ANALOG SIGNAL SAMPLING SYSTEM AND METHOD HAVING REDUCED AVERAGE INPUT CURRENT
[0001] This application claims priority of U.S. provisional application No. 60/619,007 filed on 'October 18, 2004, entitled "SAMPLING CONFIGURATION WITH REDUCED AVERAGE INPUT CURRENT, " and incorporated herein by- reference.
Related Applications
[0002] U.S. patent application No. , filed on , entitled " ANALOG SIGNAL SAMPLING SYSTEM
AND METHOD HAVING REDUCED AVERAGE DIFFERENTIAL INPUT
CURRENT" and incorporated herein by reference.
[0003] U.S. patent application No. , filed on , entitled "ANALOG-TO-DIGITAL CONVERTER WITH REDUCED INPUT AVERAGE CURRENT AND REDUCED AVERAGE
REFERENCE CURRENT" and incorporated herein by reference. . •
Technical Field
[0004] This disclosure relates to circuitry and methodology for sampling an analog input signal to . reduce an average charge taken from an input signal
source, one application of which is in analog-to- digital (A/D) converters.
Background Art [0005] An exemplary but non-limiting application of the teachings presented herein is in signal sampling for A/D conversion. A typical analog-to digital converter samples an analog input signal in order to convert it into a corresponding digital signal. During this process, the converter loads the input signal and modifies it depending upon the impedance of the signal source. Such a modification directly influences the accuracy of the conversion process and the final result. For slower-speed and lower-resolution converters, errors caused by the input signal modification are insignificant and may be safely ignored.
[0006] On the other hand, recent developments in sensor technology, improvements in converter resolution and converter speed have made such errors quadratic factors limiting further increase in conversion accuracy. Moreover, the tendency to reduce power consumption and the expansion of portable applications has spread the use of a variety of sensors with relatively high source impedances. Examples of such sensors are high-value resistive bridges used to monitor weight and pressure.
[0007] At the same time, the development of the over-sampling converter technology has pushed the resolution of the analog-to-digital conversion to a 24- bit level and higher. Typical over-sampling converters use switched-capacitor front end circuits including one or more sampling capacitors to sample an analog input
signal multiple times for each conversion cycle. During each sampling process, a certain amount of charge is transferred between the signal source and the converter front end capacitors resulting into an equivalent input current flow. As this input current passes through the signal source impedance, it causes a voltage change, modifying the original input value and creating a sampling error.
[0008] The value of the input current is directly proportional to the size of the sampling capacitors and to the sampling rate. Due to thermal noise limitations, an increase in the conversion resolution requires a quadratic increase in the size of the sampling capacitors resulting in the corresponding quadratic increase in the input current. At the same time, any increase in the overall conversion rate causes a proportional increase in the input signal sampling rate, resulting in the increased input current.
[0009] Two different strategies are typically used to deal with this problem. The first approach is to guarantee the complete settling (within the accuracy of the converter) of the front end sampling circuit including the input signal source impedance. This is a very difficult goal to achieve and it rapidly becomes impractical as the desired conversion accuracy and speed increase. The source impedance of a sensor imposes a theoretical limit on available ranges of conversion speed and resolution. Unavoidable parasitic capacitors and necessary signal filter capacitors involved in practical configurations further limit these ranges. An example of this approach is the LTC®2410 analog-to-digital converter developed by
Linear Technology Corporation, assignee of the present subject matter.
[0010] The second approach uses isolation buffers 90 and amplifiers interposed between the sensor and the converter. Such buffers can be external to the converter or may be integrated within the converter front end sampling circuits. Configurations using external buffers offer great flexibility but place an 95 unacceptable heavy burden upon the user in order to maintain the global accuracy of the measurement chain. These configurations also demand supplemental power supply rails, critical power supply sequencing circuits and an additional physical space. Integrating the
100 buffers within the converter front end sampling circuits partially resolves these issues. Nevertheless, the integrated buffers limit the analog- to-digital converter overall accuracy and dynamic range. An example of this approach is the LTC®2442
105 analog-to-digital converter developed by Linear
Technology Corporation, assignee of the present subject matter.
