Abstract: A power controller includes a plurality of power supply output circuits and a multi-level converter connected to the plurality of power supply output circuits to receive a plurality of sets of voltages. The multi-level converter is configured to supply power to a load from a plurality of output terminals. An observer module is operably coupled to the multi-level converter to selectively control multi-level converter common-mode voltage with respect to the plurality of power supply output terminals.

Abstract: A control system having an integrator component that provides an integrator output signal used to control a load, and a method controlling the same includes conditioning the integrator component during a fraction of the period in which it would tend to otherwise wind-up.

Abstract: A method and apparatus for providing electrical power through two terminals to a load wherein a portion of a LC filter circuit is connectable to each of the terminals. A feedback circuit having a sense capacitor is operably couplable to both portions of the LC output filter. The feedback circuit provides a feedback signal when a frequency of AC voltage across the sense capacitor substantially reaches the resonant frequency of the LC output filter so as to actively damp the voltage across the load.

Abstract: A high-speed constant-horsepower motor includes a rotor and a stator having a plurality of groups of windings which are separately driven. In one mode of operation, the polarity of the voltage applied to each winding is such that the windings behave as distinct electromagnets, each defining a separate magnetic pole. In another mode of operation, the polarity of voltage applied to some of the windings is reversed, such that pairs of adjacent windings behave as single electromagnets. In the latter mode, the effective number of magnetic poles is reduced by a factor of two. Thus, the effective number of magnetic poles of the motor can be varied electronically, even while the motor is operating. The invention makes it possible, in one example, to operate the motor as an eight-pole motor at low speeds, and as a four-pole motor at high speeds. The effective motor constants are changed appropriately from eight-pole to four-pole modes.

Abstract: A high-voltage switching circuit is formed of a series combination of switches, each of which is rated for use at a voltage lower than that of the power supply. The switches are connected in series with inductors, the inductors being magnetically coupled, in an anti-phase manner. A pair of capacitors, having substantially equal values of capacitance, are connected, respectively, in parallel with each series combination of a switch and inductor. The inductors and capacitors together insure that no more than one-half the total voltage of the power supply will ever appear across either switch. The series combination of switches can be used in place of a single switch, and permits the use of switches rated at low voltage to be combined to form an overall switch which works reliably at a higher voltage.

Abstract: A pulse-width modulated (PWM) circuit applies voltage to a load with high fidelity. The circuit includes switches connected in series with inductors, the inductors being magnetically coupled to other inductors connected to the gates of the switches. The arrangement of inductors and the magnetic coupling causes the turn-on of one switch to assist in the turn-off of the other switch. A pair of diodes connected across some of the inductors assures that the magnetically coupled inductors do not assist the switches to turn on. The result is a circuit which effectively eliminates the problem of "shoot through" current, without providing a "dead time" in the switching scheme, and thus without substantial power losses. The circuit can therefore be used in audio amplification, and in other applications requiring high fidelity.

Abstract: Windings for an electric motor are formed by depositing successive layers of conductive material and etching into the conductive material with a precisely-controlled high-energy beam. A computer stores information on the exact structure of the monolith of windings to be built, in the form of a series of successive planar cross-sections of the monolith. In making the monolith, one deposits a layer of conductive material onto a workpiece, bonds the conductive material to the workpiece, and cuts a pattern into the layer, using a high-energy beam in the presence of an oxidizing atmosphere. The beam causes the formation of electrically-insulating oxides which separate the layer into distinct conductive regions, according to a pattern defined by one of the stored cross-sections. In an alternative, one can also inject additional insulating material into the cuts formed by the beam. The monolith is thus built up, layer by layer, in a series of iterations.

Abstract: A pulse-width modulated (PWM) circuit alternately applies voltage from a power supply across a load, and alternately establishes one or more non-powered, closed paths for freewheeling current. The circuit includes pairs of switches connected to the load and also connected to the power supply. PWM signals derived from a command signal drive the switches in such a manner as to create the freewheeling current paths in a desired pattern and sequence. The circuit effectively reduces ripple current in the load while also substantially reducing the amount of current flowing through the power supply filter capacitor. Thus, the size and rating of the power supply can be reduced. One can use the invention to drive a load in a multiple-phase environment, and also in the context of a three-phase trapezoidal drive.

Abstract: A digital process generates a pulse-width modulated (PWM) signal for providing current to a load. In one embodiment, an input signal is repeatedly sampled at times corresponding to edges of a reference square wave. The sampled values are used to produce a PWM signal, each pulse of the PWM signal having a width corresponding to one of the sampled values. In another embodiment, the sampled values are used to produce a series of intermediate pulses, the intermediate pulses being used to construct an output PWM signal. The pulses of the output PWM signal generally have leading and trailing edges which do not coincide with the times at which sampling occurs. Therefore, in the second embodiment, sampling does not occur when the PWM signal is opening or closing a switch, and the sampling tends to occur near the average value of the input signal.

Abstract: A pulse-width modulated (PWM) circuit provides current through a load by opening and closing main electronic switches which connect the load to a power supply. Each main electronic switch connects with an auxiliary switch, connected in series with a diode, the auxiliary switch and diode being called "recovery paths". The recovery paths provide a path between the load and the associated main switch, on one side, and the power supply on the other. A first train of pulses, derived from a command signal, drives the main switches, and a second train of pulses, derived from the first pulse train, controls the switches in the recovery paths. One arranges the second train such that the recovery paths carry current only while the main switches change from on to off, or from off to on. In this way, the recovery paths carry part of the current that would otherwise flow transiently through the main switches, due to the inherent inefficiencies in semiconductor switching.

