The present invention relates to electricity supply apparatus and electricity supply methods.

Conventional electricity distribution grids typically supply sinusoidally alternating current (AC) electricity with a frequency of 50 or 60 Hz. Throughout the rest of this specification, the term 'AC is used as short-hand to refer to sinusoidally alternating current (or equivalently sinusoidally alternating voltage). However, in less developed countries or in remote regions that are currently not connected to an electricity distribution grid, electricity is often generated by techniques that produce a direct current, such as fuel cells or solar photovoltaic panels, often with batteries as storage and backup.

Conventionally, the direct current (DC) is converted to a AC using an inverter. However, most appliances, such as for cooking, lighting, radio, TV and computers, either require DC or would run satisfactorily off a DC supply. With an AC supply, the instantaneous power delivered is continuously changing over the cycle, whereas most appliances work at a constant level of power. To cope with the continuously changing power, inverters typically require large capacitors. These capacitors are bulky and can contribute up to 50% of the cost of a power inverter, making the inverter large and expensive. Even with large capacitors, there can still be a large ripple in the current drawn from the DC source, which can be damaging, and shortens the life of batteries. The electrical installation must also be able to cope with the peak current and/or peak voltage of the AC supply, which are significantly higher than the mean.

Instead of using invertors, if electricity is distributed as direct current throughout a building, such as a house, then in order to deliver power without requiring very high current and very thick wires, a reasonably high voltage is required, such as around 40 volts or higher. However, with DC, large and expensive arc breakers are required to be installed because it is difficult to safely interrupt direct current. Furthermore, distributing DC has problems such as it being more difficult to step up or step down the DC voltage; one cannot simply use a transformer.

In addition to the above issues, there are further problems. For example, at an off- grid location using a DC supply, when it does eventually become connected to a mains AC grid, there is the problem of how to install and integrate the AC supply with the existing DC system without requiring complete rewiring. Similarly, at a location that is connected to the AC grid, which then subsequently installs DC generation capacity, such as solar photovoltaic cells and batteries for storage, there is the problem of how to integrate the AC and DC systems that avoids conventional inverters and/or rewiring, and avoids rapid draining of batteries in the event of interruption of the AC supply. There is a further problem of how to provide electrical protection in the event of a leakage current fault in a system using AC and DC.

The present invention has been made in view of the above problems.

Accordingly, the present invention provides an electricity supply apparatus comprising:

an input for receiving a direct current electricity supply; and

a converter arranged to convert the electricity supply received by the input to generate, at an output, a voltage waveform of alternating polarity,

wherein the voltage waveform alternates in repeated cycles;

wherein a value A is given by the integral over a half cycle of the voltage waveform of the difference between:

(i) the square of the instantaneous voltage; and

(ii) the average value of the square of the voltage,

the integral being taken only when the square of the instantaneous voltage exceeds the average value of the square of the voltage, and

a value B is given by the integral over a half cycle of the voltage waveform of the square of the instantaneous voltage,

wherein the ratio A:B is not greater than 0.2.

Another aspect of the present invention provides an electricity supply method comprising:

receiving a direct current electricity supply; and

converting the received electricity supply to generate an output having a voltage waveform of alternating polarity,

wherein the voltage waveform alternates in repeated cycles; and

wherein a value A is given by the integral over a half cycle of the voltage waveform of the difference between:

(i) the square of the instantaneous voltage; and (ii) the average value of the square of the voltage,

the integral being taken only when the square of the instantaneous voltage exceeds the average value of the square of the voltage, and

a value B is given by the integral over a half cycle of the voltage waveform of the square of the instantaneous voltage, and

wherein the ratio A:B is not greater than 0.2.

A further aspect of the present invention provides an electricity supply interrupter, for use with an electricity supply comprising a substantially sinusoidally alternating electricity supply provided across first and second output conductors, and an output voltage waveform provided across a third output conductor and said second output conductor,

wherein the voltage waveform is of alternating polarity;

wherein the voltage waveform alternates in repeated cycles;

wherein a value A is given by the integral over a half cycle of the voltage waveform of the difference between:

(i) the square of the instantaneous voltage; and

(ii) the average value of the square of the voltage,

the integral being taken only when the square of the instantaneous voltage exceeds the average value of the square of the voltage, and

a value B is given by the integral over a half cycle of the voltage waveform of the square of the instantaneous voltage,

wherein the ratio A:B is not greater than 0.2; and

wherein the voltage waveform is in anti-phase with the substantially sinusoidally alternating electricity supply,

the interrupter comprising a sensor that senses the net current flowing through the first, second and third conductors, and interrupts the electricity supply if the net current exceeds a threshold current.

