Lag compensating controller having an improved transient response

[0002] This is a continuation of Ser. No. 09/753,120 filed Dec. 29, 2000.

[0003] The present invention relates to control systems in general, and to voltage regulators in particular.

[0004] Voltage regulators, such as DC-to-DC converters, are used to provide stable voltage sources for electronic systems. Efficient DC-to-DC converters are particularly needed for battery management in low power devices, such as laptop computers and mobile phones. Switching voltage regulators (or simply “switching regulators”) are known to be an efficient type of DC-to-DC converter. A switching regulator generates an output voltage by converting an input DC voltage into a high frequency voltage, and filtering the high frequency voltage to generate the output DC voltage.

[0005] Conventional switching regulators include two switches. One switch is used to alternately couple and decouple an unregulated input DC voltage source, such as a battery, to a load, such as an integrated circuit. The other switch is used to alternately couple and decouple the load to ground. An output filter, typically including an inductor and an output capacitor, is coupled between the switches and the load to filter the output of the switches and produce the output DC voltage.

[0006] The switches within the switching regulator are opened and closed according to commands from a closed-loop control system Control systems within DC-to-DC converters, just like control systems generally within any electronic system, need to be stabilized. Care in the design of the control system in a DC-to-DC converter must account for variations of parameters such as the it voltage, filter inductor and capacitor values, switch resistances, printed circuit board parasitics, etc. Sometimes a simple scheme such as voltage feedback alone will stably control a power supply. In other situations, extra margin of stability and higher bandwidth are gained by using current mode control techniques. Still other schemes use hystereis bands to decide how to control the switches.

[0007] In some cases, it is desired to add compensation to improve phase margin of a DC-to-DC regulator. Often phase margin can be moved by using a lag compensator, which lowers the overall bandwidth to boost phase at the crossover frequency. Unfortunately, with lowered bandwidth, DC-to-DC regulators take longer to respond to load current transients, resulting in larger output voltage deviations. As a result, many such sows use extra capacitance in the converter's output filter to improve transient response. However, using larger capacitors increases the cost of the regulator substantially.

[0008] Commercially-available hysteretic controllers trigger certain responses when the output voltage deviates too high, or too low. However, these controllers do not have a beneficial effect on nominal, steady-state performance while the voltage is within the hysteresis bands, and may add design difficulty due to their non-linear behavior.

[0009] In one aspect, the invention is directed to a method and computer program product for use in a control system controller having a control loop that includes a storage element the control loop receiving an output error signal describing an error in the output of the control system. It includes modifying the contents of the storage element according to a predetermined adjustment value when a minimum predetermined excursion occurs in the output error signal; providing a first predetermined gain to the output error signal; providing a second predetermined gain to the output of the control loop; and combining the outputs of the first and second gain elements to produce an output control signal.

[0010] Advantages that can be seen in implementations of the invention include one or more of the following. The compensator can add phase margin without requiring extra capacitance, can provide enhance stability during steady-state conditions, and does not degrade transient response.

[0011] The details of one or more of the invention are set forth in the accompanying drawings and the description below. Other features and advantages of the invention will become apparent from the description, the drawings, and the claims.

[0012] FIG. 1 shows a switching regulator according to an embodiment of the present invention.

[0013] FIG. 2 shows a controller for a switching regulator according to one embodiment of the present invention.

[0014] FIG. 3 shows a controller for a switching regulator according to another embodiment of the preset invention.

[0015] FIG. 4 shows a detector for a switching regulator controller according to an embodiment of the present invention.

[0016] FIG. 5 shows several contemporaneous waveforms that result from transients in the load of the switching regulator of FIG. 1.

[0017] Like reference numbers and designations in the various drawings indicate like elements.

[0018] Referring to FIG. 1, a switching regulator 102 is coupled to an unregulated DC voltage source 104, such as a battery, by an input terminal 106. The switching regulator 102 is also coupled to a load 108, such as an integrated circuit, by an output terminal 110. The switching regulator 102 serves as a DC-to-DC converter between the input terminal 106 and the output terminal 110. The switching regulator 102 includes a switching circuit 112 which serves as a power switch for alternately coupling and de-coupling the input terminal 106 to an intermediate terminal 114. In some applications, such as a buck converter topology, the switching circuit 112 couples the intermediate terminal 114 to ground when the intermediate terminal 114 is not coupled to the input terminal 106.

[0019] The switching regulator also includes a controller 116 for controlling the operation of the switching circuit 112. The controller 116 causes the switching circuit 112 to convert the substantially DC input voltage V_INat the input terminal 106 into an intermediate voltage having a rectangular waveform at the intermediate terminal 114.

[0020] The intermediate terminal 114 is coupled to the output terminal 110 by an output filter 118. The output filter 118 converts the rectangular waveform at the intermediate terminal 114 to a substantially DC output voltage V_OUTat the output terminal 110. The switching circuit 112 and the output filter 118 may have a buck converter topology, or another topology, such as a boost converter or buck-booster converter topology.

