PHASE CONTROL METHOD AND DEVICE FIELD OF INVENTION The present invention relates to the field of phase control, a field where there is a device in which an inductor is used to control the risetime of a waveform. The principles of the invention have many different applications other than phase control, but all related to the switching of power, voltage or current from a supply to a load. In one application, the present invention relates to a dimmer used in conjunction with loads such as a light or lighting systems. Other applications of the present invention relate to heating controls and motor speed controllers, including for fans, drills etc. The present invention, however, should not be limited to only such applications.

BACKGROUND ART Phase Control Principles, a general outline Phase Control (PC) is a technique used for controlling the transfer of electrical power between an electrical power source and an appropriate load or loads. PC is a widely used technique with many applications, including in lighting dimmers, heating controls and speed controllers. PC was originally implemented using Thyristors (of which SCRs and Triacs are two examples). However, PC is not constrained to using these devices and, in fact, virtually any switching device could possibly be used. Currently GTOs (Gate Turn-Off Thyristors), IGBTs (Insulated Gate Bipolar Transistors) and MOSFETs (Metal Oxide Semiconductor Field Effect Transistors) are commonly used in Phase Control devices, including dimmers.

Phase Control is applied to alternating current (AC) power sources whereby the input waveform is usually sinusoidal as shown in Figure 1. Figure 1 shows the waveform of the current flowing in a resistive load if it is connected to such a power source.

The power transferred into a resistive load is proportional to the area contained within the waveform and which area is shown as shaded in Figure 1.

The case illustrated in Figure 1 occurs when a load is connected directly to a

power source without any means of controlling the flow of current into the load, such as where, for example, a light globe is connected directly to a power source.

In the case of Phase Control, and as illustrated schematically in Figure 2A, a switching device 3 is inserted between the power source 1 and the load 2. The purpose of this switching device is to interrupt the flow of current into the load. By interrupting the flow of current into the load, the power transferred to the load is controlled.

In some applications, this switching device is selected to be able to switch at high speed, usually at twice the frequency of the AC power source. If the operating frequency of the switch is sufficiently high, then the frequent interruption to current flow and power delivery is usually not evident in the load or to the user.

Figure 2B shows the AC current waveform resulting when Phase Control is applied to a resistive load as shown in Figure 2A. In Figure 2B, the switching device has been arranged to switch"on"and"off'twice for each cycle of the AC power source so that is effectively"on"for approximately half the time. The amount of power that is transferred to the load is shown by the shaded area, and represents about half the time period. It can also be seen that the area enclosed by the waveform has been halved and thus the power delivered to the load has been halved also.

If the point where the switch turns on is varied, then the resulting power delivery to the load will also vary.

It can be seen from Figures 2A and 2B that Phase Control is similar to Pulse Width Modulation (PWM) whereby a switch is used to modulate the flow of power into a load and where the power delivered to the load is roughly proportional to the duty cycle of the switch. Phase Control is distinguished from PWM because it is associated with AC power sources and the switching takes place synchronously with that AC power source and usually at twice that frequency. Whereas, PWM switching usually takes place at much higher frequencies.

Some Phase Control Problems Phase Control is popular because it is relatively efficient in operation, relatively inexpensive to implement and because the technology is relatively simple and well understood. Nevertheless, Phase Control has several disadvantages and a number

of techniques have been developed over the years directed at solving these disadvantages, all with what is considered to be a limited success.

One principle disadvantage of Phase Control is evidenced by the high speed rising edge 4 (as shown in Figure 2B) of the current waveform when the switching device turns"on". Many of the Phase Control prior art attempts, which have emerged over the years, have been aimed at minimising this disadvantage. In the art, a Reverse Phase Control functions in a reverse manner to that just described, in that the switch device is arranged to turn"on"at the zero crossing point and then to switch "off'some variable time later according to output power requirements. In effect, this is the reverse of Figure 2B, where the rise time follows the sinusoidal wave, but the fall time is abrupt.

Referring to Figure 2B, it should be noticed that at the point of switch"on", the waveform is essentially vertical. It has been found that this steep rising edge 4 indicates that the output of such a device contains substantial energy in the form of harmonics of the fundamental source frequency. This characteristic of a Phase Control device is referred to a"Rise Time". These harmonics are considered detrimental to many loads for which a Phase Control device might be used to control. Since most of the usual loads which are used with Phase Control devices, such as lamps, transformers and motors, are designed for operation at normal power line frequencies (50 & 60 Hz.), they are often not compatible with the high frequency harmonics contained in the PC device output.

