Electric power conversion systems – Current conversion – With voltage multiplication means
Reexamination Certificate
2000-01-28
2001-08-07
Patel, Rajnikant B. (Department: 2838)
Electric power conversion systems
Current conversion
With voltage multiplication means
C323S284000, C307S110000
Reexamination Certificate
active
06272029
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates, in general, to integrated circuits and, more particularly, to integrated circuits having charge pump circuits generating a power supply voltage from an external power supply voltage.
2. Relevant Background
Electronic systems usually comprise ICs manufactured with a variety of technologies. This has created a need for multiple power supply voltages to be supplied on a single printed circuit board to support the various types of devices on that board. Standard IC voltages required by typical devices range from 5.0 volts to 3.3 volts, or lower voltage. However, there are a number of devices that require power at voltages in addition to the standard available voltages. These include data communications circuits that often require negative voltages, and interface circuitry such an the RS232 interface that specifies voltages ranging from +/−25V. Moreover, some ICs have different voltage requirements internally although they receive power from an industry standard power supply level. Being able to generate a range of voltage levels, including negative voltages and voltages larger in magnitude that the supplied voltage provides a great deal of flexibility to the circuit designer. Also, higher voltage levels often enable faster switching for better performance.
A practical solution to this disparity is to provide DC/DC converter circuitry that changes an input DC voltage into a higher or lower DC voltage required by another device. A negative charge pump operates to generate a negative voltage by charging a pump capacitor during a first half-cycle of a clock to the level of a source voltage. During a second half-cycle the pump capacitor is disconnected from the source and coupled, with its polarity switched, to a reservoir capacitor, thereby pumping charge to the reservoir capacitor and providing an output that is approximately the negative of the input voltage.
A positive charge pump may also operate to generate a higher voltage than the supply voltage (i.e., a “step-up” converter) by coupling the pump capacitor to the source voltage during the first half-cycle. During the second half-cycle, the pump capacitor's positive terminal is disconnected from the source voltage and the capacitors negative terminal is coupled to the source voltage in its place. The pump capacitor's positive terminal is then coupled to the reservoir capacitor to charge it to approximately twice the source voltage.
Larger high output charge pumps usually run at lower frequencies and therefore are not optimized for size. The size of a large, low frequency charge pump may be a limiting factor in obtaining the smallest IC chips as possible. It is desirable to make on-chip charge pumps as small as possible especially when the charge pumps occupy a significant area of the chip. For a target output current, the smaller the size of the charge pump, the higher proportionately the operating frequency must be. Typically, for high-current-output (e.g., greater than 5-10 milliamp) charge pumps, the operating frequency of the pump is dictated by the peak operating current and the rate of change in operating current (di/dt), as well as the size of driver and support circuitry.
One problem with a higher frequency charge pump is that regulation of the output voltage level becomes harder since it might take multiple cycles to turn on and off the charge pump which would result in an unwanted hysteresis. A charge pump whose output capacity per pump cycle is large relative to the load it is driving could change the voltage on the load an appreciable amount. In this case, waiting multiple pump cycles after a regulation point is reached to turn on or off the charge pump is not acceptable.
To remedy this problem, high speed and high power regulation methods utilizing direct current (d.c.) differential amplifiers are used. In this solution, a small portion of time at the beginning of a pump cycle is used to sense whether the voltage on the load is at or below a reference level. If the voltage is below, then a pump is initiated. If the voltage is above, then no pump occurs. With the operating frequency approaching 30 Mhz (33ns period), less than about 20% of each clock cycle (i.e., 7 ns) could be devoted to regulation. Although a fast regulation scheme can be accomplished, a large portion of the total allotted charge pump current is used for regulation in this circumstance. When using a high speed d.c. differential regulation scheme, power consumption is a problem.
If a high frequency charge pump is implemented in an IC that uses an active and standby mode, operation can become more complex. Even if power consumption during the active mode can be tolerated, power consumption by the charge pump might still be a problem during standby mode. Typically the standby mode requires much lower power consumption yet the charge pump must be operational at least some of the time. To conserve power, which is the purpose of the stand-by mode, it is desirable to turn off high powered regulation circuitry when not in use. While still in standby mode, the high powered regulation circuitry must be turned on and stabilized before entering a pump cycle requiring more complicated control and timing circuitry. This stabilization time consumes power and quite possibly increase the current consumption specification during standby.
Another method to decrease the current consumption during standby, is to totally shut-off the high powered regulator and instead use a very low power regulator that is always on. This method requires critical circuitry to ensure that multiple or partial pumps do not occur since the low power regulator takes some time for decisions to be made. In another variant, a low power, low output current pump is operational during standby mode so that a slow turn on or off would only produce a small hysteresis on the output voltage. These prior solutions all require more circuitry and complicated control logic.
SUMMARY OF THE INVENTION
The present invention involves a dynamic regulation system that is both low power and high speed. The regulator in accordance with the present invention compares a reference voltage against an input-signal and has small devices that are clocked such that internal regulator quickly nodes transition to the power rail voltages. Once the internal nodes are at the power rail voltages, little power is consumed. In accordance with the present invention, loading from subsequent circuitry is kept to a minimum so that small devices can be used to implement the internal regulator circuitry. Smaller devices enable the regulator in accordance with the present invention to be faster and consume less power. Thought the present embodiment relates to positive charge pumps, all techniques discussed can be applied to negative charge pumps.
Briefly stated the present invention involves a charge pump including an input node coupled to receive an input voltage from a power voltage source and an oscillator unit generates a periodic enable regulator signal and a periodic reset signal. A regulator clock unit is coupled to the oscillator unit generating a precharge (PC) signal and a reset regulator signal in response to the enable regulator signal. A pump clock unit receives a master clock signal and generating a plurality of pump clock signals. A charge pump unit is coupled to the input node and is operatively controlled by the plurality of pump clock signals, and coupled to an output terminal coupled to produce an output signal (V
PUMP
). A regulator unit is coupled to receive the V
PUMP
signal, the PC signal, a reference signal and the enable regulator signal, where the regulator unit is responsive to the enable regulator signal to operate in either a precharge mode or a regulation mode.
In another aspect, the present invention involves a method of charge pump regulation in which a dynamic regulator is disabled in a precharged or “ready to fire” state. In this standby state, the dynamic regulator is shut down consuming essentially
Hogan & Hartson LLP
Kubida William J.
Langley Stuart T.
Patel Rajnikant B.
United Microelectronics Corporation
LandOfFree
Dynamic regulation scheme for high speed charge pumps does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Dynamic regulation scheme for high speed charge pumps, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Dynamic regulation scheme for high speed charge pumps will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2473413