[0011] Therefore, there is a need for a new sampling technique to reduce average input current taken from an
110 input signal source.
Summary of the Disclosure
[0012] The present disclosure offers novel signal sampling system and methodology. In accordance with one aspect of the disclosure, a sampling system 115 includes a sampling device responsive to an input signal and a reference signal for providing corresponding charges, and a switching circuit controllable to supply the input signal and the
reference signal to the sampling device. The switching
120 circuit is controlled to supply the input signal and the reference signal to the sampling device so as to provide a substantially zero total charge taken by the sampling device from a source of the input signal. [0013] In exemplary applications of the principles
125 taught herein to A/D conversion, the proposed sampling architecture significantly reduces the average charge taken from the input signal source without substantially altering the instantaneous charge transferred between the input source and a modulator.
130 A charge storage device (e.g. A capacitor) can be connected at the converter input in order to supply this instantaneous charge as required by the modulator. In the same time the average current flowing through the internal resistance of the input signal source
135 (e.g. The sensor) is minimal. Consequently the effect of the input source resistance upon the measurement accuracy is greatly diminished.
[0014] The proposed architecture is based upon the observation that certain types of over-sampling
140 converters balance the charge transferred from the input signal with charge taken from the reference signal. Thus, by properly sequencing the input signal sampling.sequence, charge is transferred to and from the input signal source such that the total net charge
145 transfer is substantially zero.
[0015] In particular, a switching device may be coupled to terminals supplying an input signal and a pair of reference signals. The switching device may be controlled to provide a sampling device with the input
150 signal and one of the reference signals in each sampling phase.
[0016] in accordance with an embodiment of the disclosure, the switching circuit may be controlled by a switch controller that may produce multiple control
155 signals for controlling respective switches of the switching circuit.
[0017] A modulator, e.g., a delta-sigma modulator, may be coupled to the sampling device to produce a digital output signal. The modulator may comprise an
160 integrator responsive to the charges produced by the sampling device, and an output device, such as a comparator, for producing the digital output signal having a value determined by the output signal of the integrator. The switch controller, may produce a clock
165 signal for controlling the output device of the modulator.
[0018] Among N sampling, Ni sampling operations may result in a digital output signal having a first value, and (N - Ni) sampling operations may result in the
170 digital output signal having a second value. The switch controller may determine the number Ni of sampling operations among the total number N of sampling operations, and control the switching circuit to make the number Ni proportional to a ratio between
175 the input signal and the reference signal.
[0019] In accordance with a method of the present disclosure the following steps may be carried out for sampling an input analog signal:
- supplying a sampling device via a switching circuit 180 with the input analog signal and a reference signal, and
- controlling the switching circuit to provide a substantially zero total charge taken by the sampling device from a source of the input signal. The
185 switching circuit may be controlled during multiple successive sampling operations performed upon the input signal.
[0020] ι In accordance with another aspect of the present disclosure, a method of sampling an input
190 signal includes the steps of:
- determining number Ni of sample sequences, in which a first reference signal should be supplied, among N sample sequences,
- supplying the input signal and the first reference 195 signal in the Ni sample sequences, and
- supplying the input signal and a second reference signal in (N - Ni) sample sequences.
[0021] The number Ni may be determined based on the output digital signal to minimize total charges taken
200 from a source of the input signal. The output digital signal may have a first value in response to the Ni sample sequences, and may have a second value in response to the (N - Ni) sample sequences. [0022] In accordance with another aspect of the
205 disclosure, a system for reducing an average differential input current of a sampling circuit comprises a sampling device for sampling a differential analog input signal, and a switching circuit for supplying the differential input signal and a
210 differential reference signal to the sampling device. The switching circuit is controlled during multiple sampling operations so as to provide a substantially zero total differential charge taken by the sampling device from a source of the differential input signal.
215 [0023] In accordance with an embodiment of the disclosure, the system may further include a differential-in/differential-out voltage amplifier
responsive to an output signal of the sampling device to produce a differential output signal.