Some common problems that result from the application of phase control to incompatible loads are overheating, mechanical noise and interference with other sensitive electrical and electronic equipment. Acoustic noise can also result from the audible vibration of components that are insensitive to line frequencies but which may be readily excited by the higher frequencies contained in the output of a Phase Control device. For example, the filaments of many types of incandescent lamps are known to produce substantial noise when driven by Phase Control dimmers.

Overheating most commonly results when Phase Control devices drive loads that contain transformers. Heat is generated because the magnetic material

is often not suitable for high frequency operation. Another related problem is associated with possible asymmetry in the switching leading to a DC component in the output that can also be injurious to transformers.

Interference with other services can result because the high frequencies contained in the unfiltered Phase Control output are much more likely to be radiated than the relatively low line frequencies.

In order to address these problems, ideally the Phase Control devices would have a filter attached between the switching device and the load to eliminate all but the lowest frequency components from the output. Other means have also emerged including"Reverse Phase Control"and active rise time control, which uses linear (progressive) turn"on"of the switching device to control the rate of rise of the current in the load. This has been found to produce substantial heat in the switching device, and is thus considered undesirable. It is considered that, in such cases, they are essentially aimed at increasing the turn "on"rise time of the load current waveform.

Often the filter takes the form of an inductor located in series with the load.

The action of the inductor is to slow the rate of rise of current in the load at the point of switch"on". In this way the output harmonics are reduced. Figure 3 illustrates the load current waveform for a Phase Control device when such a filter is used. It can be seen that the rise time has been increased, that is the'rise'of the waveform is more curved, not so vertical.

In the professional lighting control industry, which includes theatres for example, it is usual to include a substantial inductive filter in dimmers because of the quantity and size of loads designed for this market. These filters are aimed at reducing harmonics and thus acoustic noise and some electrical interference with other systems, such as audio systems. Referring to Figure 4, it has become commonplace in the art to provide filters that produce current rise times between 100 and 800 microseconds. Rise times in excess of 1 millisecond have been used in extreme cases. The example of Figure 4 shows a risetime of around 700uSec when applied to a 50Hz Phase Control system.

Whilst a relatively simple inductive filter is reasonably effective there are serious trade offs which must be considered when designing Phase Control devices with such a filter. Unfortunately, as filter performance (rise time) is increased, device cost and efficiency are seriously impaired. Efficiency is reduced because energy is lost in the filter/choke in the form of heat.

Figure 5 shows a typical Phase Control circuit comprising a switch, a choke, a control circuit and a load. Figure 6 shows the output waveform of the circuit of Figure 5 and the control signal used to turn"on"and"off"the switch. In operation, the control circuit produces a control signal ("X") which is applied to the switch. This signal causes the switch to turn"on"at time"A" (Figure 6). The switch turns"off"at time"B". The control circuit determines time"A"relative to the "Zero"points (Figure 6) according to the required power output,"R". The higher the required power output of the circuit, the greater the"on"time of the switch (period"A"to"B").

One of the problems of this circuit arrangement can be appreciated by reference to Figure 6. The purpose of the choke is to increase the rise time of the output at the point of switch-on ("A"). In the example shown, the rise time is extended to about half of one division whereas in Figure 2 it will be remembered that this time is essentially zero. The problem is that for the balance of the"on" time of the switch ("A"-"B") the choke is performing no useful functions and is consuming energy.

Phase Control Filter Losses In a Phase Control dimmer, the major losses take place in the switching device and in the inductive filter (choke). Total choke losses result from both copper losses (which are the result of resistive heating in the windings of the choke and which are proportional to the square of the current), and core losses. which are considered the major power losses in a phase control dimmer, often exceeding the losses in the active switching device.

Core losses are a result of an alternating field in a core material. The loss generated for a given material is a function of operating frequency and total flux swing (DB), The core losses are due to hysteresis, eddy current and residual

losses in the core material. Eddy current loss will be the dominant loss at higher frequencies while hysteresis loss will be the dominant loss at lower frequencies.

For example, in a 2.4KW (assuming 240VAC mains supply) dimmer, the"on" state voltage drop of a power switch is dependent on package size selected and is typically 1.6 volts with a load current of about ten amps (RMS). The power loss of the power switch, at full load, is approximately 16 watts. Further, if the maximum efficiency of the dimmer was 98%, then the total losses would be 48 watts. Assuming the power loss in the other components of the dimmer to be negligible, it follows that the loss in the choke would be about 32 watts, which is higher than the loss in a power switch. Increasing the size of the wire used to make a choke can reduce power loss in the choke but this leads to increased size and cost. Another problem associated with chokes is that the choke becomes less useful but contributes more power loss when output level is close to 100% output.

SUMMARY OF INVENTION "APC"is a name given to the present invention, a new means of implementing Phase Control: Advanced Phase Control.

