Abstract
A method of detecting a fault in a power distribution system includes placing a signal on the system at a frequency F.sub.1 and then detecting a change in the signal due to a change in the impedence of the system as a result of a fault wherein the change is one of a change in phase, a change in signal tone, or a change in voltage level at the load. In one embodiment, band reject filters can be used to diminish the signal at the load or source. In another embodiment, the power source can be a periodic pulsed power source and the signal can be placed on the system during an idle phase of the periodic pulsed power.
Claims
1. A method of detecting a fault in a power distribution system, the method comprising: providing a power source; providing a load to be powered by the power source; providing conductors connecting the power source to the load; placing a signal at a frequency F.sub.1 on the conductors connecting the power source to the load; and detecting a change in the signal indicative of a change of an impedance in the system due to a fault wherein the signal change is one of a change in phase, a change in signal tone, or a change in voltage level at the load.
2. The method of claim 1 wherein the the signal change is a change in phase.
3. The method of claim 1 wherein the signal change is a change in tone.
4. The method of claim 1 wherein the signal change is a change in voltage level at the load.
5. The method of claim 1, further comprising: providing a band reject filter at the power source and a band reject filter at the load wherein both band reject filters are centered at frequency F.sub.1 to reduce the signal at the power source and at the load.
6. The method of claim 1, wherein the power distribution system provides periodic pulses of power and the signal is sent during an idle cycle of the periodic pulsed power.
Description
BRIEF DESCRIPTION OF THE DRAWINGS
[0011] FIG. 1A is a circuit diagram of an exemplary pulsed power distribution system.
[0012] FIG. 1B is a time domain view graph representing a magnitude (in voltage) over a time duration of the pulsed power that is provided by the pulsed power distribution system.
[0013] FIG. 10 is a time domain view graph representing an equivalent DC power delivered to a load by the pulsed power distribution system.
[0014] FIG. 2 is a time domain view graph representing a maximum exposure time of a power fault to a human body.
[0015] FIG. 3 is a circuit diagram of an exemplary pulsed power distribution system that uses fault detection.
[0016] FIG. 4A is a time domain view of the pulsed power distribution technique providing different functions that take place with respect to time.
[0017] FIG. 4B is a time domain view of the pulsed power distribution technique providing different functions that take place with respect to time.
[0018] FIG. 5A is a representative circuit diagram schematic of the load in a power distribution system without a fault.
[0019] FIG. 5B is a representative circuit diagram schematic of the load in a power distribution system with a fault.
[0020] FIG. 5C is a time domain view mapping a phase as a function of frequency for a power distribution system.
[0021] FIG. 6 is a block diagram of a power distribution system with fault detection and bidirectional communication capability.
[0022] FIG. 7A is a magnitude and phase frequency response centered about the frequency F.sub.1, with a fault present.
[0023] FIG. 7B is a magnitude and phase frequency response centered about the frequency F.sub.1, without a fault present
[0024] FIG. 8 is a block diagram of another type of a pulsed power distribution system with fault detection utilizing current measurement at the source and at the load.
[0025] FIG. 9 is a block diagram of another type of a pulsed power distribution system with fault detection utilizing impedance measurement at the source.
[0026] FIG. 10 is a block diagram of another type of a pulsed power distribution system with fault detection utilizing amplitude attenuation detection method.
DETAILED DESCRIPTION OF THE INVENTION
[0027] The present invention involves several methods of detecting a change in a test signal due to a change in impedance as a result of a fault in the system. FIGS. 5, 6, and 7 show a first technigue for fault detection. This technique is a continuous method of scanning to determine whether a fault is present. The method relies on the phase change at a particular frequency that occurs when the resistance (where the fault lowers the value of the resistance) of an RC circuit changes. FIG. 5A is a representative schematic of the load in a power distribution system without a fault, and FIG. 5B is a representative schematic of the load in a power distribution system with a fault. The phase response spectra of the load with and without the fault is shown in FIG. 5C. FIGS. 5A-5C disclosed below describes how the phase changes when the resistance is lowered. FIG. 5A shows the impedance and the phase of an RC circuit when no-fault is present. FIG. 5B shows the impedance and the phase of an RC circuit when the fault is present (and hence the phase is lowered due to the lower resistance). FIG. 5C shows the phase change that will occur when sampling at the “sense-frequency”. The sense-frequency should be set at the point of maximum phase change in order to get a strong signal.
[0028] A schematic of a power distribution system that incorporate this detection method is shown below in FIG. 6. FIG. 6 is a block diagram of a power distribution system with fault detection and bidirectional communication capability. The F.sub.1 Block in both the source and load represents a band reject filter at the frequency of F.sub.1. Hence at the frequency of F.sub.1, the impedance as seen looking into this block (using the person stick figure as a reference) is very high and does not pass through to the source or load. If a fault condition occurs (via a person in contact with both wires), the phase change (from normal to fault condition) will be detected (by the phase detector) and the power will be turned off for a period of time. Data communication from the source to the load utilizing 2-tone communications. Data communications from the load to the source is accomplished by the load varying the capacitive load and the source detecting the phase change. Note that this technique can work in both the idle period and the power transfer period (as long as a band reject filter at the sense-frequency is applied in front of the power source and the load).