Abstract: A pulse-width modulated (PWM) circuit provides current through a load by opening and closing main electronic switches which connect the load to a power supply. Each main electronic switch connects with an auxiliary switch, connected in series with a diode, the auxiliary switch and diode being called "recovery paths". The recovery paths provide a path between the load and the associated main switch, on one side, and the power supply on the other. A first train of pulses, derived from a command signal, drives the main switches, and a second train of pulses, derived from the first pulse train, controls the switches in the recovery paths. One arranges the second train such that the recovery paths carry current only while the main switches change from on to off, or from off to on. In this way, the recovery paths carry part of the current that would otherwise flow transiently through the main switches, due to the inherent inefficiencies in semiconductor switching.

Abstract: A pulse-width modulated (PWM) circuit alternately applies voltage from a power supply across a load, and alternately establishes one or more non-powered, closed paths for freewheeling current. The circuit includes pairs of switches connected to the load and also connected to the power supply. PWM signals derived from a command signal drive the switches in such a manner as to create the freewheeling current paths in a desired pattern and sequence. The circuit effectively reduces ripple current in the load while also substantially reducing the amount of current flowing through the power supply filter capacitor. Thus, the size and rating of the power supply can be reduced. One can use the invention to drive a load in a multiple-phase environment, and also in the context of a three-phase trapezoidal drive.

Abstract: The invention icludes a device for protecting electronic circuitry from damage, especially damage resulting from high neutron flux density produced by a nuclear explosion. The invention protects the circuitry by disabling its power supply system upon detection of the nuclear event. The power supply system is disabled both by short-circuiting the main power supply, which provides control voltages to various subsidiary powert supplies, and by short-circuiting all the output filter capacitors in the system, which may have a substantial residual charge. The invention periodically and automatically attempts to re-start the power supply system, but the system can be re-started only if the nuclear event detector is not actuated. The circuit made according to the invention is entirely analog, and therefore does not depend, for its operation, on the continued functioning of the circuitry to be protected.

Abstract: A pulse-width modulated (PWM) circuit applies a voltage across a load in accordance with a PWM command signal. The command signal is converted into first and second trains of pulses, by electronically comparing the command signal with a triangular dither signal and also with the dither signal whcih has been shifted by half its period. These two pulse trains, and two more pulse trains obtained by taking the complements of the original pulse trains, are applied to a pair of bridge circuits. The load is connected to both bridge circuits. Both bridge circuits include electronic switching devices actuated by the pulses. The bridge circuits are configured such that when a pulse is present, from either the first or the second pulse train, no net voltage appears across the load. At all other times, substantially the entire power supply voltage is applied across the load. In another embodiment, one or more additional pairs of bridge circuits are connected across the load.

Abstract: A pulse-width modulated circuit applies a voltage across a load in accordance with a command signal. The command signal is converted into first and second trains of pulses, by electronically comparing the command signal, and its inverse, with a triangular dither signal. These two pulse trains, and two pulse trains obtained by taking the complements of the original pulse trains, are applied to a pair of bridge circuits. The load is connected to both bridge circuits. Both bridge circuits include electronic switching devices actuated by the pulses. The bridge circuits are configured such that when a pulse is present, from either the first or the second pulse train, no net voltage appears across the load. At all other times, substantially the entire power supply voltage is applied across the load.

Abstract: This invention is a method and apparatus for automatically adjusting the command signal of a servomechanism, to improve the accuracy of control over the motion of an object. In one embodiment, the command signal is advanced in time, to compensate for the delayed response of the servo. The time advance is repeatedly and automatically adjusted so as to minimize the magnitude of the tracking error. Each adjustment of the time advance is based on the value of a sampled error signal, taken at selected points in time. Other parameters of the command signal, such as the amplitude and average level, can also be adjusted, and the adjustments for all parameters can be done simultaneously. For a given parameter, the error signal is sampled at those points at which misadjustment of that parameter is most likely to contribute to the error of the system. Thus, in general, different parameters of the command signal are adjusted based on different sets of sampled points of the signal.

Abstract: A circuit is disclosed for controlling the position or velocity of a moving object. The motion of the object may be linear or rotational. The object is moved by an electric motor, or linear actuator, which is driven by a power amplifier. The current output of the amplifier, which drives the motor or linear actuator, is controlled by an input signal based at least in part on the estimated velocity of the object. The control circuit includes a pair of integrators connected in series, the inputs to the integrators being dependent both on the current output from the amplifier, and on the difference between the output of the second integrator and the actual sensed position of the object. The velocity estimation circuit is thus based on a form of Kalman filter. The velocity estimator uses a model of the motor which is extremely simple, and which contains only constants. The velocity estimator can control the motor very accurately without detailed knowledge of the model of the motor.

Abstract: The invention provides a method and apparatus for determining the angular displacement of a rotating shaft, and for commutation of a brushless motor. At least two analog sensors, mounted near the shaft, generate phase-displaced analog signals responsive to the angular position of the shaft. A logic circuit produces a digital code indicating the region of the cycle in which the shaft is instantaneously located. An analog multiplexor selects a unique pair of analog signals from the group of the original analog signals and their analog complements, according to the value of the digital code. An analog to digital converter then generates a digital signal proportional to the quotient of the analog signals produced by the multiplexor, and this digital signal, together with the digital code, is used to activate an address in a memory device. The memory device, such as an EPROM, is programmed with a large number of binary digits, which define a train of pulses.