Switched-capacitor circuits and methods with improved settling time and systems using the same

[0002] This application for patent is related to the following applications for patent:

[0003] Pending U.S. patent application Ser. No. 09/904,940, filed Jul. 12, 2001 by inventor Yu Qing YANG and entitled “SWITCHED-CAPACITOR SUMMER CIRCUITS AND METHODS AND SYSTEMS USING THE SAME”; and

[0004] Pending U.S. patent application Ser. No. 09/870,900, filed May 30, 2001 by inventor Yu Qing YANG and entitled “SWITCHED-CAPACITOR SUMMER CIRCUITS AND METHODS AND SYSTEMS USING THE SAME”.

[0005] 1. Field of the Invention

[0006] The present invention relates in general to switched-capacitor circuitry and methods and in particular to switched-capacitor circuits and methods with improved settling time and systems using the same.

[0007] 2. Description of the Related Art

[0008] Switched-capacitor circuits are used in a wide range of applications, including analog-to-digital converters (ADCs), digital-to-analog converters (DACs) delta-sigma modulators, filters, power supplies, and voltage regulators. Generally, switched capacitor techniques allow for the construction of accurate, compact, frequency tunable circuits without the use of resistors. This is particularly advantageous in integrated circuit designs, where the construction of precise resistor-based circuits is difficult and unnecessarily consumes chip area.

[0009] In the basic switched-capacitor integrator, an input voltage is sampled onto a sampling capacitor during a first clock phase. During a second (non-overlapping) clock phase, the charge on the sampling capacitor is transferred to the integrator capacitors. The output of the integrator is fedback to the summing node with a integrator capacitor. The impedance of the circuit depends on the size of the sampling and integrator capacitors and the frequency of the clock.

[0010] One of the key performance parameters in switched-capacitor circuits is the settling time. In particular, it is important that the circuit output ramp to its final value quickly and smoothly. This is particularly true in high performance applications, such as audio analog-to-digital and digital-to-analog conversion circuits where discontinuities in the circuit response can affect the ultimate quality of the audio signal presented to the listener.

[0011] According to one such embodiment, a switched capacitor circuit is disclosed which includes an operational amplifier having an input and an output, a sampling capacitor and a set of switches. During a first phase, a set of switches samples an input voltage by charging sampling capacitor. During a first portion of a second phase, the operational amplifier input is electrically coupled to sampling capacitor with a first path having a first time constant. During a second portion of the second phase, the operational amplifier input is electrically coupled with the sampling capacitor through a second path having a second time constant, the second time constant being smaller than the first time constant.

[0012] The present inventive principles substantially reduce the settling time when a switched capacitor techniques are being practiced. Specifically, circuits and methods are disclosed which substantially reduce the negative output spikes in switched capacitor circuits by controlling the rate of voltage change at the switch capacitor circuit summing nodes. In particular, multiple sets of switches are provided for transferring charge to the summing node. One set of switches has a large time constant such that the voltage change is relatively slow and a second set of switches has a smaller time constant such that the change in voltage at the summing node is relatively rapid.

[0013] For a more complete understanding of the present invention, and the advantages thereof, reference is now made to the following descriptions taken in conjunction with the accompanying drawings, in which:

[0014] FIG. 1A is a functional block diagram of a single-chip analog to digital converter suitable for illustrating the present inventive principles; invention

[0015] FIG. 1B is functional block diagram of a 5th Order delta-sigma modulator suitable for use in the analog to digital converter of FIG. 1A:

[0016] FIG. 1C is a functional block diagram of a digital radio employing 1C and analog to digital converter such as shown in FIG. 1A;

[0017] FIG. 2A illustrates an exemplary fully-differential conventional delayed switched-capacitor integrator;

[0018] FIG. 2B illustrates the ideal response;

[0019] FIG. 2C illustrates the embodiment in which this creates a negative spike in the output

[0020] FIG. 3 illustrates a first technique for substantially reducing or eliminating the negative voltage spike shown in FIG. 2 according to the present inventive concepts; and

[0021] FIG. 4 illustrates an alternate switched-capacitor network 400 according to the inventive principles.

[0022] The principles of the present invention and their advantages are best understood by referring to the illustrated embodiment depicted in FIGS. 1-4 of the drawings, in which like numbers designate like parts.

[0023] FIG. 1A is a high level functional block diagram of a single-chip audio analog-to-digital (A/D) 100 suitable for practicing the principles of the present invention. A/D converter 100 is only one of a number of possible applications requiring switched-capacitor integrator stages. Other examples include digital to analog converters (DACs) and Codecs.

[0024] A/D converter 100 includes two conversion paths for converting left and right channel analog audio data respectively received at the left and right analog differential inputs AINL+/− and AINR+/−. The analog inputs are passed through an input gain stage 101 and then to a 5th order delta-sigma modulator.

[0025] Each delta-sigma modulator is represented in FIG. 1 by a summer 102, low-pass filter 104 (quantizer), comparator 105 and DAC 106 in the feedback loop. The outputs from the delta-sigma modulators are passed through a decimation filter 107 and a high pass filter 108.

[0026] The resulting left and right channel digital audio data are output through a single serial port SDOUT, timed with a serial clock SCLK and a left-right clock LRCLK in accordance with the Digital Interface Format (DIF). The SCLK and LRCLK clocks can be generated externally and input to converter 100 or can be generated on-chip, along with the associated data, in response to a received master clock MCLK.

[0027] If 5th order delta-sigma modulators are used in system 100, each low pass filter 104 comprises 5 integrator stages 109 as shown in FIG. 1B. Integrator stages 109 are one of a number of possible applications of switched capacitor integrators 300 and 400 according to the inventive concepts discussed below.

[0028] One possible application of A/D converter is in a digital radio 110, such as that shown in FIG. 1C.

