Power supply regulation can be an important design consideration in modern integrated circuit devices. Designers of such devices often take into account power supply voltage fluctuations, as well as other types of process and component performance variations, in attempting to arrive at a robust system design.
With continued reductions in size of integrated circuit elements, smaller devices and lower power supply voltages are often implemented. Threshold voltage levels, however, do not generally tend to decrease as fast as other main design parameters. For example, some semiconductor memory designs use sense amplifiers to sense relatively small voltage differentials in order to detect data bits stored in memory cells. These voltage differentials can be on the order of about 50-100 millivolts (mv). In these and other applications, power supply voltage fluctuations can adversely affect the ability of such systems to function reliably.
Decoupling capacitance can be added to such designs in an effort to reduce power supply voltage fluctuations. Decoupling capacitance can be implemented by distributing small capacitors throughout the circuit which are permanently installed between various power supply lines and reference lines, such as ground. During operation, each capacitor accumulates charge from the supply line when the voltage overshoots the nominal voltage level, and dumps charge to the supply line when the voltage undershoots the nominal voltage level.
Because the voltage difference between the capacitors and the fluctuating supply voltage can be relatively small, a relatively large number of capacitors may be required to maintain the supply voltage within acceptable voltage fluctuation tolerances. This can increase the cost of the design, in that the capacitors take up overhead space on a semiconductor chip that could be utilized for more valuable functions. Because the capacitors are a permanent feature of the circuitry, the capacitors can also increase the amount of leakage current within the circuit, leading to higher power consumption levels.
Various embodiments of the present invention are generally directed to a method and apparatus for compensating for voltage fluctuations on a voltage supply line in an integrated circuit device.
In accordance with some embodiments, the method generally comprises sensing a voltage on the voltage supply line, and using a switch to actively connect a charge storage device (CSD) to the supply line when the sensed voltage passes outside a predetermined voltage range.
In accordance with other embodiments, the apparatus generally comprises a voltage fluctuation sensor configured to sense a voltage on the voltage supply line, and a compensation circuit comprising a switch and a charge storage device (CSD). The switch actively connects the CSD to the supply line when the voltage sensed by the voltage fluctuation sensor passes outside a predetermined voltage range.
These and other features and advantages which characterize the various embodiments of the present invention can be understood in view of the following detailed discussion and the accompanying drawings.
The device 100 includes a controller 102 which provides top level control of the device. The controller may be a programmable or hardware based processor. Data I/O operations are carried out using an interface (I/F) circuit 104 which communicates with the host device. Data are transferred between the host device and a data storage array 106. The storage array 106 can comprise an array of volatile or non-volatile memory cells.
A power supply 108 supplies electrical power in the form of various supply voltages to the controller 102, I/F circuit 104 and storage array 106 to facilitate operation of these devices. These supply voltages may be at different nominal voltage levels depending on the requirements of the device, and may be on the order of about +3.0V, ±5.0V, +20.0V, etc. The power supply 108 supplies the electrical power from a separate power source (not shown), such as a battery or a power supply cable from the host device.
As explained below, the regulation circuit 110 operates to detect and compensate for fluctuations in the supply voltage on line 114 in order to maintain the voltage on the line substantially equal to a nominal supply voltage level VDD within some predetermined tolerance range (such as VDD=+3.0V±Δ). The regulation circuit 110 includes a voltage fluctuation sensor 116, an overshoot compensation circuit 118, and an undershoot compensation circuit 120.
The fluctuation sensor 116 detects fluctuations in voltage on the line 114 above (overshoot) and below (undershoot) selected threshold levels. An exemplary voltage threshold range may be ±10% of the nominal line voltage level (e.g., 2.7V to 3.3V), or some other value.
When an overshoot condition is detected (e.g., VDETECT>3.3V), an overshoot signal (OVER) is provided via path 122 to the overshoot compensation circuit 118. In response, the circuit 118 operates to lower the voltage level of the line 114 back within the acceptable voltage range via an active connection 124. Similarly, when an undershoot condition is detected (e.g., VDETECT<2.7V), an undershoot signal (UNDER) is provided via path 126 to the undershoot compensation circuit 120, which in turn operates to raise the voltage level of the line 114 to the acceptable voltage range via an active connection 128.
The sensor 116 further includes a comparator stage 142 with respective first and second comparators 144, 146 which operate as 1-bit analog-to-digital converters (ADCs). A first reference voltage is supplied to the first comparator 144 characterized as an overshoot reference input, or OS_REF. A second reference voltage is supplied to the second comparator 146 characterized as an undershoot reference input, or US_REF. The outputs of the respective comparators 144, 146 provide the OVER and UNDER signals on paths 122 and 126 in
The amplifier stage 130 operates to amplify noise variations (AC components) appearing on the VDD and GND terminals 138, 140, and supply an amplified signal to the respective comparators 144, 146. The first and second voltage references OS_REF and US_REF are selected at appropriate levels so that, when the noise exceeds the associated threshold, a 1-bit digital signal will be output. That is, the OVER or UNDER output will transition to a selected logical state, such as high (logical 1). The magnitudes of the voltage references and the gain of the differential amplifier can be empirically determined to provide the requisite threshold range about the nominal VDD voltage.
