GB2284100A - Electrical switch - Google Patents

Abstract

In one arrangement, a hybrid relay comprises the series combination of a semiconductor switch 32 and a resistor 34 and/or inductor coupled in parallel across electromechanical switched relay contacts 36 arranged such that the semiconductor switch closes before the relay contacts. Preferably the time of energisation of one or both of the relay coil 37 and semiconductor switch are controlled relative to zero crossings of an ac power supply L, N. High in rush current and associated transients on switching on can be reduced, component ratings can be lowered, and/or faster acting circuit breakers can be employed. In another arrangement, a microprocessor (78, Fig. 7) is used to energise a relay and to sense the instant of contact closure, whereby to alter the start of energisation of the relay if the instant of contact closure lies outside a predetermined time range relative to the a.c. supply cycle so that contact closure occurs within or closer to the predetermined range during a subsequent operation of the relay. <IMAGE>

Description

Electrical Switch This invention relates to electrical switches, and, in particular, to electrical switches comprising an electromechanical relay such as an electromagnetic relay.

The most common types of electrical switch are those incorporating mechanical contacts which are moved to make or break a circuit (hereafter referred to as mechanical switches), and solid state switches in which the state of a semiconductor device is altered between a generally insulating and a generally conducting condition, often in response to an electrical control signal.

Mechanical switches can be made to have an extremely low resistance in the ON condition, which is desirable particularly when significant current flow is involved insofar as energy loss and heat generation within the switch itself are low. It is relatively easy to design a mechanical switch capable of withstanding a high voltage between the open contacts in the OFF condition, and in such condition the resistance is effectively infinite.

Furthermore, such switches are often capable of surviving abuses, such as short lived exposure to current flows well above their continuous rating.

Disadvantages of mechanical switches are that they possess mechanical inertia, so that the precise time of operation of the contacts tends to be indefinite, and many exhibit cont-act bounce on closure. They also wear out.

The life of mechanical switch contacts is primarily dependent upon the operating conditions when switching occurs. Contacts normally have to be sized to withstand the transient phenomena which occur on making and breaking the circuit.

The most damaging transient phenomenon is the formation of an arc between the contacts. Arcs cause erosion of contact material by localised evaporation of material from the contact surfaces. In addition, they may cause the contacts to weld together, so preventing them from opening normally.

These problems are normally minimised by increasing the mass of the contacts, by careful design and by careful selection of the contact materials.

Arc erosion is a particular problem where highly inductive or capacitive circuits are being switched, as, for example where large banks of power factor corrected fluorescent lights are installed in commercial or industrial premises.

In the particular case of capacitive loads, the high inrush current which occurs when the circuit is made causes extremely damaging arcs to form between the switched contacts when they inevitably bounce on making. The effect is exacerbated by the inductance of associated wiring, etc., which will produce an overvoltage on contact bounce.

A high inrush current can also be associated with other types of load, e.g. incandescent lamps (where the resistance is initially low, but then increases) and electric motors.

It is commonly necessary for special materials such as precious metals, and materials such -as cadmium which are hazardous to health and safety, to be used for the switched contacts, and these can lead to expense and difficulty in manufacturing. The more onerous the switching conditions, the more expensive the alloys that need to be used, and the larger the contacts have to be. In addition, larger forces required to operate the contacts (for increased contact pressure, speed of operation) lead to higher power requirements for the magnetic circuit of an electromagnetic relay, for example.

Unlike mechanical switches, solid state switches possess no significant inertia and so can be switched at precisely determined moments. They also have no contacts to wear out. However, in the ON state they still retain a degree of resistance which leads to greater energy dissipation and heat generation within the switch itself; and they tend to be destroyed relatively rapidly once their ratings are exceeded.

When it is desired to switch a load which presents problems such as those outlined above, using an electromechanical relay, clearly it is possible to specify a relay sized to be capable of dealing with the conditions imposed during switching, but this approach can be expensive, and fails to solve associated problems.

When a simple electromechanical relay is used to switch a load, the switching action itself, and the accompanying bounce of mechanical contacts, will both tend to create transients which can be transmitted both to the load and back through the power supply.

Furthermore, the time of switching relative to an ac supply is not precisely controlled. For many loads, inrush current will be maximised if the circuit is completed when the ac voltage is at a positive or negative peak, and may be so large as to trip a circuit breaker which is conventionally provided in the power supply. For this reason, slower acting circuit breakers are often provided where the initial current is expected to be much greater than the continuous current, but this in itself may lead to a need to uprate associated components, including the wiring from the power supply itself.

Figures 1 and 2 show time/current characteristics of types 2 and 3 circuit breakers to BS 3871 Part 1, which are commonly used, the latter particularly when high inrush currents might occur, to provide a delayed tripping action (there are other types, e.g. type 4, for even greater delay before tripping). The maximum continuous current accommodated by each type lies in the region at and just above the unit current multiplier at the top of each curve.

As the current exceeds this value, the time to trip the breaker decreases along the curved region, until a generally vertical region appears at which the time to trip decreases very sharply. For type 2 breakers the current in this lower vertical region is no more than about 7 times the continuous current tripping figure, and the corresponding ratio for type 3 breakers is no more than about 10. Thus a current which is about 12 times the maximum continuous current figure should be effective to trip either of these common types of circuit breaker in a very short period, even if the less common types would take longer to trip.

Hybrid relays are known in which switched contacts of an electromechanical relay and power terminals of a semiconductor switch are coupled in parallel, the control inputs of the relay and switch being coupled together.

Because the semiconductor switch operates faster than the electromechanical relay, such an arrangement serves to overcome the problems of contact bounce and arc erosion when the switch is closed. In addition, the lower ON resistance of the switched contacts effectively bypass the semiconductor switch during continuous operation, and it is known to turn the semiconductor switch OFF during such operation, e.g. to reduce heating in the semiconductor switch due to continued application of the control input.