[0021] The output voltage is regulated, or maintained at a substantially constant level, by controller 116. Controller 116 measures electrical properties of the output, such as output voltage and/or output current, and compares these properties to a control electrical property, such as voltage V_REFat terminal 120. Based on this comparison, controller 116 provides a current command I_COMMANDto the switching circuit 112 at terminal 122.

[0022] Switching circuit 112 operates its switches according to the current command I_COMMAND. Switching circuit 112 can control its switches based not only on the current command, but also on the output current delivered by switching circuit 112 to output filter 118. Other embodiment; employ direct feedback without the use of current commands.

[0023] Controller 116 includes a control loop including a storage element that stores a nominal value under nominal conditions. However, under certain predetermined transient conditions, the cones of the storage element are modified as described in detail below.

[0024] In one embodiment, the storage element is loaded with a predetermined adjustment value under predetermined transient conditions. Referring to FIG. 2, controller 116 includes a combiner 202 that receives reference voltage V_REFat terminal 120 and output voltage V_OUTat terminal 110, and produces an error voltage V_EER=V_REF−V_OUTat terminal 222. A gain element 204 applies a gain Gp to V_EERto produce a current I_PROPat terminal 230 that is proportional to V_EER.

[0025] Controller 116 also includes a lag compensator that includes a control loop and a gain element 216 that applies a gain Gi to the output of the control loop to produce a current I_INTat terminal 228. Combiner 218 adds currents I_PROPand I_INTto produce current command I_COMMANDat terminal 122.

[0026] The control loop includes gain elements 206 and 208, combiner 210, delay element 212, and storage element 214. Gain element 206 applies a gain 1-Ki to V_ERR, where Ki is the discrete time pole, in the Z-domain unit circle, of the lag compensator. Selection of an appropriate value for Ki will be apparent to one skilled in the relevant art.

[0027] Combiner 210 combines the output of gain elements 206 and 208. Storage element 214 loads the output VNOM of combiner 210 at terminal 224 during nominal operation (that is, when excursions of V_EERdo not leave a predefined envelope).

[0028] However, when a minimum predetermined excursion occurs in ouput voltage V_OUT, error voltage V_ERRleaves the predefined envelope. This event is detected by detector 220, which asserts a LOAD signal at terminal 234 and a predetermined adjustment value V_ADJat terminal 232. The LOAD signal causes storage element 214 to load predetermined adjustment value V_ADJat terminal 232, causing the predetermined adjustment value to appear at terminal 226 as the output Acc of storage element 214.

[0029] In digital implementations, storage element 214 can be implemented as an accumulator. In analog implementations, storage element 214 can be implemented as an integrating capacitor.

[0030] In digital implementations, detector 220 can be implemented as an A/D converter to determine V_OUT. The A/D converter is centered at analog reference voltage V_REF, and outputs a monotonically increasing four bit rear versus V_OUTwithin the predetermined voltage envelope for V_REF. Below or above that range, the A/D converter clips, or saturates. When the A/D converter saturates, it causes the storage element 214 (here, an accumulator) to preload the predetermined adjustment value. In other embodiments, A/D converters of widths other than four bits are used.

[0031] Delay element 212 applies a predetermined delay to the output of storage element 214. Gain element 208 applies gain Ki to the output of delay element 212.

[0032] In another embodiment, the contents of the storage element are incremented by a predetermined adjustment value under predetermined transient conditions. Referring to FIG. 3, the adjustment value V_ADJis combined with Acc by combiner 302. When detector 220 asserts the LOAD signal, the output of combiner 302 is loaded into storage element 214, thereby incrementing the contents of storage element 214 by the predetermined adjustment value.

[0033] FIG. 4 is a functional block diagram of a detector 220 for a switching regulator controller according to an embodiment of the present invention. The detector includes comparison elements 406A and 406B associated with predetermined preload values 402A and 402B, respectively, and switches 404A and 404B, respectively. Each comparison element compares V_ERRto a predetermined voltage range. When V_ERRfills within a comparison element's range, the comparison element triggers a switch, thereby supplying a predetermined preload value as the adjustment voltage V_ADJ.

[0034] In the embodiment of FIG. 4, detector 220 implements a single envelope bounded by thresholds t1 and t2. When V_ERRfalls below threshold t1, comparison element 406A triggers switch 404A, thereby supplying v1 as adjustment value V_ADJat terminal 224.

[0035] When V_ERRexceeds threshold t2, comparison element 4061B triggers switch 404B, thereby supplying voltage v2 as adjustment value V_ADJat nominal 224. When either V_ERRfalls below threshold t1, or when V_EERexceeds threshold t2, OR gate 408 asserts LOAD signal at terminal 224, thereby causing storage element 214 to load

[0036] Under nominal operations V_ERRfalls between thresholds t1 and t2. Therefore no LOAD signal is generated. Consequently storage element simply loads V_NOMunder nominal operations.

[0037] In one embodiment, detector 220 implements more than one predetermined envelope. The error voltage V_ERRis compared to a plurality of ranges, each associated with an envelope. Each range is associated with a predetermined preload value. When V_ERRfalls within a particular range, the predetermined preload value associated with that range is supplied to storage element 114 as the predetermined adjustment value.

