BACKGROUND

[001] For driving a high-power load, power can be adjusted by switching on/off an AC power signal to generate an AC output power. The switching operation can create harmonics, which may cause fluctuations in the AC power signal. The harmonics also can interfere with additional loads connected to the AC power source, such as a mains electrical supply. Switching on/off the AC power signal additionally may create flicker, depending on the switching frequency. A device including the driven load may not pass voltage variation regulations

BRIEF DESCRIPTION OF DRAWINGS

[002] For a more complete understanding, various examples will now be described with reference to the accompanying drawings in which:

[003] Fig. 1 is a block diagram of an AC power control system according to an example;

[004] Fig. 2a and 2b are diagrams for illustrating the switching operation of a triac device which can be used in the system of fig. 1, according to an example;

[005] Fig. 3a to 3g illustrate different switching sequences according to various examples of an AC power control method;

[006] Fig. 4a and 4b illustrate a switching sequence according to another example and a comparative switching sequence;

[007] Fig. 5 is a flow diagram of a method according to one example. DETAILED DESCRIPTION

[008] Examples described herein relate to a system and a method for controlling AC power to be delivered to a load. - -

[009] Fig. 1 shows one example of a control system 10 for delivering AC power to a load 30. The system 10 comprises a zero crossings detection circuit 12, a controller 16, and a switching circuit 18. The zero crossings detection circuit 12 is connected to an AC power line 14 to detect zero crossings of a sinusoidal or approximately sinusoidal AC current and to output a zero crossing detection signal to the controller 16. The zero crossing detection circuit 12 can be a circuit or combination of circuit plus filtering software that manages the time base required for a switching decision of the controller 16.

[0010] The controller 16 can be a microcontroller, ASIC, or other control device, including control devices operating based on software, hardware, firmware or a combination thereof. It can include an integrated memory or communicate with an external memory or both. The controller 16 outputs a trigger or activation signal to the switching circuit 18. The switching circuit 18 can be a triac based switching circuit and also is connected to the AC power line 14. The switching circuit 18 may include a triac device 20 for switching the AC power based on the activation signal from the controller 16. The load 30 is connected to the output of the switching circuit 18 and may comprise a high-power load, such as a heating lamp.

[0011] Just as one example, the load 30 can be a heating device or a group of heating devices provided, for example in a printer, such as a 3D printer. For example, the load can be a lamp such as an infra-red lamp, any type of filament based heating device, a thermal based curing device or any other heating device. The load also can be another type of device, such as an alternating current motor, any other alternating current load. The load further can be a high- power load operating on an AC power supply having a frequency of 50 or 60 Hz and having power consumption in the range of 10 to 100 kW, or 10 to 50 kW, or 10 to 20 kW, to just give some examples.

[0012] The switching circuit 18 can be a triac based control circuit. Operation of a triac device 20, which can be used in the switching circuit 18 fig. 1, is illustrated with reference to fig. 2a. A triac, i.e. a triode for alternating current, is a three terminal electronic component that conducts current in either direction when triggered. A triac allows current flow in both directions and can be triggered by applying either a positive or negative voltage to its gate. Once triggered, a triac continues to conduct, even if the gate current ceases, until the main current drops below a certain level called the holding current. This usually is the case when the line voltage changes phase, at or close to a zero crossing point. Accordingly, a triac can be activated at any phase of the sinusoidal waveform but cannot be switched off until the next zero crossing. This process is illustrated in fig. 2a. A triac is a simple and reliable component for switching on/off alternating voltage, e.g. from the mains line power grid. Other switching devices can be used, including those which are switched off actively at the zero crossing point.

[0013] In certain applications, such as in some 3D printers, the power delivered to a fusing or heating lamp is to be controlled very accurately. This can be achieved by providing small adjustment steps, such as adjusting the power in steps of 10% or less of the maximum power, or by providing a continuous power adjustment. This type of control code be achieved, for example, by activating only every N ^th cycle or every N ^th semi-cycle of a sinusoidal AC power signal, or an integer multiple thereof, to obtain an adjustment resolution of 1/N. In this approach, the AC power signal is switched at the AC zero crossing level which can be implemented by simple electronics. However, at low power levels, the switching frequency will be reduced below the frequency of the AC power which can create a high level of flicker. Flicker can comprise regulatory flicker and visual flicker. Visual flicker is an issue when using heating lamps which contain a filament radiating light in the visible spectrum.

[0014] According to another approach, it would be possible to activate only a fraction 1/n, or an integer multiple thereof, of each cycle or semi-cycle to obtain an adjustment resolution of 1/n. This type of control can achieve a higher level of resolution at the expense of creating harmonics. Harmonics may be generated particularly when switching the power load at high frequency. Further, harmonics may be caused by sharp rising voltage edges created when the triac device is switched ON during one semi-cycle. Whereas switching the power load at high frequency can provide high adjustment resolution, it is recommended to use additional input filters to meet regulations concerning the suppression of harmonics and emission.

[0015] In an electrically noisy environment, which frequently is found in industrial or heavy machinery environments, the input AC power signal may contain disturbances. An example is shown in figure 2b. Such noise may lead to wrong zero crossing detection and, as a consequence, erroneous triac switching. Given the intrinsic triac operation, i.e. the functionality that a triac cannot be switched off until a next zero crossing, in the case that small output power shall be provided, i.e. turning-on the triac very close to the next zero crossing, this noise can produce an additional effect: if the actual switching happens just after the actual zero crossing, due to a wrong zero crossing detection, the triac will be enabled for the next entire cycle. - -

[0016] When supplying a high-power load from the mains line power grid, regulatory issues need to be observed. For example, harmonics generated by switching the AC power signal shall not exceed a threshold for harmonic current emission. Similarly, there are thresholds for regulatory and visible flicker which should be observed to provide an acceptable control apparatus. In some examples, the regulation may be the "Electromagnetic Compatibility (EMC) - Part 3-2: Limits - Limits for harmonic current emission (equipment input currents 16 A per phase)", IEC 61000-3-2, and/or the "Electromagnetic compatibility (EMC) - Part 3-3: Limits - Limitation of voltage changes, voltage fluctuations and flicker in public low- voltage supply systems, for equipment with rated current <16 A per phase and not subject to conditional connection", IEC 61000-3-3.