In its rudimentary form, however, a hybrid relay does not address the other problems created by an inrush current, nor does it cope with arc formation on opening of the hybrid relay, should the switched relay contacts open after the semiconductor switch. Furthermore the semiconductor switch is vulnerable to the effect of the high inrush current when it alone is conducting and so must be of a sufficient rating to cope therewith.

A modified hybrid relay, using a triac in parallel with switched electromechanical relay contacts, is shown in JP 57-27527 A. The relay has two pairs of contacts which are closed sequentially. The first pair of contacts to close is connected in the gate circuit of the triac so that the triac is then essentially immediately turned on to couple the supply to the load. The second pair of contacts closes later and provides a second path for coupling the supply to the load.

In this arrangement, the triac does not necessarily switch on close to zero crossings of the supply voltage, and so must have a sufficiently high switching capacity to cope with switching (and any associated large inrush currents) at any part of the supply voltage waveform. After switchon, once the current is flowing through the load, the lower contact resistance of the switched relay contacts provides greater efficiency, and less heat generation.

Nevertheless, the triac is turned on all the time that the relay is energised, and this does give rise to energy loss and heat generation in the gate drive circuit.

On switching off, the second pair of relay contacts open, and subsequently the gate current through the triac is turned off (in that order, due to the contact arrangement), but the triac will continue to conduct until the next time zero current flows therethrough, which may be advantageous in certain instances (for example, inductive loads).

British Patent Application No. 20907d2 discloses a switch arrangement for a three phase power circuit intended to minimise DC current offset generated by the voltage integrating property of an inductive load. Hybrid relays in each of the three phases are operated (opening and closing) at predetermined times with respect with respect to the power supply voltage waveform. In particular, prior to the closure of the electromechanical relay contacts, in one phase the semiconductor switch is turned ON at 100 before a zero voltage crossing, and in the other two phases closure is at about 900. During de-energisation, the semiconductor switches are opened at the same time, following opening of the electromechanical relay contacts.

Optionally the semiconductor switches are opened during continuous operation, and closed only for periods which bridge the operation of the switched contacts.

While this arrangement may fulfil its intended purpose, the inductive nature of the load should prevent the development of a large inrush current. Substitution by a capacitive load potentially could lead to high inrush currents and destruction of the semiconductor switches, etc.

Embodiments of the present invention in its several aspects each seek to deal with at least one of the problems outlined above.

In a first aspect the present invention provides a hybrid relay for coupling a power supply to a load, comprising an electromechanical relay having at least one pair of switched contacts for coupling to the supply and the load respectively, a series combination of a semiconductor switch and a resistor and/or inductor being coupled in parallel across said pair of switched contacts, a relay control input of the electromechanical relay and a switch control input of the semiconductor switch being coupled to a command input of the hybrid relay to receive respective energisation signals such that the semiconductor switch closes before the switched contacts.

In this arrangement, when the semiconductor switch initially turns on, the resistor and/or inductor will limit the initial current, and the rating of the semiconductor switch may be correspondingly reduced. With an ac supply the energisation signal may occur at any time relative to the supply waveform. The switched contacts of the relay will subsequently close, but the voltage across the relay contacts will have been reduced by the action of the semiconductor switch, so that arc erosion will be reduced or avoided. The value of the resistor and/or inductor is a compromise between limiting the inrush current through the semiconductor switch so as to protect the latter, and providing sufficient current flow (a) to significantly reduce the voltage across the relay contacts in the case of a correct load, and (b) to lead to rapid operation of a circuit breaker in the case of a short circuit load. This aspect will be discussed later.

Preferably the resistor is a wirewound component.

In general, the semiconductor switch of a hybrid relay is operated such as to provide electrical isolation between the power source and the command signal, for example by the use of optical isolating means such as optocouplers and opto-diacs, or transformer coupling. However, the use of the term "hybrid relay" herein is not intended to specify that such isolation, although often desirable, is an essential part of the invention.

Where the hybrid relay is intended to be used with an ac power supply, each of the signals provided to the control inputs of the semiconductor switch and electromechanical relay may, or may not, be controlled to commence with a predetermined timing relative to the ac voltage waveform.

Where no such timing control is provided, so that both the semiconductor switch and the electromechanical relay are energised simultaneously, the inrush current can be potentially large, even if limited by the resistor and/or inductor in series with the semiconductor switch. Where the time of closure of the semiconductor switch can be precisely controlled and/or when the time of closure of the switched relay contacts can be at least approximately controlled, e.g. relative to voltage zero crossings, it is possible to limit the amplitude of the inrush current by arranging that the voltage across the semiconductor switch is also low when it is first closed, and/or to limit the voltage across the switched contacts as they close.

It is preferred to operate the electromechanical relay as soon as possible after operation of the semiconductor switch, for example, so as to minimise power loss in the circuit due to the voltage drop across the semiconductor switch. Thus preferably the relay is such that the contacts will close within one ac cycle, more preferably within half an ac cycle (20 milliseconds and 10 milliseconds respectively for a 50 Hz supply), of the commencement of said energising signal.

However, when the timing of the control input to the semiconductor switch is controlled, it may then become necessary to delay the signal to the input of the electromechanical relay in order to ensure that the semiconductor switch closes first.

It is also preferred to terminate operation of the semiconductor switch within a short time after switching of the electromechanical relay has occurred. Thus it is typically operated only for 1M to 2 .cycles (at 50 Hz) of the ac supply. Where it is a triac, for example, this avoids the loss of power that would otherwise occur in the gate circuit of the triac. Assuming that the relay contacts provide a very low resistance when closed, the loss of power across the main terminals of the triac should be minimal once the relay contacts have closed. Thus preferably the semiconductor switch is turned on only for a predetermined period which is sufficiently long to ensure that the relay contacts have closed, e.g. for no longer than 2 cycles of the ac supply (40 milliseconds at 50 Hz).