[0038] In general, the magnitude of the adjustment corresponds to the magnitude of the envelope. For example, when a small excursion in V_ERRoccurs, a small adjustment value is supplied to storage element 214. When a large excursion in V_ERRoccurs, a large adjustment value is supplied to storage element 214.

[0039] Selection of an appropriate adjustment values and thresholds will be apparent to one skilled in the relevant art. In general, the adjustment values should be chosen to quickly reduce the output error signal V_ERRto a desirable value. The threshold values should be chosen such that nominal operation of the control system is not unnecessarily disturbed.

[0040] The behavior of the lag compensator can be described in the time domain. The lag compensation appears as a change in the equation for the current command I_COMMAND. Without lag compensation,

[0041] I_COMMAND=GpV_ERR

[0042] where Gp is the proportional gain and V_ERR=V_REF−V_OUTis the error term from the outer voltage loop. For the discrete-time lag compensation technique discussed above, there is an additional term GiAcc so that

[0043] I_COMMAND=GpV_ERRGiAcc

[0044] and

[0045] Acc[n]=Ki*Acc[n−1]+(1−Ki)V_ERR

[0046] where Acc is the output of the accumulator, which acts as storage element 214 in a discrete time implementation of the system.

[0047] The lag compensation pole is defined by Ki. The zero falls out from the combination of these equations in the increased order system and will always be a higher frequency than the pole for non-zero Gi. The ouput Acc of the accumulator will reach in steady-state the value V_EER. Therefore of stead-state,

[0048] I_COMMAND=GpV_EER+GiV_EER

[0049] Thus the DC gain is now Gp+Gi

[0050] FIG. 5 shows several contemporaneous waveforms that result from transients in the load of the switching voltage regulator described above with reference to FIG. 1. An envelope is defined to limit excursions of V_ERRto a high of V_HIGHand a low of V_Low. From time t₁to time t₄, nominal operation is depicted (that is, V_ERRdoes not reach either limit V_HIGHor V_LOWof its envelope).

[0051] At time t₅, a transient from zero load to full load occurs. In response, V_OUTdecreases, and so V_ERRincreases. The proportional part of the current command, I_PROP, changes with V_ERR. There is no delay between a V_EERchange and an I_PROPchange. I_INT, on the other hand, changes slowly due to the Ki pole.

[0052] At time t₆, V_ERRreaches limit V_HIGHof its envelope. Before V_ERRreaches V_HIGH, I_INTchanges slowly with its Ki pole. However, once V_ERRreaches an envelope limit, controller 116 determines that a severe load transient has occurred. The controller 116 then step changes I_INTto a predetermined final value by preloading storage element 214. The preloading adjusts the total current I_COMMANDto the value it would have reached given much more time.

[0053] Two hypothetical waveforms are shown for comparison with the I_COMMANDwaveform. I_COMMANDis shown as a solid line. Waveform 504 depicts how I_COMMANDwould behave without preloading. Waveform 502 depicts the ideal I_COMMAND.

[0054] At time t₇, a transient from full load to zero load occurs. In response, V_OUTincreases, and so V_ERRdecreases. The controller behaves in a manner similar to that described above for the zero load to fill load case.

[0055] The invention has been described in terms of particular embodiments. Other embodiments are within the scope of the following claims. For example, the steps of the invention can be performed in a different order and still achieve desirable results. In addition, embodiments of the controller of the present invention can be used in control systems other than DC-to-DC converters.

[0056] The invention can be implemented in digital electronic circuitry, or in computer hardware, firmware, software, or in combinations of them. Apparatus of the invention can be implemented in a computer program product tangibly embodied in a machine-readable storage device for execution by a programmable processor; and method steps of the invention can be performed by a programmable processor executing a program of instructions to perform functions of the invention by operating on input data and generating output. The invention can be implemented advantageously in one or more computer programs that are executable on a programmable system including at least one programmable processor coupled to receive data and instructions from, and to transmit data and instructions to, a data storage system, at least one input device, and at least one ouput device. Each computer program can be implemented in a high-level procedural or object-oriented programming language or in assembly or machine language if desired; and in any case, the language can be a compiled or interpreted language. Suitable processors include, by way of example, both general and special purpose microprocessors. Generally, a processor will receive instructions and data from a read-only memory and/or a random access memory. Generally, a computer will include one or more mass storage devices for storing data files; such devices include magnetic disks, such as internal hard disks and removable disks; magneto-optical disks; and optical disks. Storage devices suitable for tangibly embodying computer program instructions and data include all forms of non-volatile memo, including by way of example semiconductor memory devices, such as EPROM, EEPROM, and flash memory devices; magnetic disks such as internal hard disks and removable disks magneto-optical disks; and CD-ROM disks. Any of the foregoing can be supplemented by, or incorporated in, ASICs (application-specific integrated circuits).

[0057] Embodiments of the controller of the present invention are not limited to lag compensators, but can also be practiced within other types of compensators, such as lead-lag compensators. Further, although the switching regulator is discussed in the context of a buck converter topology, embodiments of the invention are also applicable to other switching regulator topologies, such as a boost converter topology or a buck-boost converter topology.