Embodiments of the invention will now be described, by way of example only, with reference to the accompanying drawings in which:

Fig. 1 is a graph of the voltage waveform of a conventional AC supply;

Fig. 2 is a graph of the square of the voltage of an AC supply; Fig. 3 is a schematic graph of the voltage output of an electricity supply

apparatusaccording to an embodiment of the invention;

Fig. 4 is a schematic graph of the square of the voltage output of an apparatus according to an embodiment of the invention;

Fig. 5 is a schematic of an electricity supply apparatus according to an embodiment of the invention;

Fig. 6 is a more detailed schematic circuit diagram of a portion of an embodiment of the apparatus of Figure 5;

Fig. 7 is a schematic of an electricity supply apparatus according to a further embodiment;

Fig. 8 is a schematic of an electricity supply apparatus according to a further embodiment;

Fig. 9 is a more detailed schematic circuit diagram of a portion of the apparatus of Figure 8;

Fig. 10 illustrates a pulse width modulation (PWM) pulse train and its filtered sinusoidal output according to the embodiment of Fig. 9; and

Fig. 11 is a diagram of an electricity supply interrupter according to a further embodiment. To assist understanding of the present invention, comparison will be made between voltage waveforms of a conventional supply and those used in one embodiment of the invention. Figs. 1 to 4 are purely schematic and the axes are in arbitrary units. Figure 1 shows the voltage waveform of a conventional AC supply. Figure 2 shows the square of the voltage of Figure 1; the power supplied being proportional to the square of the voltage. In Figure 2, the average value M of the square of the voltage is shown by a horizontal line. If the AC supply is produced using an inverter, then capacitors or similar must store energy when the instantaneous square of the voltage is less than the average of the square of the voltage, and must supply energy when the instantaneous square of the voltage is greater than the average square of the voltage. The area indicated A represents the amount of energy that must be shifted within a half cycle (i.e. needs to be temporarily stored within a capacitor and/or inductor). Area A is defined as the integral over a half cycle of the difference between: (i) the square of the instantaneous voltage; and (ii) the average value of the square of the voltage, the integral being taken only when the square of the instantaneous voltage exceeds the average value of the square of the voltage.

The total power supplied over one half-cycle is proportional to the area indicated by B in Figure 2 (i.e. the integral over one half cycle of the square of the instantaneous voltage), and this is, of course, also equal to the area indicated by C (the integral of the average value of the square of the voltage over one half cycle).

One metric of the performance of the supply is given by the ratio of A to B because this represents the relative provision for energy storage that must be made. For a conventional DC to AC inverter, the ratio is just over 0.3 (i.e. just over 30%).

An output supply voltage waveform produced by an apparatus and a method according to one embodiment of the invention is illustrated in Figure 3. It is a substantially square wave, or equivalently can be considered as a DC supply that periodically alternates in polarity. Figure 4 shows the corresponding square of the voltage (proportional to the power). The areas labelled A, B and C have the same definition as explained above with reference to Fig 2. In practice, the corners of the square wave are never perfect, and the reversal of polarity is not instantaneous, so the average value of the square of the voltage is just below the peak value of the square of the voltage, and a very small area A is present. However, the ratio of A to B is less than 0.3. According to embodiments of the invention, the ratio of A to B is not more than 0.2, such as less than 0.1, and in some embodiments is less than 0.05. For a perfect square wave the area A would be zero, and so the ratio A:B would also be zero.

Other examples of the voltage waveform are possible according to other

embodiments of the invention; for example there may be ripple on the horizontal portions of the square waves, or the waveform could represent just a few or more of the lower odd harmonics of the fundamental frequency, such as the first plus third or first plus third plus fifth and so forth. However, these would still reduce the ratio of A:B to 0.2 or less, and so reduce the need for capacitors or other energy storage.