[0029] FIG. 7A-7B shown below shows the magnitude and phase spectra of a power distribution system using the phase shift technique. FIG. 7A is a magnitude and phase frequency response centered about the frequency F.sub.1, with a fault present. FIG. 7B is a magnitude and phase frequency response centered about the frequency F.sub.1, without a fault present.
[0030] Data communication can be accomplished as follows. In order to communicate from the transmitter to the receiver, the oscillator in the transmitter can transmit two frequencies (one for a logic one and another frequency for a logic zero). This technique is referred to as a two-tone communication system. The receiver simply detects these tones. In order to communicate from the receiver to the transmitter, the receiver can either employ the two-tone system or change the load characteristic at a particular frequency such that the transmitter can detect these load characteristic changes.
[0031] FIG. 8 shows a second technique for fault detection. This technique is a continuous method of scanning to determine whether a fault is present and is described in FIG. 8 shown below. This method relies on two (one from the transmitter and one from the receiver) of the four shown current measurements (I.sub.src-p & I.sub.load-p or I.sub.src-n & I.sub.load-n or I.sub.src-p & I.sub.load-n or I.sub.src-n & I.sub.load-p) and a communication technique between the receiver and the transmitter. The technique works by determining the magnitude of the fault current I.sub.fault by subtracting the two measured currents. The transmitter must receive the measured current from the receiver side in order to perform the current subtraction and if the calculated fault current is high enough the power source in the transmitter is opened to prevent injury. The communication technique can be the two-tone method described in method #1. FIG. 8 shows a block diagram of another type of a pulsed power distribution system with fault detection utilizing current measurement at the source and at the load. Note that if there is a fault, the fault current will cause the current Isrc-p and Isrc-n to increase while not effecting the currents Iload-p nor Iload-n. Hence monitoring one of the currents in the source and one in the load will be sufficient in determining whether a fault has occurred. The fault detection determination can occur during the power or the idle cycle. Also note that the pulsed power during the idle time does not have to go to zero, instead it could go to a lower continuously safe level (e.g., 60 VDC).
[0032] FIG. 9 shows a third technique for fault detection. This technique is a continuous method of scanning to determine whether a fault is present and is described in FIG. 9 shown below. This method relies on a signal-tone that is transmitted towards the receiver. The magnitude of the signal-tone will depend if the fault is present or not. If the fault is not present, the impedance as seen by the signal-tone is very high due to the notch filters centered at the tone frequency. If a fault is present, the impedance as seen by the signal-tone will be lower and hence detectable within the transmitter circuitry.
[0033] The communication technique can be the two-tone method described in method #1. FIG. 9 is a block diagram of another type of a pulsed power distribution system with fault detection utilizing impedance measurement at the source. The impedance measurement is made at a frequency F1 during the power or idle cycle (outside of the transition areas of the pulsed power waveform). The additional notch filter shown in the source block is centered at the pulsed power switching frequency to provide additional attenuation of the frequencies induced from the switching edges. Data communication from the source to the load utilizing 2-tone communications. Data communications from the load to the source is accomplished by the load varying the resistive load and the source detecting this impedance change.
[0034] FIG. 10 shows a fourth method of fault detection. This technique is a method of fault scanning to determine whether a fault is present and is described in FIG. 10 shown below. The detection of a fault is determined during the idle portion of the periodic pulsed power waveform. During this interval, a 10 V periodic pulsed signal is applied. Under normal conditions, the 10V peak voltage at the source gets voltage divided down to 5V at the load. If a fault is present, this 5V at the load will be reduced down to approximately 1.5V (due to the 2 kΩ presence of the fault). This reduction from 5V to 1.5V can be detected in the power source equipment and appropriately keep the power off. The power should stay off for a period of time that is determined to be a safe interval, and then return to a normal state.
[0035] Data communication can be accomplished in this method as follows. In order to communicate from the transmitter (power source) to the receiver (load), a data signal from the source operating from 10-12V is used in a pulse width modulation (or pulse coded modulation) scheme. The load would have the 50 kΩ resistor switched in during source to load communication. In order to communicate from the receiver (load) to the transmitter (source), the 50 kΩ resistor is turned on and off to create a pulse width (or pulse code) modulated signal. Here the source would detect the amplitudes between 5 and 4 volts (here the transmitter is injecting a constant 10 volts). Of course this is just one way of providing a communication channel between the source and the load, there are many other ways of implementing this. FIG. 10 is a block diagram of another type of a pulsed power distribution system with fault detection utilizing amplitude attenuation detection method. During the idle portion of the pulsed power waveform, the 10V supply is being modulated by the load resistance. Under normal conditions the minimum of the periodic signal is ½ of the 10V power supply (5V) and under a fault, the minimum voltage drops to data communication from the source to the load utilizing the 10-volt pulsed waveform (e.g., number of pulses modulation, pulse width modulation, pulse code modulation, . . . ). Data communications from the load to the source is accomplished by the load varying the resistive load and the source detecting this amplitude change.