220 [0024] Additional advantages and aspects of the disclosure will become readily apparent to those skilled in the art from the following detailed description, wherein embodiments of the present disclosure are shown and described, simply by way of
225 illustration of the best mode contemplated for practicing the present disclosure. As will be described, the disclosure is capable of other and different embodiments, and its several details are susceptible of modification in various obvious 230 respects, all without departing from the spirit of the disclosure. Accordingly, the drawings and description are to be regarded as illustrative in nature, and not as limitative.
Brief Description of the Drawings
235 [0025] The following detailed description of the embodiments of the present disclosure can best be understood when read in conjunction with the following drawings, in which the features are not necessarily drawn to scale but rather are drawn as to best 240 illustrate the pertinent features, wherein:
[0026] FIG. 1 is a diagram that illustrates the charge taken from an input signal source in a sample operation of the present disclosure.
[0027] FIG. 2 is a diagram illustrating sampling of 245 an input signal and a pair of reference signals in accordance with the present disclosure.
[0028] FIGs. 3A and 3B are diagrams illustrating A/D conversion arrangements of the present disclosure.
[0029] FIG. 4 is a diagram illustrating an 250 embodiment of an A/D converter of the present disclosure.
[0030] FIG. 5 is a diagram illustrating a system for sampling a differential input signal so as to reduce an average differential input current.
255 Detailed Disclosure of the Embodiments
[0031] The present disclosure will be made using the example of an over-sampling analog-to-digital (A/D) converter. It will become apparent, however, that the concepts described herein are applicable to any type of
260 circuit that implements sampling of analog signals. [0032] The instantaneous charge required from an input signal source in any single sampling operation is proportional to the size of the sampling capacitors as well as to the amount of charge stored in these
265 capacitors prior to the sampling operation. For example, as shown in FIG. 1, one terminal of a sampling capacitor C is connected to a reference voltage, e.g., to ground. The second terminal may be supplied through a switch Si with a voltage Vi or through a switch S2
270 with a voltage V2. The voltages Vi and V2 are defined with respect to the established ground level. [0033] Assuming that initially the switch Si is closed and the switch S2 is open the charge Qi stored in the sampling capacitor C is:
[0034] In the second phase, the switch Si is open and subsequently the switch S2 is closed. At the end of
this process the sampling capacitor C will have an accumulated charge Q2:
280 Q2 = V2*C.
[0035] During this sampling operation, the signal source V2 provides a charge amount dQ which can be calculated as :
285 [0036] In another example, illustrated in FIG. 2, the sampling capacitor C has a first terminal connected to ground, and a second terminal supplied with an input voltage Vi through the switch Si, with a first reference voltage VL through the switch SL, and with a second
290 reference voltage VH through the switch SH. At any- given time only one of the three switches SL, SI and SH is closed, while the remaining two are open. [0037] A first sampling sequence of the input voltage Vi starts with the switch SL closed in the first
295 phase and continues with the switch Si closed in the second phase. The amount of charge dQL taken from the input signal Vi is:
[0038] A second sampling sequence of the input 300 voltage Vi starts with the switch SH closed in the first phase and continues with the switch Si closed in the second phase. The amount of charge dQH taken from the input signal Vi in the second sampling sequence is:
dQH = (Vi - VH) *C.
305 [0039] It may be assumed that in a set of N consecutive sample sequences of the input signal Vi, Ni sample sequences are of a first type and the remaining No = (N - Ni) sample sequences are of a second type. In particular, the first type of the sample sequences may
310 be associated with supplying the reference voltage VH, together with the input signal Vi, and the second type may be associated with supplying the reference voltage VL, together with the input signal Vi. Therefore, the total charge dQw taken from the input signal source Vi
315 during these N consecutive sample sequences is:
dQN = Ni* (Vi - VH) *C + No* (Vi - VL) *C dQN = Ni* (Vi - VH) *C + (N - Ni) * (Vi - VL) *C dQN = N* (Vi - VL) *C - Ni* (VH - VL) *C
If the following condition is imposed: 320
VH >= VI si= VL (1)
than Nl can be selected as the digital representation of the input signal VI with respect to the reference 325 signal VH - VL with a resolution of N counts. This relation can be written as:
Nl = N * (VI - VL) / (VH - VL) (2)
330 Using. this value of Nl in the above calculation of dQN we obtain:
dQN = 0.