[0029] Digital radio 110 includes an analog section or front-end 111 which receives radio frequency (RF) signals from an associated antenna 112. Analog front-end 111 is preferably a conventional RF down-converter including a low noise amplifier (LNA) 113 for setting the system noise figure, a bandpass filter 114 and mixer 115 driven by an analog local oscillator 116. The mixed-down analog signal is then converted into digital form by analog to digital converter 117.

[0030] The digitized data output from A/D converter 117 is passed to digital processing section 118. A pair of mixers 119a,bgenerate in-phase (I) and quadrature (Q) signals from a corresponding pair of clock phases from crystal oscillator 120. The I and Q signals are next passed through bandpass filters 121aand 121band on to digital baseband processor 122. The processed digital signal is then re-converted to analog (audio) form by D/A converter 123.

[0031] A switched mode (Class D) audio power amplifier (APA) 124 is used to drive an external set of speakers or a headset. Preferably, at least some of the components of digital radio 110 are powered by a switched mode power supply (SMPS) 124.

[0032] An exemplary fully-differential conventional delayed switched-capacitor integrator 200 is shown in FIG. 2A. Integrator 200 generally works as follows. During clock Phase 1 (φ₁) switches 201a,band 204a,bclose and the input voltage V_Inis sampled onto sampling capacitors Cs 203a,b.During clock Phase 2 (φ₂), switches 201a,band 204a,bopen and switches 202a,band 205a,bclose. Consequently, the charge on sampling capacitors Cs to the summing nodes of op amp 206 and onto capacitors 207a,b.

[0033] The size of sampling capacitor Cs, and accordingly the amount of charge transferred, will be very large in high performance A/D and D/A converters, and similar applications. Specifically, the charge being transferred is in accordance with the relation:

[0034] C=Q(V_LEFT−V_RIGHT)  (1)

[0035] where:

[0036] C is the capacitance, Q is the charge stored on the capacitor, V_LEFTis the voltage on the left capacitor plate, and V_RIGHTthat on the right capacitor plate. In the present example, the charge Q and capacitance C are fixed.

[0037] During the initial period of Phase 2, (after switches 202a,band 205a,bhave closed), the voltage on the left plate of sampling capacitors C_Schanges from V_IN(i.e., the final voltage at the end of Phase 1) to V_cm. At the same time, the charge on sampling capacitors C_Sremains the same, just before it begins to be transferred out. According to Equation 1, the right plate voltage on sampling C_S(i.e. the summing nodes) also must change by the same amount as the left plate voltage, which normally is a larger voltage swing. Consequently, op amp 206 must pull the summing nodes back to common mode voltage (V_CM) before it can start to ramp-up the output voltage V_Outto its final value. This creates a negative spike in the output as shown in FIG. 2C (the ideal response is shown in FIG. 2B for reference). Because the length of Phase 2 is fixed, this phenomena seriously affects the settling of the circuit output

[0038] Moreover, in order to improve the settling behavior of the conventional integrator shown in FIG. 2, switches 202a,band/or 205a,bare usually selected to be large so as to decrease the time constant of the switched-capacitor network (i.e. speed up the response). However, as the speed of the network increases, the negative spike shown in FIG. 2C also grows larger.

[0039] FIG. 3 illustrates a first technique for substantially reducing or eliminating the negative voltage spike shown in FIG. 2 according to the present inventive concepts. In this embodiment, the time constant of the switched-capacitor network 300 is varied using sets of parallel switches 301a,band 302a,bin place of switches 202a,bin the embodiment of FIG. 2A.

[0040] In network 300, switches 301a,bare selected to be small such that they have a large resistance while switches 302a,bare selected to be large such that they have a small resistance. Small switches 301a,btherefore increase the time constant and make network 300 slower. On the other hand, large switches 302a,bdecrease the time constant and make the network faster. Network 300 preferably operates as follows.

[0041] During Phase 1, switches 204a,bclose and the input voltage V_Inis sampled onto sampling capacitors C_S. At the same time, switches 205a,bare open such that the summing nodes of op amp 206 are at the voltage on integrator capacitors C_I.

[0042] Small switches 301a,bclose first during Phase 2 in response to the clock signal φ₂and switches 204a,bopen. When only small switches 301a,bare closed, and the network 300 is slower (i.e. the RC time constant is increased) which forces the voltages the summing inputs to integrator 206 to change more slowly. Consequently, the magnitude of “negative” spike at the integrator output becomes much smaller.

[0043] Large switches 302a,bare subsequently closed, after the introduction of a delay, by the clock signal φ_2D. In particular, after op amp 206 has pulled the summing nodes almost back to common mode voltage VCM, large switches 302a,bare closed such that network 300 becomes significantly faster. The electric charge on sampling capacitors C_Sis transferred to the integrator capacitors C_Iat full speed.

[0044] FIG. 4 illustrates an alternate switched-capacitor network 400 according to the inventive principles. In this case, switches 205a,bin the conventional embodiment are replaced by sets of parallel switches 401a,band 402a,b.Small switches 401a,bare closed during the start of Phase 2 in response to clock signal φ₂to slowly bring the voltages at the summing nodes the common node voltage and minimize the negative voltage spike. Large switches 402a,bare subsequently closed after a delay by delayed clock signal φ_2Dto transfer the charge on sampling capacitors C_Son to the integrator capacitors C_Irapidly. In other words, as was done in network 300, a large time constant is used during the initial phase to minimize spikes in the output while a smaller time constant is used to fully charge the integrator capacitors and ramp the output voltage.

[0045] Although the present invention and its advantages have been described in detail, it should be understood that various changes, substitutions and alterations can be made herein without departing from the spirit and scope of the invention as defined by the appended claims.