and g a and gb are gains of the digital multipliers (i.e., 312). Because the output from the stages (i.e., 304-1) Di ,a and Di , b should be the same, the difference AD would then be:
• adjustment circuit 204 can adjust the transfer function of each of the pipeline ADCs (i.e., 204-1 and 204-2 of FIG. 2), and there are several ways to adjust these transfer functions.
• FIGS. 4 and 5 illustrate examples of one of the stages 304-1 to 304-N (hereinafter
• Stage 304-A and 304-B for FIGS. 4 and 5, respectively) which can be adjusted by the adjustment circuit 204 can be seen.
• Stage 304-A generally comprises a sub-ADC 402-1, a DAC 404, an adder 408 (which operates as a subtractor), and a residue amplifier 406, while stage 304-B includes sub-ADC 402-1 and analog multiplier 502.
• the an analog signal (either from the T/H circuit 302 or a previous stage) is converted by sub-ADC 402-1 (or 402-1) to a digital signal. This digital signal is provided to adder 310 and DAC 404.
• the DAC 404 converts the signal back to an analog signal, and the analog signal from the DAC 404 (which can introduce a DAC gain) is subtracted from the analog signal from the T/H circuit 302 or a previous stage by adder 408 to generate a residue signal.
• This residue signal is amplified by residue amplifier 406 (which can introduce an inter-stage gain).
• sub-ADC 402-1 or 402-2 can be one of a variety of types of ADCs but is typically a flash ADC (as shown).
• This flash ADC 402 generally comprises a voltage divider 604 (which generally includes resistors Rl to R(M+1) coupled in series with one another) and comparators 602-1 to 602-M.
• each comparator 602-1 to 602-M is coupled to voltage divider 604 and receives an analog input signal to generate a digital output signal.
• adjustments to the transfer function can be accomplished by making direct adjustments to the sub-ADC 402-1.
• adjustments to the transfer function can be performed by shifting the comparators 602-1 to 602-M (i.e., shifting the reference voltage REF) within sub-ADC 402-1 with an adjustment signal ADJ.
• the transfer functions for each of pipeline ADCs 202-1 and 202-2 are supposed to match (as shown in FIG. 7A) match, but to resolve the estimation ambiguity described above, the comparators 602-1 to 602-M for one or more of the stages each of pipeline ADCs can be shifted by predetermined amounts. For example and as shown in FIG.
• the comparators 602-1 to 602-M for a the first stage (i.e., 304-1) for pipeline ADC 202-1 can be shifted by +1 ⁇ 4 of a least significant bit (LSB), while the comparators 602-1 to 602-M for a the first stage (i.e., 304-1) for pipeline ADC 202-2 can be shifted by -1 ⁇ 4 of an LSB. By doing this, however, some resolution in digital redundancy is lost.
• LSB least significant bit
• the analog signal from the T/H circuit 302 or from the previous stage can be modified before being converted.
• a multiplier 502 can be included in the signal path.
• the adjustment circuit 206 can provide a gain MUL (or alternatively a signal) to achieve substantially the same goal as directly adjusting ADC 402-1 described above.
• a weighted difference 5D can be defined as follows using equations (8) and (9) above:
• D b K b (D l b + ga D 2 b ) + OS b .
• K(k + 1) K(k) - K VJ K ,
• the adjustment circuit 204 can estimate (and compensate for) inter-stage gain error, DAC gain error and gain/offset mismatches at about the same time.
• FIGS. 8 A through 9 show the results of some example simulations.
• each of pipelined ADCs 202-1 and 202-2 is a 16-bit pipelined ADC having four stages.
• Each of the four stages in the pipeline for this example has 4, 5, 5, and 5 bits for each stage, respectively.
• the optimal inter-stage gains are 8 and 16 for the first and second stages.
• both pipelines have gain errors in the first two stages, and the gain implemented for pipelined ADCs 202-1 and 202-2 are shown in Table 1 below.
• FIGS. 8 A and 8B the SFDRs before and after the calibration are shown, and it can be observed that the SFDR is improved from 74dB to 113dB.
• the convergence curve of the digital multipliers (i.e., 312) for each of pipelined ADCs 202-1 and 202-2 can be seen in FIG. 9, showing a convergence at about 40,000 samples when using a 1/10 LSB shift instead of 1 ⁇ 4 LSB shift (that would result in a longer convergence time), which is about 100 times faster than any other known approach and which is not restrictive (unlike some energy-free approaches).

Landscapes

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

Priority Applications (2)

Applications Claiming Priority (2)

Publications (2)

Family

ID=46652285

Family Applications (1)

Country Status (4)

Families Citing this family (25)

Family Cites Families (15)

• 2011
  • 2011-02-22 US US13/032,457 patent/US8451152B2/en active Active
• 2012
  • 2012-02-22 WO PCT/US2012/026022 patent/WO2012116006A2/en not_active Ceased
  • 2012-02-22 JP JP2013555511A patent/JP6076268B2/ja active Active
  • 2012-02-22 CN CN201280009792.8A patent/CN103392297B/zh active Active

Also Published As

Similar Documents

Legal Events