This result is independent of the order and succession 335 of the two types of sample sequences within the set N.
[0040] The accuracy of this relation is limited by the quantization accuracy of the input signal VI with respect to the reference signal VH - VI within an N counts representation. Thus, while systems and methods
340 according to the invention reduce the current drawn from the analog signal source to substantially zero, . nevertheless, some current is drawn from the analog signal source. The amount of current drawn from the analog signal source is preferably within the limits
345 set by the quantization accuracy of the input signal. [0041] It has been shown that, within the limitations of (1) and using prior or concomitent knowledge of the magnitude of VI with respect to VH and VL as expressed by (2) , the proposed strategy reduces
350 significantly the average charge required from the input signal source. This reduction is proportional with the resolution of the digital representation of the input signal (2) and it is particularly useful in high resolution analog-to-digital converters.
355 [0042] An immediate implementation of this proposal is shown in FIG. 3A. One terminal of the sampling capacitor #10 is connected to the input signal terminal VI and the reference signal terminals VH and VL through the analog switch block #20. The other terminal of the
360 sampling capacitor #10 is connected to the "High accuracy converter" #30.
[0043] The magnitude of input signal VI is evaluated with respect to the reference signals VH and VL by the "Low accuracy ADC" #50. This converter can be
365 implemented using a variety of well known analog-to- digital conversion techniques and, because of its relative low accuracy with respect to high-accuracy converter 30, it does not present a substantial load to
the input signal VI. The analog-to-digital converter
370 #50, using the input signal VI and the reference signals VH and VL produces an equivalent digital representation DLA of input signal VI. The digital signal DLA is a serial binary stream as described by equation (2) and has a resolution of N counts.
375 Depending upon the conversion method used by the converter #50, this stream can be produced directly or converted from a parallel format through common digital techniques. [0044] The DLA data stream is used by switch
380 controller #40 to direct the operation of analog switch block #20. During each sample operation, the analog switch #20 connects the sampling capacitor #10 in' two successive phases to one of the reference terminals VH and VL and to the input signal terminal VI. Switch
385 controller #40, using the information contained in the digital data stream DLA, selects the appropriate sampling sequence such that the total charge taken from the VI signal source during the conversion process is substantially zero.
390 [0045] The "High accuracy" converter #30 uses the charge sampled on capacitor #10 during a minimum of N succesive sampling steps together with the sampling sequence information containded in the DLA data stream to produce the output data Dout. Dout is a high
395 accuracy representation of the input signal VI.
[0046] In the above description the "High accuracy" and "Low accuracy" are relative terms that relate directly to the different potential loadings of the input signal (corresponding to the larger capacitors
400 required for higher accuracy resolution as described above) by the two converters. The terms "low accuracy"
and "high accuracy" as defined herein are intended only to describe the relative relationship of the two analog-to-digital converters and are not intended to
405 limit the scope of the invention, or of either of the converters, to any particular objective accuracy range. [0047] The operation of converters #30 and #50 can be simultaneous and synchronized or converter #50 can produce its output at any time prior to the utilization
410 of the output in the sampling process.
[0048] The implementation of analog-to-digital converters, sampling capacitors, analog switches and switch controllers is well known and widely described in the technical literature. The sampling capacitor,
415 shown as a single device in fig.3a, may be in an actual implementation a set of capacitors which simultaneously perform the input and reference sampling operations required by the over-sampling converter as well as additional scaling and calibration functions. Similarly
420 the analog switch may be implemented using multiple physical switches in various parallel and serial configurations supporting simultaneous sampling, scaling and calibration functions. [0049] Furthermore a single "High accuracy"
425 converter #30 may be connected to multiple sampling capacitors C and receive each of the respective corresponding data streams DLA where each capacitor and its coresponding "Low accuracy" converter producing the DLA samples a distinct input signal. The converter
430 #30 preferably combines each of the multiple capacitors respective charges in the analog domain and produces Dout as a digital representation of the ratio of the multiple input signals.