It has been found that when switching loads such as power factor corrected fluorescent lights, it is not always necessary to switch off at zero supply volts because problems such as arcing at the switched contacts are far less significant than those occurring at switch on.

Nevertheless a switch-off which is timed relative to the supply, or otherwise controlled, may still be advisable for other reasons, e.g. to impose minimum disturbance on the supply.

In addition, when other loads are encountered, for example loads which are inductive and/or not power factor corrected, the timing of the switch off may also assume importance.

Therefore the hybrid relay may be arranged to respond to a second state of the command signal to de-energise the electromechanical relay, and to turn said semiconductor switch on before the switched relay contacts have opened, and to turn said semiconductor switch off after the contacts have opened.

This may require additional components arranged to sense when the control signal changes from "ON" to "OFF". Deenergisation of the relay leads to opening of the switched contacts after a short time delay arising from mechanical and magnetic inertia, and the opening occurs at a time which is indeterminate relative to the ac supply voltage zero crossings. The additional components serve to switch the triac gate current on again before the relay contacts open, so as to conduct the current until the next time zero current flows in the triac after the contacts have opened.

Thus, as with switch on, arcing can be substantially reduced or eliminated, enabling the use of a cheaper and/or lower rated relay.

Operating the semiconductor switch at controlled instants and/or for controlled periods also leads to lower demands therefrom, enabling this component to be of a lower rating than would otherwise be the case.

For further features of the invention in its first aspect, reference should be made to the detailed description and to the appended claims.

European Patent Application No. 86102183.0 (serial number 0192258) discloses a circuit in which closure of an operating switch leads to the generation of pulses coincident with zero crossings of the ac supply. These pulses are delayed in two separate delay paths, and combined by logic circuitry to produce a single pulse for application to the coil of a relay. One delay path imposes a delay which is reflected in the start of this single pulse, which is timed such that the relay contacts are intended to close at the next zero crossing of the supply voltage after the operating switch has been closed. The single pulse ends after the operating switch has been opened, and the other delay path imposes a delay to the end of the single pulse such that opening of the relay contacts occurs at a time intended to coincide with the next zero crossing of the supply voltage.

However, the operating characteristics of commercially available relays, including the time between energisation of the relay and the time at which the contacts close, vary considerably. A circuit in which the relay is operated with a fixed timing or timings relative to zero crossings of the ac supply, with the intention that the contacts close and/or open substantially at a zero crossing time, can take no account of this, unless it is adapted to the particular relay in use, e.g. by suitable adjustment of the individual circuit. This is not, in general, a practical proposition. Moreover, it takes no account of the fact that the operating characteristics of a relay may vary over the lifetime of the relay.

In a second aspect the invention provides an electrical switch for altering the coupling between an ac supply to a load in response to a command signal, comprising an electromechanical relay having at least one pair of switched contacts for coupling to the supply and the load respectively, and a control circuit, wherein said control circuit comprises first control means for altering the state of energisation of said relay at a first time in response to a change in said command signal, said first time being controlled relative to the ac supply cycle, second control means for detecting the instant when the switched contacts have changed state relative to the ac supply cycle, and third control means for adaptively altering said controlled time if said instant lies outside a predetermined range, such as to bring said instant within, or closer to, said predetermined range, during a subsequent operation of said relay.

In this aspect, the invention is not necessarily in the form of a hybrid relay. Without a parallel semiconductor switch, there is no need to take measures to protect this relatively vulnerable component. The result is a switch which possesses the robustness associated with mechanical contacts, and which is operated in such a way that contact wear due to arc erosion can be greatly reduced. In addition, transient generation and the requirement for relatively highly rated components including the relay, the power wiring, and the type of circuit breaker, are also reduced. Alteration of the relay characteristics over time are also taken into account.

Where contact bounce is significant, the second control means may comprise circuitry such as logic circuitry for determining an instant when the switched contacts are considered to have changed state.

In this second aspect, it is recognised that different commercial relays, even of the same type, will exhibit variations in switching speed as a result of manufacturing tolerances, and that the speed of a single relay may vary over its lifetime, for example due to the effects of wear.

In this second aspect, the ac supply voltage cycle is monitored, and means are arranged to energise the relay at a controlled time during the cycle. The instant at which the switched contacts change state is measured relative to the ac supply cycle, and the controlled time to be used for a subsequent energisation is altered in a way such as tends to reduce any difference between the said instant and a predetermined value, preferably to zero. Alteration of the controlled time may be done whenever the said instant is measured and found to lie outside a predetermined range, or whenever the instant is determined not to coincide with a predetermined value, for example.

Measurement of the said instant may occur upon every energisation, or at selected or predetermined intervals.

In the limit this might be a single measurement or set of measurements (setting up the switching time) prior to use of the relay.

The change in the command signal may be such as to cause the load and supply to be coupled by closure of the switched contacts, an associated controlled time hereafter being referred to as a contact closure time, and an associated predetermined range being a predetermined closure range. Preferably this predetermined closure range includes the zero crossing of the ac supply voltage.

Where contact bounce occurs upon closure, the instant at which the switched contacts are considered to have changed state may be taken, for example, as the first closure, the cessation of bounce or an intermediate time, as desired, and means may be provided to measure the time during which bounce occurs. Where bounce is prolonged, it is preferable to arrange that an ac zero crossing occurs after the first closure, in such a way that the highest voltage across the switched relay contacts while bounce is occurring is minimised.

Alternatively, or additionally, a change in the command signal may be such as to cause the load and supply to be decoupled by opening of the relay contacts, an associated controlled time being a contact opening time, and an associated predetermined range being a predetermined opening range. The latter may include the zero crossing of the ac supply voltage; alternatively the control circuit also comprises fourth control means arranged to calculate an optimum period relative to the supply voltage cycle at which the energy in the circuit is at or close to a minimum, for use as said predetermined opening range.