Figure 5 illustrates schematically an electricity supply apparatus according to an embodiment. The converter 10 receives a DC electricity power supply at input 12 and generates a voltage waveform of alternating polarity that is a substantially square waveform at output 14. The output 14 is connected to a plurality of electrical outlets 16 provided at a location, such as a building 18 e.g. a house. The converter 10 can, of course, be located within the building 18. The converter 10 provides power to the outlets 16. The output 14 can be connected to the outlets 16 by any suitable wiring configuration, such as a ring circuit or radial spurs. The electrical outlets 16 can also be of any suitable type or mixture of types, such as sockets for receiving plugs, or hard-wired outlets for specific appliances.

The construction of the converter 10 according to one embodiment is illustrated in Figure 6. It consists of an H-bridge of four power switches Jl, J2, J3 and J4 connected as shown. Integrated power modules are commercially available in which all of the power switches are integrated into a single chip. Signals applied to the gates Gl, G2, G3 and G4 of the respective switches Jl to J4 are used to switch the switches on and off (i.e.

conductive and non-conductive, or active and passive).

In use, for one half of a cycle, Jl and J4 are switched on (with J2 and J3 off), and for the other half cycle J2 and J3 are switched on and Jl and J4 are switched off. From the DC supply at input 12, this produces a voltage waveform of alternating polarity at output 14 in which the polarity switches every half cycle. The output is a substantially square wave, and the peak-to-peak voltage amplitude is double the input DC voltage.

A microcontroller 20 produces the gate driving signals. In the configuration shown, a common first gate driving signal can be applied to gates Gl and G4, and a common second gate driving signal can be applied to gates G2 and G3. The first and second gate driving signals are essentially low voltage square wave signals in anti-phase with each other. The duty cycle is preferably set to slightly less than 0.5 such that the "on" time is less than the "off time so that there is a dead time when none of the switches Jl to J4 is on. This ensures that there is never a short-circuit condition in which say both Jl and J2 were both in the conductive state simultaneously.

The frequency of the gate driving signals produced by the microcontroller 20, and hence the frequency of the fundamental frequency of the voltage waveform at the output 14 could be any suitable frequency, such as in the range from 20 to 200 hertz. However, the frequency is typically set to correspond to existing electricity distribution systems, such as 50 hertz or 60 hertz. The microcontroller 20 can have its own oscillator to provide a reference frequency for governing the gate driving signals, or can be provided with a dedicated separate oscillator, or can be fed with a signal having a reference frequency from an external source. The peak-to-peak voltage of the waveform at the output 14 in preferred embodiments for electrical power applications is at least 40 volts. In one specific embodiment, the input DC voltage is 48 volts (this is a value typically supplied by DC sources, and is a multiple of 12 volts, so is compatible with lead-acid batteries). This produces a peak-to-peak voltage output of 96 volts.

An embodiment of the electricity supply apparatus for use in electrical power applications should be capable of supplying at least 100 watts of electrical power, such as at least 400 watts. In an embodiment in which the input voltage is 48 volts, the output is rated at about 10 amps, so could deliver a maximum theoretical output of 480 watts.

Although the voltage waveform output by the electricity supply apparatusembodying the invention is based essentially on a constant DC voltage, the fact that it alternates in polarity means that any arc formed across the contacts of a circuit breaker is rapidly extinguished because the polarity changes every half cycle. Therefore the apparatus does not require expensive DC arc breakers, but can use conventional AC circuit breakers.

The fact that the polarity of the voltage alternates also means that the voltage can be stepped-up or stepped-down using a transformer, because the current is not constant.

Other circuits for changing the voltage can, of course, be used, such as a switched mode power supply. Some of the electrical outlets 16 can be provided with circuitry to convert the substantially square wave supply to a low voltage DC supply, such as 5 volts DC for use with powering electrical equipment, such as via a USB connection.

An electricity supply apparatus according to a further embodiment is illustrated in context in Figure 7, and a still further embodiment is illustrated in Figures 8 and 9. Many of the features of these embodiments are the same as Figures 5 and 6, so description thereof will not be given to avoid repetition. The same reference numerals are used to indicate corresponding items. The properties, features and uses of these embodiments are generally the same as the previously described embodiments, except as explained below.