The switching device 150 respectively connects the capacitor 152 between a ground terminal 154 connected to electrical ground (or other reference utilized at 140 in
When the OVER signal on path 122 transitions high, the inverter 148 inverts this signal to provide an input to the switching device 150, which actively connects the capacitor 152 to the VDD supply line terminal 156. The capacitor 152 will be in an initial non-charged state, and so will begin to accumulate charge from the VDD supply line 114, thereby lowering the voltage of the supply line.
The switching device 150 will continue to actively maintain the capacitor 152 coupled to the supply line until the amplified noise signal from the differential stage 130 (
The undershoot compensation circuit 120 of
When the UNDER signal on path 128 is asserted high, the inverter 162 signals the switching device 160 to actively connect the precharged capacitor 162 to the terminal 166, which is coupled to the voltage supply line 114. The prestored charge on the capacitor 162 from the voltage boosting circuit 168 is transferred to the supply line 114 in relation to the differential voltage between the capacitor and the line. This state will continue until the sensed voltage on the supply line 114 exceeds the US_REF threshold,
In some embodiments, the voltage boosting circuit 168 of
The charge pump circuit 170 is configured as a Dickson charge pump and includes serially connected switching devices 172 respectively connected to charging capacitors 174. The switching devices can be n-channel MOSFETS or can take some other form, such as static charge transfer switches. Time-varying clock inputs Φ and ΦB are supplied on paths 176, 178 and are 180 degrees out of phase. An input voltage VIN, such as VDD, is supplied on line 180 and a voltage VOUT is output on line 182. Generally, VOUT will be greater than VIN (VOUT>VIN), and may be characterized as:
where N is the number of stages (in this case, 4), C is the capacitance of each of the capacitors 174, CS is a measure of stray capacitance associated with the MOSFETs 172, VΦis the voltage magnitude of the input clock signals, IOUT is the output current, fOSC is the frequency of the input clock signals, and VTN is the CMOS threshold voltage of the MOSFETs. Variations to the circuitry in
Decision step 204 queries whether an overshoot condition has been detected; if so, overshoot compensation is applied at step 206 as set forth in
Decision step 210 in
In this way, the active regulation circuit 110 initiates and maintains compensation of both overshoot and undershoot conditions in a closed-loop fashion until the fluctuations in the supply line voltage are returned to the specified threshold range.
A straightforward analysis shows that the amount of charge provided by the active operation of
Further advantages of the various embodiments illustrated herein include reduced areal overhead through the elimination of permanently connected, passive decoupling capacitors as used in the prior art. The active regulation provides reduced variation in supply voltage levels over prior art approaches, leading to improved performance and device reliability. It will be appreciated that the various embodiments discussed herein have numerous potential applications and are not limited to a certain field of electronic media or type of data storage devices.
While both overshoot and undershoot compensation blocks (see 118, 120 in
The use of digital logic such as the DAC and inverter combinations 134/148 and 136/158, provides hysteresis control and noise rejection stability to the control loop. Other control mechanisms can readily be used, however, including other combinations of logical gates. Various other types of charge storage devices (CSDs) can be used to respectively accumulate and dump charge to the supply lines apart from the discrete semiconductor transistors disclosed herein, including capacitive charge planes, inductors, etc.
It will be further appreciated that the regulation circuit 110 operates in accordance with the various embodiments to augment the voltage being generated by the voltage source 112, and not to serve as a substitute therefor such as in the case of a power shutdown operation. However, in further embodiments it is contemplated that circuitry as embodied herein could be used to detect larger changes in voltage, such as −30% of VDD (0.7VDD), and to switch in the addition of charge to signal and initiate a short term recovery operation, such as the transfer of data from a volatile memory location to a non-volatile memory location before the supply voltage reaches a level (such as zero volts) where further device operation is inhibited.
It is to be understood that even though numerous characteristics and advantages of various embodiments of the present invention have been set forth in the foregoing description, together with details of the structure and function of various embodiments of the invention, this detailed description is illustrative only, and changes may be made in detail, especially in matters of structure and arrangements of parts within the principles of the present invention to the full extent indicated by the broad general meaning of the terms in which the appended claims are expressed.
This application makes a claim of domestic priority under 35 U.S.C. §119(e) to U.S. Provisional Patent Application No. 61/103,738 filed Oct. 8, 2008.
Number | Date | Country | |
---|---|---|---|
61103738 | Oct 2008 | US |