When there is control of both opening and closure instants, the predetermined opening and closure ranges may be the same or different.

In a preferred embodiment, the control circuit comprises a microprocessor which serves to function as some or (more preferably) all of the control means. The microprocessor can be taught the characteristics of the relay at the time of manufacture, or at installation. Particularly when use is for switching a potentially high inrush current, this teaching phase is preferably performed with a substitute resistive load.

The use of a microprocessor has the advantage that it can be used in the sensing of the circuit characteristics, as referred to above, and calculate the optimum timing for switching the circuit off to ensure that the amount of residual energy in the circuit is at a minimum when the relay contacts open. (Often this will be at or close to zero current in the load.) Thus arcing may be reduced upon switching on and off.

The control circuit may also contain sensor means arranged to sense and to respond to fault conditions in the switch.

For example, it may sense fault conditions in the switch circuit and/or sense imminent failure or overheating of the electromechanical relay; in response thereto, it may sound an alarm and/or maintain the switch in a non-conductive state, for example. This function may also be controlled by the (or a) microprocessor.

The invention extends to an electrical installation or apparatus comprising a hybrid relay according to the first aspect or a switch according to the second aspect.

Embodiments of the invention will now be described with reference to the accompanying drawings, in which: Figures 1 and 2 show the respective time/current characteristics of type 2 and type 3 circuit breakers according to BS 3871 Part 1, as described above; Figure 3 shows embodiments of the invention according to the first aspect, mainly in block diagram form; Figure 4 shows a more detailed embodiment of the invention according to the first aspect; Figure 5 illustrates typical waveforms at three points in the circuit of Figure 4, when the switch is turned on.

Figure 6 shows two examples of how a hybrid relay such as that of Figure 3 or Figure 4 or Figure 7 may be incorporated into a circuit; and Figure 7 shows an embodiment of the invention according to the second aspect in block diagram form.

In Figure 3, a hybrid relay 30 is shown comprising a triac 32 and an electromechanical relay with switched contacts 36. The main terminals of the triac 32 are coupled in series with a resistor 34 between an ac power supply line 40 and a terminal 44 for connection to one side of a load 46. Also coupled between line 40 and terminal 44 are the switched contacts 36 of a relay operated by energisation of a coil 37. A circuit breaker 48 would normally be included in the power line, but such could form part of the switch.

The hybrid relay has a command input 50 which is coupled to the gate of triac 32 and the relay coil 37. While it might be possible in certain circumstances to have a direct coupling, it is generally necessary to match the command input to at least one of the triac and relay, insofar as they commonly require different inputs for operation. As already mentioned, it is also often desirable to isolate triac 32 from the command input 50.

Accordingly at least one of circuits 52, 54 will be provided to receive and condition the command input and provide appropriate output signals 53, 55 for operating the triac and relay. In the first aspect of the invention, there is no control over the timing of signals 53, 55 relative to the ac supply voltage waveform. Thus, although different in form, these outputs may be arranged to commence simultaneously, or one of the circuits 52, 54 could provide a delay relative to the other provided that it is ensured that the triac 32 always conducts before the contacts 36 are closed.

A dc power supply unit 60 may be connected across line 40 and a neutral line 42 to provide power for circuits 52, 54 and also for optional circuit 62 to be described later in connection with a development of the basic circuit.

The value of resistor 34 needs to be selected such that it is small enough not to produce too large a voltage across contacts 36 as they close, even if the ac voltage is then at a peak, and preferably such that if the load 46 presents a short circuit the circuit breaker 48 is rapidly tripped.

Resistor 34 also needs to be capable of handling the power therein before the contacts have closed or the breaker has tripped. Nevertheless, its value must not be so small as to fail to limit the inrush current to a degree which the triac is capable of handling (generally only for the period prior to closure of contacts 36, and thus generally a reflection of the short term maximum (peak) current rating of the triac).

Ideally, the continuous current rating of relay contacts 36, the load 46 and the breaker 48 are all suitably matched. This should mean that if the load requires a (maximum) continuous current Ic, the breaker trips at a slightly larger current, and the relay is rated at a slightly higher current still, although in practice the differences should be greater to take account of tolerances, ageing etc.

As mentioned above, both type 2 and type 3 breakers can be reliably be expected to operate very rapidly with an applied current of at least 12 times their continuous rated current handling ability. For rapid operation of the breaker with a short circuit load, this leads to a value for resistor 34 equal to at most one twelfth of the value of the maximum load impedance which the hybrid relay 30 is intended to serve. Given a supply voltage of 240 volts, it also leads to a value for resistor 34 of 240/(12.Itrip) = 20/Itrip, where 1trip is the minimum continuous current which trips the breaker 48 (or, equally, 20/iris, where 1rly is the maximum continuous current rating of the relay contacts 36) Thus far, the circuit of Figure 3 conforms to the first aspect of the invention, and there is no predetermined time relative to the ac supply at which the relay switch contacts 36 and/or triac 32 are energised.

The latter feature appears in developments of the circuit, and in Figure 3 there is provided a zero volt detector 62, which is connected between line 40 and line 42 (or alternatively across the relay at point 41 and point 43), and one or both circuits 52, 54 are adapted to receive the output therefrom and condition the timing of the outputs 53 and 55.

In one arrangement, circuit 52 is adapted such that the output 53 commences at, or close to, a zero crossing point, whereby the triac 32 turns ON with at most only a few v

Alternatively, or additionally, circuit 54 is adapted such that the relay coil 37 is energised at a time whereby the relay contacts close when the ac voltage is close to zero.