An object of the present invention is to alleviate the trade off between, on the one hand, filter performance and, on the other hand, costs and efficiency.

The present invention seeks to enable a Phase Control device to have high filtering performance realised with improved efficiency and lower cost.

The present invention provides, in a device having at least a circuit arrangement including at least a first switch means and filter means coupled between a source of power and a load in order to provide control of power transferred from the source to the load for a first period of time, an improvement including a second switch means coupled in parallel with the circuit arrangement and being adapted to provide control of power transferred from the source to the load for a second period of time.

The present invention also provides a Phase Control, or dimmer control device adapted to provide dimmer control of at least one light source, the device having a first switch means coupled to a filter means, forming a first circuit arrangement for coupling between a source of power and a load, and a second

switch means coupled across the first circuit, the first circuit and the second switch being adapted to provide control of the at least one light source for a first period of time and a second period of time, respectively.

The present invention further provides a method of providing control of the transfer of power from a source of power to a load, the method including : a) enabling a first path to provide power from the source to the load, b) enabling a second path to provide power from the source to the load.

Preferably, the first path is enabled for a first period of time, the second path is enabled for a second period of time, and/or the first and second path are enabled for periods of time which overlap. Preferably, the first path includes a first switch and a filter means, and the second path includes a second switch.

Furthermore, preferably, the method is used in conjunction with a phase control device.

In essence, the present invention is based on using a first circuit arrangement having a first switch and filter to initially turn'on'the power to the load, and within a relatively short period of time thereafter, turning'on'a second switch, coupled in parallel to the first circuit, thus also providing power to the load.

In this way, the filter inefficiencies normally associated with the prior art due to all the power flowing through the filter for a relatively long period of time, are bypassed by enabling most of the power flowing through the second switch, which has substantially smaller or minimal losses associated with it.

Although the present invention will be described with reference to a lighting control dimmer, many applications of the present invention are possible, whether to heavier loads or smaller, domestic load or industrial situations. It is considered that the present invention can be applied to virtually any Phase Control based device, including variants like Reverse Phase Control and asymmetric PC schemes. Equally, the present invention should not be limited to sinusoidal waveforms: for example triangle or square waves, or in fact any shape waveform would also work. The waveform need not be symmetrical for the invention to work either. Further, the switching need not be constrained to being symmetrical

(essentially identical"on"time for alternating negative and positive half cycles) : There are some phase-control variants where this is so.

Advanced Phase Control A new phase control strategy is shown in Figure 7. Without switch #2, this is a prior art phase control circuit. By adding switch #2 as shown and turning"on" switch #2 a certain time delay after turning"on"switch #1, the switch #2 will conduct the major portion of the current that would otherwise have flowed through switch #1 and the choke.

Figure 8 illustrates the relative timing of switching"on"and"off"switches #1 and #2. The timing for switch #1 is identical to that described above for the standard phase control case. Referring to Figures 7 & 8, signal"X"is the normal phase control signal and signal"Z"is the control signal for controlling the additional switch #2. Switch #2 is arranged to switch"on"at time"C". The delay time between"A"and"C"has been designed to be equal to, or slightly longer than, the rise time of the load current as described elsewhere.

When switch #2 turns"on", because of the lower impedance of this current path, the current flowing through switch #1 and the choke gradually reduces to near zero and switch #2 eventually carries the load current. Figure 9 shows how the current in switch #1 reduces whilst the current in switch #2 increases after time"C". This figure illustrates the principles of operation of the circuit. The actual waveforms for each embodiment of the invention would depend on the specific components selected by the designer.

Therefore, the Advanced Phase Control (APC) system of the present invention illustrated in Figure 7 will maintain the same performance (risetime) during the switching period but provides much less power loss during the stable condition after the switching period. This efficiency gain will become more pronounced as the output power approaches 100% because, due to the gradual rise of current in switch #2, switch #2 will carry proportionally more of the current in these circumstances.

The preceding circuit description and diagrams are based on a circuit constructed using Thyristors for switch #1 and switch #2. Those skilled in the art would appreciate that the application of other switching devices would result in

subtly different circuit control signal timing and operation. The invention can equally be applied to circuits comprising combinations of switching devices. For example, one might choose to use an IGBT for switch #1 and a TRIAC for switch #2.

As an alternative, it is contemplated that switch #2 is arranged to bypass the choke only. That is, referring to figure 7, switch #2 is arranged in parallel with only the choke, not switch #1 and the choke. It has been found, however, that the operation of this alternative arrangement is less efficient than the arrangement above. Firstly, when closing switch #1 the high dv/dt across the choke may (depending on the package selected) cause switch #2 to false trigger which would lead to incorrect operation. Secondly, with two switches in series the arrangement would be less efficient.