[0050] An over-sampling converter can greatly
435 benefit from this sampling configuration being able to perform simultaneous the function of both converter #50 and converter #30 of FIG. 3A. During the conversion process such a converter samples the input signal N times (where N is the over-sample ratio) and generates
440 a stream of digital data which is subsequently- processed in order to obtain the conversion result. This stream of digital data contains information about the ratio between the input signal and the reference signal and such information can be used to control the
445 sampling sequences in order to substantially reduce the average charge taken from the input signal source. [0051] Another proposed configuration according to the invention is shown in FIG. 3B. FIG. 3B is a diagram illustrating components of an A/D converter 10
450 of the present disclosure that comprises an analog switch 22, a sampling capacitor C, an over-sampling converter 32 and a switch controller 42. The analog switch 22 connects one terminal of the sampling capacitor C to an input signal terminal Vi and reference
455 signal terminals VH and VL. The other terminal of the sampling capacitor C is connected to the over-sampling converter 32.
[0052] A conversion cycle of the over-sampling converter 32 that produces an output data stream Dout
460 consists of a set of N successive sample operations where N is the over-sampling ratio. The output data stream Dout, which is a digital representation of the ratio between the input signal and the reference signal, is supplied to the switch controller 42 to
465 direct the operation of the analog switch 22. During each sample operation, the analog switch 22 connects
the sampling capacitor C in two successive phases to one of the reference signal terminals VH and VL and to the input signal terminal Vi. The switch controller 42
470 uses the information contained in the digital data stream Dout to select the appropriate sampling sequence such that the total charge taken from the Vi signal source is substantially zero. [0053] One skilled in the art would realize that the
475 analog switch 22, over-sampling converter 32, sampling capacitor C, and switch controller 42 may be implemented using various arrangements. For example/ the sampling capacitor C may be represented by a set of capacitors which simultaneously perform the input and
480 reference sampling operations required by the over- sampling converter as well as additional scaling and calibration functions. Similarly, the analog switch 22 may be implemented using multiple physical switches in various parallel and serial configurations supporting
485 simultaneous sampling, scaling and calibration functions.
[0054] An exemplary implementation of an A/D converter 100 of the present disclosure is shown in FIG. 4. The A/D converter 100 comprises a sampling
490 capacitor Q, a voltage amplifier 110 having an integrating capacitor Cf in its feedback loop, a comparator 120, and a switch controller 140. An analog switch Si is provided for supplying a reference voltage VR to a first terminal of the sampling capacitor Ci. An
495 analog switch S2 is arranged for connecting an input voltage Vi to the first terminal of the sampling capacitor Ci. For simplicity, a ground potential is selected as the second reference voltage (VL = 0) . An analog switch S3 connects this ground potential to the
500 first terminal of the sampling capacitor Ci. Analog switches S* and Ss are provided to connect a second terminal of the sampling capacitor Ci to a ground terminal, and the amplifier 110, respectively. [0055] The integrating capacitor Cf, voltage
505 amplifier 110, comparator 120 and analog switches S4 and S5 represent a first-order delta-sigma modulator. The amplifier 110, together with the capacitors Ci and Cf, and the switches S4 and Ss, represent a switched- capacitor implementation of an analog integrator. For
510 simplicity of explanation, a ground potential is selected as common mode voltage references for the integrator and comparator circuits.
[0056] The output of the amplifier 110 is connected to the comparator 120 controlled by a clock signal CIk
515 generated by an internal clock of the switch controller 140 to produce a single-bit output digital signal Dout. In particular, when the comparator 120 is triggered by the CIk signal, it produces output digital value Dout = 1 if the output of the amplifier 110 is
520 positive, and output value Dout = 0 if the output of the amplifier 110 is negative.
[0057] The data signal Dout is used by the switch controller 140 to control operations of the first-order delta-sigma modulator. In particular, the switch
525 controller 140 produces switch drive signals Si to Ss for controlling the respective switches. Further, the data stream Dout may be processed by a digital filter (not shown) connected to the output of the comparator 120 to calculate the conversion result.
530 [0058] Based on the data signal Dout and an internal clock signal produced by the internal clock, the switch controller 140 controls the first-order delta-sigma
modulator to maintain the total charge taken from the input signal source substantially zero for N 535 consecutive sample operations of one conversion cycle.