In this case, because the delay between coil energisation and contact closure is variable, depending upon the particular relay and its age, a precise contact closure time cannot be guaranteed; nevertheless, it should be possible to arrange circuit 54 to alldw for a typical delay between energisation and contact closure, whereby there is some assurance that contact closure will not occur at an ac voltage peak. Because of the undesirable effect of contact bounce, and because there is still some voltage produced across the resistor and relay switch contacts when the triac is conducting, it is preferable to arrange that the relay switch contacts initially close somewhat before a zero crossing, in order that the bounce is accommodated in a period including zero volts. This is more important when only circuit 54 is adapted for timing, as the triac 32 and resistor 34 will be required to carry a significant inrush current, albeit one which is limited by the provision of the resistor 34. In such case, the ratings of triac and resistor must also be correspondingly great.

When both circuits 52, 54 perform timing functions, the presence of circuit 52 may permit a lower rated triac and resistor to be used, the presence of circuit 54 may allow lower rated relay switched contacts to be used, transient production may be reduced, a type 2 or type 3 circuit breaker may be employed in place of a slower acting device, and requirements for the associated power wiring may be lowered.

In tests of the basic circuit, using a triac 32 with a maximum continuous current rating of 25 amps and a short term (10 milliseconds) peak current rating of 240 amps, it was found that a one ohm wirewound resistor 34 rated at 50 watts performed satisfactorily under short circuit load conditions. It restricted the current sufficiently to protect the triac prior to closure of the relay contacts while providing sufficient current to cause very rapid tripping of a 16 or 20 amp type 2 or type 3 circuit breaker (however, this might not be the case if a slower acting beaker was used, in which case the dissipation in the triac and resistor might destroy these components).

In Figure 4, an ac control signal S at supply line frequency and phase is applied between terminals 2 and 4 of a bridge BR2, for full wave rectification. Terminal 2 of the bridge is connected to the neutral line N of the supply. Output terminal 3 of the bridge is coupled to a line B and output terminal 1 is coupled via series connected resistors R3, R4 and a diode D1 to a line A. A capacitor C7 is connected across lines A and B begins to be charged upon commencement of the ac signal. Plot CH1 of Figure 5 illustrates the voltage waveform at output terminal 1 of BR2.

A 56 volt zener diode ZD4 and a resistor R14 are connected in series between the lines A and B, with the base of a transistor T4 coupled to their junction. The collector of transistor T4 is coupled by a resistor R15 to line A, and its emitter is coupled to line B. While the voltage V, between lines A and B is less than 56 volts, no current flows through the zener diode ZD4, the transistor T4 remains OFF. The base of a transistor T5 is coupled to the junction of resistor R15 and the collector of transistor T4, and its emitter is connected to line B, so that transistor T5 is turned on by current through R15 as V, rises above its VBE and the voltage at its collector (point C) assumes a value close to line B. However as soon as VAB is sufficient to cause conduction through ZD4 and the baseemitter path of transistor T4, the latter turns ON, causing T5 to turn OFF. Only when this happens can the voltage at point C assume a value which differs significantly from line B.

A relay RL2 and parallel catching diode D2 are connected in series with a silicon controlled rectifier SCR1 between lines A and B. The gate of SCR1 is coupled through a resistor R8 to the point C so that while transistor T5 remains ON, SCR1 remains OFF. A capacitor C3 prevents false triggering of SCR1 by stray high frequency voltage spikes. Circuit values are arranged so that transistor T5 is turned ON before the voltage at point C has risen sufficiently to turn SCR1 ON.

The output of bridge BR2 is also coupled to line B through a resistor R5, a 5.6 volt zener diode ZD2 and a resistor R6, all in series. The diode D1 prevents these components from being directly affected by the rising voltage on capacitor C7. The base of a transistor T1 is coupled to the junction of ZD2 and R6, its emitter is coupled to line B and its collector is connected to point C. A collector load resistor R7 is provided between point C and line A.

The full wave rectified voltage at the output of the bridge BR2 is sufficient for most of its cycle to ensure that conduction occurs through zener diode ZD2, and that the transistor T1 is ON. However, during a short period encompassing the zero point of the waveform from the bridge, the zener diode ZD2 ceases to conduct, and transistor T1 turns OFF.

When the voltage VZ across capacitor C7 is about 56 volts, the zener diode ZD4 commences to conduct, turning transistor T4 ON, and thus causing transistor T5 to turn OFF. The voltage at point C is then controlled by transistor T1, and in this condition a series of short pulses could be generated corresponding to the short periods when T1 is OFF, i.e. at and close to the zero crossing points of the ac voltage applied to bridge BR2.

However, upon energisation of the relay coil (see below), the voltage on line A falls below 56 volts, T5 turns on again, and voltage at the point C is again brought close to line B. Therefore it is common for a single pulse only to be produced at point C, although it might be possible to generate more than one such pulse under certain conditions (e.g. dependent upon the rate of discharge of capacitor C7 and the rate of charge from the bridge BR2).

This pulse is sufficient to trigger SCR1 thereby to energise the relay RL2. However although triggering of SCR1 occurs at or close to a zero crossing point, it is well known that the closing of the relay contacts RLA will occur some time later, as a result of the electromechanical properties of the relay, and there is no guarantee that closure of the contacts will coincide with a zero crossing point. This is not a factor which can be readily dealt with by incorporating a fixed time delay or advance into the circuit, since it is to be expected that no two relays will require precisely the same timing correction.

The electromechanical relay is preferably such that it will have operated within one half cycle of the ac supply voltage waveform.

Also coupled between the line A and SCR1 are series connected resistors R9 and R10, the base of a transistor T2 being coupled to the junction between these resistors. The emitter of T2 is coupled to line A and its collector coupled to line B via series connected resistor R11, a 39 volt zener diode ZD3 and a resistor R12. A transistor T3 has its base coupled to the junction of ZD3 and R12, its emitter is coupled to line B, and its collector is coupled to an input K of an opto-coupler OPC (for example of the type CNX21 manufactured by Phillips), the other input A thereof being coupled to line A through a resistor R13.

Triggering of SCR1 has an initial effect of turning T2, T3 and the light emitting diode and phototransistor of the opto-coupler OPC ON.