The main change of the embodiment of Figure 7 with respect to the embodiment of Figure 5 is that a mains AC supply is present in addition to the DC supply. The AC supply may be derived, for example, from an electricity distribution grid. The two conductors are labelled L and N for live and neutral respectively. The converter 10 is additionally connected to the AC supply via an input 30. The input 30 is passed to circuitry in the converter 10 that detects the phase of the AC supply, such as a phase detector, a phase- locked loop, or a zero-crossing detector. The result of the phase detector is passed to the microcontroller 20 (see Figure 6) which is then used to control the timing of the gate driving signals such that the switches Jl to J4 are switched to generate the substantially square wave output that has the same fundamental frequency as the AC supply. One of the conductors 32 from the output 14 of the converter 10 is connected to the neutral conductor N of the AC supply. The phase of the switching of the gate drive signals is such that the waveform on the other output conductor SQ relative to N is in anti-phase with respect to the voltage waveform on the live conductor L relative to the neutral conductor N. The small waveform sketches in Figure 7 illustrate this.

A further embodiment of an electricity supply apparatus is illustrated in Figure 8.

This can be used where only a DC supply is available, such as an off-grid location where mains electricity is not yet provided. The converter 10 receives the DC input and generates output on conductors L, N and SQ that have the same properties as those in the example of Figure 7, namely, a sinusoidal voltage across L and N, and a substantially square wave, in anti-phase with the sinusoidal wave, across conductors SQ and N.

An embodiment of the converter 10 for generating this output is illustrated in more detail in Figure 9. In this case, power switches Jl to J6 are used, but configured as three half-bridges: Jl & J2; J3 & J4; and J5 & J6. To generate the sinusoidal wave, pulse width modulation (PWM) is used. In this technique, the microcontroller 20 generates gate driving signals for Jl and J2 that comprise a high frequency pulse train of varying duty cycle. For a 50 hertz output, the pulse train frequency in this embodiment is 250 hertz, i.e. at least several multiples of the desired output frequency. The frequency of variation in the duty cycle matches the desired frequency of the output sinusoidal wave. The variation in output power of the pulse train therefore matches the variation in output of a sinusoidal wave. The resulting power pulse train, obtained by alternately switching Jl and J2 in opposite phase to each other, is passed through a low pass filter (LPF) 40 with a cut off frequency above the fundamental frequency of the required sinusoidal output. The high frequency switching of the pulse train itself is filtered out leaving only the low frequency variations in the duty cycle in the form of a sinusoidal wave. Figure 10 illustrates a pulse width modulated pulse train and the resulting filtered sinusoidal wave output.

To generate the substantially square wave output on conductor SQ, switches J5 and J6 are alternately switched on and off in anti-phase with each other by gate driving signals supplied by the microcontroller 20. These gate driving signals are square waves with a duty cycle of 0.5, and a frequency the same as the output sinusoidal wave on the conductor L, such as 50 hertz. The gate driving signals are synchronised such that the substantially square wave output is in anti-phase with the sinusoidal output. The LPF 40 may introduce a phase shift in the sinusoidal wave, so, if required, phase angle correction can be employed by the microcontroller 20 when generating the gate driving signals (for either or both of these half bridges) in order to set the phase difference between the square and sinusoidal wave outputs to 180 degrees or close thereto, such that they are substantially in anti-phase.

The half bridges described so far, of course, each only has one output, and so is unable to reverse the voltage direction (polarity) on its own. What is required is a further conductor to which the load can be connected which acts as a return wire or virtual ground, relative to which the square and sinusoidal voltage waveforms on the conductors SQ and L swing both positive and negative. This can be achieved using a further half bridge comprising switches J3 and J4 which are driven in anti-phase by a pulse train with a duty cycle of 0.5 at a significantly higher frequency for example in the range of 1 kilohertz to 100 kilohertz, such as 25 kilohertz. This produces an output that rapidly switches between the positive and negative DC supply voltages, which is then passed through a low-pass filter 42 to produce a mean voltage half way between the DC peak to peak voltage. This mean voltage is then presented on the output conductor N. Alternatively, if the DC source comprises two DC sub-sources of equal voltage (e.g. two batteries or battery banks) that can be connected in series in the same voltage direction, then the point of connection between them can provide the return or neutral contact at a potential half way between the positive and negative terminals.