[0059] Sampling operations of the A/D converter 100 are described below. Every pulse of the internal clock signal starts the following two-phase sampling operation sequence: 540 1. Trigger the voltage comparator 120 using the clock signal CIk.
2. Open switch Ss.
3. Open switch S2.
4. Close switch S4. 545 5. If Dout = 0 close switch S3, if Dout = 1 close switch
Si.
6. Wait for settling of the first phase samples.
7. Open switch Si.
8. Open switch Si and S3. 550 9. Close switch Ss.
10. Close switch S2.
11. Wait for settling of the second phase samples. [0060] The time allocated for each one of the above steps may be determined in accordance with a particular
555 switched capacitor implementation.
[0061] The amount of charge QM transferred into the integrating capacitor Cf during such a sampling sequence is:
560 when Dout = 1, QM = (Vi - VR) *Ci.
[0062] The amount of charge dQ taken from the input signal source Vi during such a sampling sequence is:
when Dout = 1, dQ = (Vi - VR)*Ci.
565 [0063] It is assumed that among N consecutive sampling operations in a sampling process, for Ni sampling operations Dout = 1 and for No = (N - Ni) sampling operations Dout = 0. Hence, in each of Ni sampling operations, the reference voltage VR is
570 supplied together with the input voltage Vi, and in each of No sampling operations, the reference voltage VL (set to a ground potential in this particular example) is supplied together with the input voltage Vi. Therefore, the total charge QMTOT transferred into the integrating
575 capacitor Cf during the set of N consecutive sampling operations is:
QMTOT = No*Vi*Ci + Ni* (Vi - VR) *Ci = (N - Ni) *Vi*Ci +
QMTOT = (N*Vi - Ni*VR)*Ci.
580 [0064] The delta-sigma modulator operates so as to minimize the total charge accumulated in the integrator. Hence, within the resolution of the modulator,
if QMTOT = 0, Ni = N*VI/VR.
585 [0065] Hence, the ratio Ni/N representing the density of "1" bits contains information on the ratio VI/VR between the input signal and the reference signal. Therefore, the digital output data stream Dout provides this information to the switch controller 40.
590 Accordingly, the total charge taken from the input signal source can be calculated as:
dQτoτ = Nc,*Vi*Ci + Ni* (Vi - VR) *Ci = (N*Vi - NI*VR) *Ci
and within the resolution of the converter
if N,=N*VI/VR dQTOT = 0.
595 Therefore, the total charge taken from the input signal source is substantially zero.
[0066] One skilled in the art would understand that the concept of the present disclosure is also
600 applicable to higher-order modulators that: may have additional integrator stages between the output of the amplifier 110 and the input of the comparator 120. [0067] In addition, one skilled in the art would realize that the disclosed technique is not limited to
605 over-sampling converters producing single-bit digital data streams. It may also be applicable to converters generating multi-bit output data streams. In this case, multi-bit output data streams may be converted into multiple single bit data streams such as binary
610 weighted or thermometer encoded streams, which may be used to control multiple equivalent weighted input signal sampling capacitors.
[0068] Further, the disclosed sampling front-end configuration with reduced average input current may be
615 integrated with other well known delta-sigma modulators, such as MASH or band-pass modulators. [0069] FIG. 5 illustrates an A/D converter 200 having a differential front-end sampling configuration
that substantially reduces an average differential 620 input current. The A/D converter 200 comprises 16 analog switches Soi to Sis for supplying to first terminals of 4 sampling capacitors Ci to C4 a differential input signal defined by voltages VIP, VIN, and a differential reference signal defined by voltages 625 VRP, VEN. Second terminals of the sampling capacitors Ci to CA are connected to a differential first-order delta- sigma modulator including 12 analog switches S21 to S35, a differential-in/differential-out voltage amplifier 210, integrating capacitors C21 and C22, and a
630 voltage comparator 220.