Circuit values are chosen such that once SCR1 has been turned ON, and current commences to flow through the winding of relay RL2, the voltage V starts to fall from the 56 volt level to a new level at which current from the bridge balances the current flowing from line A to line B.

This level needs to be sufficient that a voltage at or greater than the holding voltage is imposed across the winding of relay RL2. However it is also such that the zener diode ZD3 ceases to conduct, and transistor T3 and the optocoupler OPC are turned OFF. Optocoupler OPC is therefore turned ON for a short period only, but sufficient to allow the relay contacts RLA to be closed, for example one and a half to two full cycles of the ac waveform. Plot CH2 of Figure 5 shows how the collector voltage at T3 falls to a low value during this period, and plot CH3 shows the anode voltage of the SCR, indicating that it continues to conduct after the triac has been turned off, to enable a sufficient holding current to pass through the winding of the relay.

The contacts RLA of relay RL2 are connected between a line L at line voltage and a switched line SL, in parallel with a series combination of a capacitor C4 and a resistor R18.

There is a further parallel connection consisting of a triac U1 and a low value resistor R24. The latter provides an impedance which prevents excess heat generation in the triac while it is turned ON with a short circuit load.

A point F is coupled to the line L by a parallel combination of a 15 volt zener diode ZD5, a resistor R21 and a capacitor C5. Diodes D4 and D5 are coupled between point F and line L, and a parallel combination of a voltage dropper capacitor C6 and a resistor R22 (for permitting discharge of the capacitor) are coupled between the junction of the diodes D4, D5 and a resistor R23. The latter is coupled to the line L via a voltage dependent resistor VDR2. The neutral line N also contains a fuse F2.

These components provide a 15V dc bias between line L and the point F, and provide a protective function, for example against very high voltage spikes of high energy tending to damage the electronic components.

One output (collector) terminal of optocoupler OPC is coupled through series connected resistors R19, R20 to line voltage, and the other (emitter) terminal is coupled to the point F. The base of a transistor T6 is coupled to the junction of resistors Rl9 and R20, its collector is coupled to the point F, and its emitter is coupled in series with resistors R18 and R17 to the line L. The gate of triac U1 is coupled to the junction of resistors R17 and R18.

The use of an optocoupler provides one way of transferring the signal from transistor T3 to the triac U1 despite the voltage shift on passing from that part of the circuit between lines A and B to the part of the circuit containing the triac. However, it should be evident to the skilled reader that any practical way of transferring the signal could be used. The particular optocoupler employed contains a phototransistor, but other photosensitive devices, e.g. a photodiode, phototriac, or optothyristor could be used. An alternative form of signal coupling element providing electrical isolation is a transformer, e.g. a pulse transformer.

When the optocoupler OPC is turned ON, at or close to a zero crossing point of the ac waveform, the transistor T6 is also turned on, enabling the triac U1 to conduct. There is no significant lag as in the case of the relay RL2, so that the switched line can assume line voltage substantially immediately. Shortly thereafter, the relay contacts RLA will close, and shortly after that the optocoupler turns OFF. In turn, transistor T6 turns OFF and the triac U1 ceases to be conductive at the next zero crossing point.

The use of a triac in the short term has the advantage that the switching time can be well controlled. By switching at close to the zero crossing point, no large inrush current occurs. Because the electromechanical relay is subsequently switched on in parallel with the triac it does not have to cope with a large inrush current, and consequently the requirements therefor are reduced, enabling a cheaper and/or smaller relay, capable of coping with the steady state current to be employed.

The use of the electromechanical relay in the long term provides a more efficient electrical conduction path to the load, and the turning OFF of the triac within a short period saves dissipation in the triac and prevents losses through its gate and through the path between its main terminals T1 and T2. Because of this and the switching at zero crossing the triac itself can be a relatively small and/or low energy component.

The ac control signal S can be derived by any suitable means, including the use of direct switching of the supply, and indirect switching, as by the use of relays and remote controls, e.g. infra-red/ultrasonic transmitters. As shown in Figure 4, A relay RL1 forms part of a two-way switching system. The relay is energised in response to a signal CL via a series connection of capacitor C1, resistor R1 and full wave bridge BR1, all protected by a voltage dependent resistor VDR1 in parallel. The output of the bridge is smoothed and controlled by a capacitor C2 and a zener diode ZD1, and applied to the relay RL1. The signal CL may be derived by operation of a switch acting on the supply voltage, and a fuse F1 is inserted in the line carrying the signal CL.

The contacts RLB of the relay RL1 form part of a two-way switching circuit, two examples of which are shown in Figure 6. The circuit shown diagrammatically in Figure 6 comprises upper and lower bus bar systems, each comprising earth (E), neutral (N), line (L1 or L2), switched line (SL), and control signal (2W) conductors. The switched line conductors are broken by switches, but the others are continuous. Loads, such as banks of fluorescent lights, are associated with separate segments of the conductors.

The left hand side of the Figure illustrates two way switching using the same set of bus bars, while two way switching between the two sets of bus bars is shown on the right. It will be seen that operation of a single pole switch serves to connect a line L1 (or L2) to the signal bus 2W. Thereupon a relay, such as RL1 in Figure 4, is actuated. The contacts of the relay form one half of a two way switching circuit, with a manually operated switch providing the other half.

Each of the bus bar systems shown in Figure 6 could be (but is not necessarily) part of a control and connection system of the type described in our copending British Patent Application No. 9403609.2 (serial number GB 2275835 A), which itself is a development of the system described in our British Patent Application No. 9110813.4 (serial number GB 2255865 A), and reference should be made to these applications. Furthermore, although the relays illustrated form part of a switch according to the first aspect of the invention, it should be understood that a similar relay which forms part of a switch according to the second aspect of the invention (an embodiment of which is described in more detail below) could be used in like manner.