Those skilled in the art will recognise that there are several options for controlling the timing of switch SW2. Further, the choices of device for switch SW2 (thyristor, transistor etc) influences the choice of timing control.

Figure 9 illustrates Standard Control and is the timing method wherein a thyristor is employed for SW2. In this method a pulse (longer than t9'-specified by the thyristor manufacture) is generated to turn"on"SW2, a delay period after turning"on"SW1, where the delay is fixed and determined to be longer than choke rising time. The SW2 triggering pulse must be after SW1 being turned on and must be removed from SW2 before the choke voltage zero crossing.

A modification of Standard Control is to use a longer pulse whereby SW2 is turned"on"as described above and turned"off"before the choke voltage zero crossing (refer to Figure 9). This type of control is often termed"hard firing"when applied to thyristors and is also applicable to devices like transistors.

A third option is Adaptive Control. Adaptive Control can be applied to both the pulse and hard-fired methods described above. Referring to Figure 9, with this method SW2 is turned"on"after both turning"on"SW1 and when the voltage across choke has fallen to less than a predetermined value (say, a few volts). In this way the timing can be adapted to the actual rise time of the choke and the predetermined value may be selectable or preset depending on the

particular application of the invention. Adaptive Control does not necessarily need preset delay values for different chokes.

TESTING RESULTS The following test results were recorded when using a Phase Control lighting dimmer constructed according to Figure 7. The dimmer was connected to a nominal 3kW load and supplied from a 240Vac, 50Hz power supply. The switching devices (#1 & #2) were suitably rated Triacs and the choke was 4mH.

This filter is typical for low to medium performance professional dimmers and produces a rise time of around 200uSec. The"APC"switch #2 delay was set to 200uSec.

Test 1 : The dimmer was arranged to ramp the output current"up"and"down" continuously with a two minute cycle time. The test results were recorded after the temperature of the measured components had stabilised. (Note: the temperature of Switch #2 was about the same as for switch #1) Temperature(C°) With Switch #2-"APC"Without Switch #2 Ambient.26 26 Choke 56 90 Switch#1 48 68 Table 1 Temperature on Switch #1 and Choke with or without controlling Switch #2 Test 2: In this test, the dimmer was set to constant output level of nearly full (1 OAmps RMS) and the input and output power were recorded. Power Consumption With Switch #2-"APC"Without Switch #2 Input 2. 41 KW 2. 40 KW Output 2. 39 KW 2. 35 KW Power Loss 20 Watts 50 Watts Efficiency 99. 2% 97. 9%

Table 2 Power consumption with or without controlling Switch #2 Test 3: * the third test is done based on the same general conditions as above, however measurements are taken at a number of output levels. The temperature of the choke and switches has been measured for varying output power settings. The temperatures on different parts (given as the order of choke/switch #1/switch #2) are shown as Table 3: Output Level (%) Temperature (C°)--Amb : 28 C° With Switch #2-"APC"Without Switch #2 30 68/58/34 87/69/-- 40 87/67/40 90/71/-- 50 81/69/51 104/79/-- 60 78/70/62 114/88/-- 70 70/67/79 123/94/-- 80 55/55/101 127/99/-- 90 40/42/122 130/99/-- 100 30/37/124 133/98/-- (See Note)

Table 3 Temperature on Choke, Switch #1 and Switch #2 with or without controlling Switch #2 CONCLUSIONS The testing results from test1 show that the average temperature on both Switch #1 and choke are lower when"APC"technology is employed.

The testing results from test2 show that the power loss for the whole circuit (of figure 7) is 30W lower if it is with new phase control at full output level.

The results from test3 show that the higher the output levels, the cooler the temperature on the choke as a result of employing the new advanced phase control of the present invention. Note: the temperature disparity between Switch 1 and Switch 2 is because the two switches were mounted on different heatsinks in the test.

The circuit will achieve maximum efficiency (minimum power) loss if the output level is more close to 100% with controlling Switch #2. This can be seen from test 3. The temperature on the choke changes from 30°C to 133 °C with or without controlling switch #2.

The maximum power loss for the choke should happen at about 100% output level without"APC"and switch #2 and at around 40% output level with"APC". (This can be varied from case to case). With"APC"the output level for the maximum power loss in a choke is dependent on the delay time and load condition.

The advantages of"APC" (adding Switch #2) are as follows : the circuit will run much cooler especially the output level is close to 100%, the wire size of the choke can be reduced, (cross area can be reduced at least 50% for the same or less power loss) reduced weight and volume if the wire size of the choke is reduced, * reduce the total voltage drop on the full output level, * cost may be reduced (with changing choke)