The low pass filters 40 and 42 can be of any suitable design, such as a T-section filter or an LC filter, with a cut-off frequency above the supply frequency and below the frequency of the PWM pulse train.

It will be recognised that all of the preceding embodiments effectively produce a two-phase power supply, with one phase being sinusoidal and the other phase being a substantially square wave. If the loads on the two phases are approximately balanced, or even partially balanced, then the current flow on the "return" conductor (i.e. the neutral conductor N) is substantially reduced. This represents a saving in ohmic losses in the wiring at the location such as the building 18, compared to having totally separate wiring for the two phases, and so the efficiency is improved.

In certain applications, the electricity supply can be distributed using conventional three-core wiring for the L, N, and SQ conductors. Many modern appliances are adequately insulated and so their plugs connect only to the live and neutral terminals of a conventional electrical outlet, and they are not earthed. In existing installations, the potential of the neutral conductor is often very close to earth in any case. However, a separate earth conductor can be provided if appropriate.

It is envisaged that existing appliances would connect via a conventional plug to only the L and N conductors of an electrical outlet. Other appliances able to use the

substantially square wave electricity supply have a plug that connects to the N conductor and to the third conductor (SQ), but has a dummy pin for the L conductor in the electrical outlet, for example a suitably shaped pin made of plastic or other non-conductive material. Yet further appliances may be adapted to use both the sinusoidal AC and the substantially square wave supplies, and so would connect to all three conductors in the electrical outlet. For appliances such as a washing machine or air conditioning unit, the sinusoidal supply could be used for motors and other higher power loads, and the substantially square wave which can be readily converted to DC could be used to power the sensors and control electronics.

With the embodiment of Figure 7, in which a conventional AC supply is integrated with a DC supply, and the DC supply includes batteries or fuel cells, then the supply apparatus can act as a so-called uninterruptable power supply (UPS). In this case it is envisaged that high power, low value loads (heating, motors etc) be connected to use the conventional AC supply (either as stand-alone appliances or within hybrid appliances as just described), and low power, high value loads, such as computers and audio visual equipment are connected to be powered by the substantially square wave supply derived from the DC source. If there are power blackouts on the AC distribution grid, then important loads can still be powered from the DC source, which is typically locally generated. Furthermore, the higher power loads are not driven by the DC supply, and so when there is a power cut on the AC grid, the local batteries will not be rapidly drained by high power loads. A further aspect of the invention concerns how to provide electrical safety in a hybrid electricity supply such as in the embodiments of Figures 7 or 8. Over-current protection can be provided by conventional circuit breakers, but protection is also desirable against current leakage. A protection device for this latter type of fault is known under various names, such as a residual-current device (RCD), a residual-current circuit breaker (RCCB), a ground fault circuit interrupter (GFCI), ground fault interrupter (GFI) or an appliance leakage current interrupter (ALCI). Such a device will be referred to herein simply as an interrupter. An embodiment of an interrupter is illustrated in Figure 11. It comprises a differential current transformer having a magnetisable core 100 and a secondary winding 102 around the core. The three conductors L, N, and SQ as described in the previous embodiments of the electricity supply apparatus pass through the core 100. In normal operation, the currents flowing and returning through these conductors at any instant sum to zero, and so no current is induced in the secondary winding 102. The total sum of the currents in the three conductors, taking into account the direction of current flow, is also referred to as the net current. In a fault condition, when current passes to ground for example through a person touching live wiring, then this leakage current means that the net current is not zero, and so a current is induced in the secondary winding 102. The induction in the secondary winding 102 activates electronics 104 (such as a solenoid) to open the contacts 106 to interrupt the electricity supply, if the leakage current exceeds a threshold. The core 100, secondary winding 102 and electronics 104 can together be considered to constitute a sensor that senses the net current flowing through the three conductors.

The speed of operation and minimum leakage current necessary to cause the sensor to interrupt the electricity supply can be in accordance with conventional safety devices to avoid electrocution, for example the supply can be interrupted in under 40 milliseconds if any leakage current (i.e. the magnitude of the net current) greater than 30 milliamps is detected.