Accordingly the invention extends to an electric control and connection system comprising line, neutral and earth electrical conductors, wherein at least two additional conductors are provided for control purposes and wherein at least one of the additional conductors is provided with switch means at one or more outlet points, such switch means being operated by the insertion of one of at least two possible different styles of connector whilst at least one other style of connector leaves the switch mechanism unoperated, wherein at least one connector comprises or is coupled to a switch according to the first or second aspect of the invention.

However, it should be noted that a switch according to the invention may be incorporated in other types of switching circuit, and that means for switching'other than the manual switches illustrated in Figure 6 may be used, as indicated above.

The switch of the invention provides advantages when used in the switching of reactive loads, as has been explained above. However, it should be noted that its use is not restricted thereto. It can be used in most cases where it is desired to energise a load from an ac supply, whether or not the load is reactive, for example in circuit breakers and in remotely operated switches which do not have to interrupt fault level currents. In this context it should be noted that even purely resistive loads can give rise to arcing during making or breaking of the relay contacts, and also that even when the load is purely resistive, the associated wiring may provide a certain amount of reactance.

Furthermore the capability to switch on reliably at zero ac supply volts, particularly if combined with a capability of switching off at an optimum time, is advantageous with any form of load, in the sense that there is minimal disturbance to the ac supply.

In Figure 7, an electrical switch 70 is shown, comprising switched contacts 72 which are operated by an electromagnetic coil 74, a low voltage power supply 76, a zero volt detector circuit 77, and a micro-processor 78.

An ac power supply is connected across a line terminal L and a neutral terminal N. The switched contacts 72 are arranged such that they may be used to control a load connected to an output terminal 82. The low voltage power supply 76 supplies power to the zero volt detector circuit 77, and the micro-processor 78.

When a command signal is applied to terminal 80, which is connected to the micro-processor 78, the micro-processor 78 is arranged to operate the switched contacts 72 by energising the coil 74 at a particular time delay after a zero-voltage. The micro-processor is arranged to measure the operating time of the switch, by means of the feedback connections 86 and 87. The operating time is measured from the instant when the operation of the switch contacts by energising the coil 74 was initiated to the instant that the switch contacts 72 close. The micro-processor is arranged to use this information on the operating time of the switch so that the next time a control signal is applied to terminal 80 the particular time delay is such that the switch contacts close when the voltage across them is at or nearly at zero. This ensures that as the operating time of the relay varies with wear arising from normal usage, that the switch contacts always close at as near the optimum point on the wave as possible.

In a similar manner the micro-processor is also arranged to measure the switch-off time, which is taken from the instant when the coil 74 is de-energised to the instant that the switch contacts 72 open. The micro-processor is arranged to use this information on the switch-off time of the switch so that the next time the control signal is removed from terminal 80 the switch-off time delay is such that the switch contacts open when the voltage across them is at or nearly at zero. This provides the benefit that the switch contacts always open at as near the optimum point on the wave as possible.

A further benefit of this arrangement is that not only does this ensure that any changes in the operating characteristics of the relay over time are automatically compensated for, but also there is no requirement to adjust values of components on the assembly line, or require tight tolerances on the timing characteristics of the switch contacts. The micro-processor preferably has a non-volatile means of remembering the operating time and the switch-off time of the switch contacts, so that it may be automatically set on the assembly line when it is tested for operation. Preferably this first operation is conducted with a high value resistive load, as the first operation of the switch is unlikely to be at a current zero, and hence there is a risk of damage to the switch contacts.

Preferably the speed of operation of the switch contacts is such that the micro-processor may have limits set on the permissible values for the operating and switch-off times.

Should the micro-processor not be able to achieve switching within the desired time limits at the predetermined voltage at or nearly at zero, then the micro-processor will cease to attempt to operate the switch contacts and will signal a fault condition by means of an indicator lamp 84 or other suitable means. Preferably the micro-processor will ensure that the switch contacts remain in the open state once it has signalled a fault condition. Preferably the micro-processor will also be arranged to monitor other parameters of the switch, such as the voltage drop across the switch contacts 72 measured between the feedback connections 86 and 87. Should this voltage drop rise above a predetermined threshold when the contacts are closed, this will indicate that excessive heat is being generated within the relay, hence the micro-processor will be arranged to open the contacts and preferably signal a fault condition.

Claims (38)

1. A hybrid relay for coupling a power supply to a load, comprising an electromechanical relay having at least one pair of switched contacts for coupling to the supply and the load respectively, a series combination of a semiconductor switch and a resistor and/or inductor being coupled in parallel across said pair of switched contacts, a relay control input of the electromechanical relay and a switch control input of the semiconductor switch being coupled to a command input of the hybrid relay to receive respective energisation signals such that the semiconductor switch closes before the switched contacts.

2. A hybrid relay according to claim 1 including a first conditioning circuit coupled to receive said command input and having an output coupled to said switch control input for energising said switch.

3. A hybrid relay according to claim 2 wherein said first conditioning circuit includes a coupling providing electrical isolation between the command input and said semiconductor switch.

4. A hybrid relay according to claim 2 or claim 3 for use with an ac power supply, wherein the timing of the output of said first conditioning circuit is arranged to be responsive to the timing of the ac voltage waveform.

5. A hybrid relay according to claim 4 wherein said first conditioning circuit includes a detector for detecting zero crossings of the ac voltage waveform.

6. A hybrid relay according to claim 4 or claim 5 wherein the output of said first conditioning circuit is arranged to energise the semiconductor switch only within a predetermined range of time encompassing a zero crossing of the ac voltage waveform.

7. A hybrid relay according to claim 4 or claim 5 wherein the output of said first conditioning circuit is arranged to energise the semiconductor switch voltage waveform substantially at a zero crossing of the ac voltage waveform.

8. A hybrid relay according to any one of claims 2 to 7 wherein said first conditioning circuit is arranged to turn said semiconductor switch off after a predetermined period which is sufficiently long to ensure that the switched contacts have closed.