[0070] The voltage amplifier 210 has a pair of inputs for supplying a differential input signal, and a pair of outputs for producing a differential output signal. Together with integrating capacitors C21 and
635 C22, the voltage amplifier 210 forms a differential integrator circuit. The voltage comparator 220 is controlled by a clock signal CIk generated by a switch controller 210 to produce a one-bit digital data stream Dout . In particular, when the comparator 120 is
640 triggered by the CIk signal, it may produce output digital value Dout = 1 if the differential output of the amplifier 210 is positive, and output digital value Dout = 0 if the output of the amplifier 210 is negative. The data signal Dout is used by the switch
645 controller 240 to produce switch drive signals Soi to Sis and S21 to S35 that control the respective switches Soi to Sis and S21 to S3s. A control sequence generated by the switch controller 240 using its internal clock signal provides N consecutive samples of the input
650 differential signal while maintaining the total differential charge taken from the input signal source
at a substantially zero level. Therefore, the switch controller 240 controls the sampling sequence to reduce the average input differential current of the A/D
655 converter 200 to a substantially zero value.
[0071] Although the illustrated A/D converter 200 contains a first-order delta-sigma modulator, one skilled in the art would realize that the concept of the present disclosure is applicable to any modulator.
660 Further, to simplify explanation of the present concept, the A/D converter 200 is shown with the input common mode voltage of the amplifier 210 set to a ground potential. [0072] Sampling operations of the A/D converter 200
665 are described below. At every pulse of the internal clock signal, the switch controller 240 carries out the following sequence composed of 8 sampling phases:
I. Trigger the voltage comparator 220 using signal CIk.
670 2. Open switches S2i, S25, S30, S34.
3. Open switches Soi, So*, Soe, Sn, Si3, Sis.
4. Close switches S23, S27, S31, S35.
5. Close switches Sos, Sn.
If Dout = 0, close switches So4, S13. 675 If Dout = 1, close switches Soi, Sie.
6. Wait for settling of phase 1 samples.
7. Open switches S23, S27, S3i, S3s.
8. Open switches Soi, S04, Soe, Sn, Si3, Si6.
9. Close switches S21, S26, S29, S34. 680 10. Close switches S02, Sis.
If Dout = 0, close switches Soβ, Sos. If Dout = 1, close switches Sos, S12.
II. Wait for settling of phase 2 samples. 12. Open switches S21, S26, S29, S34.
685 13. Open switches S02, Sos, Soβ, So9, S12, Sis.
14. Close switches S23, S27, S31, S3s.
15. Close switches So2, Sis.
If Dout = 0, close switches So5, S12.
If Dout = 1, close switches Soβ, S09.
690 16. Wait for settling of phase 3 samples.
17. Open switches S23, S27, S31, S3s.
18. Open switches S02, Sos, Soβ, Sos, S12, Sis.
19. Close switches S22, S26, S29, S33.
20. Close switches S07, S10.
695 If Dout = 0, close switches So*, S13. If Dout = 1, close switches S01, Sis.
21. Wait for settling of phase 4 samples.
22. Open switches S22, S26, S29, S33.
23. Open switches S01, So4, So?, S10, S13; Sis. 700 24. Close switches S23, S27, S31, S35.
25. Close switches S07, S10.
If Dout = 0, close switches S01, Sis. If Dout = 1, close switches S04, S13.
26. Wait for settling of phase 5 samples. 705 27. Open switches S23, S27, S31, S3s.
28. Open switches S01, S04, S07, S10, S13, Sis.
29. Close switches S22, S25, S30, S33.
30. Close switches S03, SM.
If Dout = 0, close switches Sos, S12. 710 If Dout = 1, close switches Soβ, S09.
31. Wait for settling of phase 6 samples.
32. Open switches S22, S25, S30, S33.
33. Open switches S03, Sos, Soβ, So9, S12, Si*.
34. Close switches S23, S27, S31, S3s. 715 35. Close switches So3, Si4.
If Dout = 0, close switches Soβ, S09. If Dout = 1, close switches Sos, S12.
36. Wait for settling of phase 7 samples.
37. Open switches S23, S27, S31, S35.
720 38. Open switches S03, Sos, Sos, S09, S12, Si4.
39. Close switches S21, S25, S30, S34.
40. Close switches Soe, Sn.
If Dout = 0, close switches S01, Sie.
If Dout = 1, close switches So4, S13. 725 41. Wait for settling of phase 8 samples.