9. A hybrid relay according to claim 8 wherein said predetermined period is no more than 2 ac cycles.

10. A hybrid relay according to claim 8 or claim 9 arranged to respond to a second state of the command signal to de-energise the electromechanical relay, said first conditioning circuit being arranged to turn said semiconductor switch on before the switched contacts have opened and to turn it off again after the contacts have opened.

11. A hybrid relay according to any preceding claim including a second conditioning circuit coupled to receive said command input and having an output coupled to said relay control input for energising said electromechanical relay.

12. A hybrid relay according to any preceding claim for use with an ac power supply wherein the timing of the output of said second conditioning circuit is arranged to be responsive to the timing of the ac voltage waveform.

13. A hybrid relay according to claim 12 wherein said second conditioning circuit includes a detector for detecting zero crossings of the ac voltage waveform.

14. A hybrid relay according to claim 12 or claim 13 wherein the output of said second conditioning circuit is arranged to commence energisation of the electromechanical relay only within a predetermined range of time relative to a zero crossing of the ac voltage waveform.

15. A hybrid relay according to claim 12 or claim 13 wherein the output of said second conditioning circuit is arranged to commence energisation of the electromechanical relay such that the switched relay contacts first close shortly before or substantially at a zero crossing of the ac voltage waveform.

16. A hybrid relay according to claim 12 or claim 13 wherein upon closure the switched contacts bounce over a bounce period substantially less than half a cycle of the ac waveform, and the output of said second conditioning circuit is arranged to commence energisation of the electromechanical relay such that a zero crossing of the ac voltage waveform falls within the bounce period.

17. A hybrid relay according to any preceding claim wherein said electromechanical relay is such that the switched contacts will first close within one ac cycle from the commencement of energisation of the electromechanical relay.

18. A hybrid relay according to claim 17 wherein said electromechanical relay is such that the switched contacts will first close within half of cne ac cycle from the commencement of energisation of the electromechanical relay.

19. A hybrid relay according to any preceding claim wherein said switched contacts have a maximum continuous current rating of 1rly amps and said resistor has a value R, where R is less that or equal to 20/idly ohms.

20. A hybrid relay according to any preceding claim wherein said semiconductor switch is a triac or thyristor.

21. A hybrid relay according to any preceding claim wherein said resistor coacts with said semiconductor switch to limit the current through the semiconductor switch to a value no greater than the maximum short term (peak) current rating of the semiconductor switch when the series combination is subject to 240 volts ac.

22. A hybrid relay according to any preceding claim wherein said electromechanical relay is an electromagnetic relay.

23. A hybrid relay according to any one of claims 2 to 22 adapted to operate with an ac command signal, and including means for converting said command signal to a dc power supply for energising said conditioning circuit(s).

24. A hybrid relay substantially as hereinbefore described with reference to Figure 3 or Figure 4 of the accompanying drawings.

25. A power supply circuit comprising a source of ac voltage, a hybrid relay according to any preceding claim, and a circuit breaker in series between said source and one said switched contact, wherein said circuit breaker has a maximum continuous current rating 1trip amps, and said resistor has a value less than or equal to 20/Itrip ohms.

26. An electrical switch for altering the coupling between an ac supply to a load in response to a command signal, comprising an electromechanical relay having at least one pair of switched contacts for coupling to the supply and the load respectively, and a control circuit, wherein said control circuit comprises first control means for altering the state of energisation of said relay at a first time in response to a change in said command signal, said first time being controlled relative to the ac supply cycle, second control means for detecting the instant when the switched relay contacts have changed state relative to the ac supply cycle, and third control means for adaptively altering said controlled time if said instant lies outside a predetermined range, such as to bring said instant within, or closer to, said predetermined range, during a subsequent operation of said relay.

27. A switch according to claim 26 wherein a said change in said control signal causes the load and supply to be coupled by closure of the switched contacts, an associated said controlled time is a contact closure time, and an associated said predetermined range is a predetermined closure range.

28. A switch according to claim 27 wherein said predetermined closure range includes the zero crossing of the ac supply voltage.

29. A switch according to any one of claims 26 to 28 wherein a said change in said command signal causes the load and supply to be decoupled by opening of the relay contacts, an associated said controlled time is a contact opening time, and an associated said predetermined range is a predetermined opening range.

30. A switch according to claim 29 wherein said predetermined opening range includes the zero crossing of the ac supply voltage.

31. A switch according to claim 29, wherein said control circuit also comprises fourth control means arranged to calculate an optimum period relative to the supply voltage cycle at which the energy in the circuit is at or close to a minimum, for use as said predetermined opening range.

32. A switch according to any one of claims 29 to 31 as dependent on claim 27 or claim 28, wherein said predetermined opening and closure ranges are the same or different.

33. A switch according to any one of claims 26 to 32 wherein said control circuit contains fifth control means arranged to sense and to respond to fault conditions in the switch.

34. A switch according to any one of claims 26 to 33 wherein said control circuit comprises a microprocessor which serves to function as some or all of said control means.

35. A switch substantially as herein described with reference to Figure 3, or Figure 4, or Figure 7, of the accompanying drawings.

36. An electrical installation or apparatus comprising a hybrid relay according to any one of claims 1 to 24, or a switch according to any one of claims 26 to 35.

37. An electrical installation according to claim 36 and substantially as herein described with reference to Figure 5 of the accompanying drawings.

38. An electric control and connection system comprising line, neutral and earth electrical conductors, wherein at least two additional conductors are provided for control purposes and wherein at least one of the additional conductors is provided with switch means at one or more outlet points, such switch means being operated by the insertion of one of at least two possible different styles of connector whilst at least one other style of connector leaves the switch mechanism unoperated, wherein at least one connector comprises or is coupled to a hybrid relay according to any one of claims 1 to 24, or a switch according to any one of claims 26 to 35.