[0073] The time allocated for each one of the above steps is determined in accordance with a particular switched capacitor implementation.
[0074] The amount of differential charge QD 730 transferred into the integrator presented above is:
if Dout = 0 , QD = - 2 * (Ci + C2 +C3 + C4) * [ (VIP - VIN) + (VRP ; if DOUt = 1 , QD = - 2 * (Ci + C2 +C3 + C*) * [ (VIP - VIN) - (VRP . 735 If :
Then :
740 if Dout = 0, QD = 2*Ceq*(- R - V), if Dout = 1, QD = 2*Ceq*(R - V) .
[0075] The total amount of charge dQp taken from the input signal source positive terminal VIP during the sampling sequence is :
745 if Dout = 0, dQp = Ceq* (VIP - VRN) , if Dout = 1, dQp = Ceq* (VIP - VRP) .
[0076] The total amount of charge dQw taken from the input signal source negative terminal Vm during the sampling sequence is:
750 if Dout = 0, dQN = Ceq* (Vm - VRP) , if Dout =1, dQN = Ceq* (VIN - Vrø) .
[0077] It can be assumed that a set of N consecutive sampling operations representing one conversion cycle comprises Ni operations for which Dout = 1, and 755 Mo = (N - Ni) operations for which Dout = 0. Thus, the total differential charge transferred into the integrator during the N consecutive sample operations is :
QDTOT = 2*Ceq* [Ni* (R-V) + (N - Ni) * (- R - V) ]
760 [0078] Since the modulator control loop operates so as to minimize the charge accumulated by the integrator, within the resolution of the system, it can be stated that:
if QDTOT = 0, NI = 0.5*N*(l + V/R) = 0.5*N*[l + (Vip -
[0079] This relation shows that the Ni/N ratio representing the density of "1" bits contains information about the ratio V/R between the differential input signal and the differential 770 reference signal. The digital output data stream Dout provides this information to the switch controller 240 for producing proper switch drive signals. The total charge dQpτoτ taken from the input signal source positive
terminal VIP during a set of N consecutive sample 775 operations is equal to:
dQpxoT = Ceq* [N*VIP - Ni* VRP - (N - Ni) *VRN] dQpTOT = Ceq* 0 . 5*N* [ (VIP + VIN) - (VRP + VRN) ] .
[0080] Similarly, the total charge dQuτoτ taken from the input signal source negative terminal VIN during a 780 set of N consecutive sample operations is equal to:
dQmoT = Ceq* [N* VIN - Ni* VRN - (N - Ni) *VRP] dQmoT = Ceq* 0 . 5*N* [ (VIP + VIN) - (VRP + VRN) ] .
[0081] Since dQpτoτ = dQ.Nτoτ, the average differential current taken from the input signal source including a
785 set of N consecutive sample operations is zero within the resolution of the system. Hence, the switch controller 240 controls the sampling procedure to make dQpTOT = dQuTOT. Accordingly, the sampling technique of the present disclosure reduces the average input
790 differential current of the A/D converter 200 to a zero level .
[0082] One skilled in the art would realize that the disclosed sampling configuration may be used not only with single-bit output data streams Dout but also with
795 multi-bit output data streams, and may be integrated with higher-order modulators by inserting additional integrator stages between the output of the differential integrator 210 and the input of the . comparator 220. Further, the disclosed system for
800 reducing an average input differential current may also be implemented using other delta-sigma modulators such as MASH modulators .
[0083] The foregoing description illustrates and describes aspects of the present invention.
805 Additionally, the disclosure shows and describes only preferred embodiments, but as aforementioned, it is to be understood that the invention is capable of use in various other combinations, modifications, and environments and is capable of changes or modifications
810 within the scope of the inventive concept as expressed herein, commensurate with the above teachings, and/or the skill or knowledge of the relevant art. [0084] The embodiments described hereinabove are further intended to explain best modes known of
815 practicing the invention and to enable others skilled in the art to utilize the invention in such, or other, embodiments and with the various modifications required by the particular applications or uses of the invention.
820 [0085] Accordingly, the description is not intended to limit the invention to the form disclosed herein. Also, it is intended that the appended claims be construed to include alternative embodiments.