[0017] Fig. 3 a to 3 g illustrate examples of different sequences, including different switching modes, according to various examples for controlling a switching circuit, such as the triac based switching circuit 18 of fig. 1. The different sequences, enumerated as, can be used for controlling the switching circuit to output an AC power of 0% to 100% of a maximum power. In fig. 3, the black area indicates a state where the triac device 20 is turned ON, current may flow and the triac device hence transmits power; the white area indicates a state where the triac device 20 is turned OFF and hence blocks current and power; and the shaded area represents a triac switching adjustment zone, i.e. a phase range of the sinusoidal AC power signal within which switching is allowed to occur. Each of the switching sequences illustrated in fig. 3 includes one or more of the following switching modes: (a) switching ON the AC power at the start of a semi-cycle, and (b) switching ON the AC power during a predetermined first portion of a semi-cycle, i.e. in the triac switching adjustment zone; and (c) not switching ON the AC power during a semi-cycle; wherein, at the end of a respective semi-cycle, the AC power is switched OFF.

[0018] The triac device 20 is not switched at a phase angle falling within a second portion of the semi-cycle, wherein the second portion of the semi-cycle spans the rest of the semi-cycle, from the end of the predetermined first portion to the next zero crossing. In other words, if the triac device has not been switched ON during the predetermined first portion of a semi-cycle, it will remain OFF until the end of the respective semi-cycle. The predetermined first portion of the semi-cycle, or triac switching adjustment zone, can be defined to span a phase range from 0° to about 90°, or 0° to about 100°, or 0° to about 110°, or 0° to about 120°, or 0° to about 135°, under the assumption that the entire semi-cycle spans a phase range from 0° to 180°, just to mention a number of examples. The predetermined first portion of the semi-cycle - - is selected so that switching of the triac device occurs at a phase angle which is sufficiently far apart from the next zero crossing to avoid that the triac device is erroneously switched at the beginning of the next semi-cycle. As explained above, such erroneous switching can occur if the AC power signal is very noisy and if a switching trigger signal is received close to the zero crossing point. The safe distance can be determined under consideration of the particular application and the noise to be expected.

[0019] The different switching modes are distributed over sequences of semi-cycles to adjust the AC output power, such as the output power of the switching circuit 18, in such a way that harmonics and flicker are avoided or reduced. For example, in one approach, switching modes (a) and (b) are distributed equally over one sequence of semi-cycles as far as possible. Looking at this in another way, in one approach, switching modes (a) and (b) are distributed over one sequence of semi-cycles in such a way that the triac ON phases are balanced or resemble or approximate the sinewave signal as much as possible. Clustering of triac activation periods should be avoided. This can be explained with reference to the example of fig. 3a to 3g.

[0020] As shown in fig. 3a to 3g, the AC output power can be adjusted from 0% to 100% of a maximum power by combining different switching modes in a switching sequence. In the example of fig. 3 a to 3 g, a switching sequence comprises five semi-cycles or half waves; however, a different number of semi-cycles can be selected for a switching sequence, such as three to six semi-cycles. The numbers of semi-cycles in a sequence can be selected under consideration of the desired adjustment resolution and adjustment speed, avoidance of visible flicker and harmonics, among others. Selecting a switching sequence of five semi-cycles is a good compromise between fast power regulation capability and elimination of human visual flicker at the standard power grid frequencies of 50 Hz or 60 Hz. Human visual flicker can occur with lamps which radiate light in the visible spectrum when the lamps are switched at a switching frequency below the standard power grid frequency. For other types of lamps, such as infrared lamps, where visible flicker cannot occur, sequences of more than five semi-cycles can be selected. More generally, the current examples are not limited to any specific number of semi-cycles per sequence.

[0021] In the example of fig. 3a to 3g, it is assumed that the system is supplied from the mains line power grid, providing an AC power signal at 50 Hz. If the AC output power shall be zero, switching mode (c) is selected for each semi-cycle, as shown at fig. 3 a. In such a case, there will be no harmonics and flicker. This provides a fixed output power of 0%. - -

[0022] If the AC output power shall be above zero, the switching method of this example can provide a first fixed power value of about 10% of the maximum power. To obtain this 10% value, one of the semi-cycles in each sequence of five semi-cycles, e.g. the first semi-cycle, will be selected to switched ON AC power at 90°, or in the middle of the semi-cycle, so that AC output power is provided during the second half of this one semi-cycle; see fig. 3b. The next time the AC power signal shall be switched will be during the first semi-cycle of the next sequence. This provides a fixed output power of about 10% of the maximum power. Switching ON the triac at 90° of the first semi-cycle may create a certain amount of harmonics, as explained in further detail below. Further, switching ON AC power only at each fifth semi- cycle (each first semi-cycle of subsequent sequences) may create a certain amount of flicker. However, as the overall output power is low, the overall disturbance caused by said harmonics and flicker may be negligible.

[0023] For example, turning on a heating lamp emitting visible light once every fifth semi- cycle, at an operating frequency of 50 Hz, means that the heating lamp is turned on 10 times per second, creating optical flicker noticeable to the human eye. However, as the overall power is only one 10 ^th of the maximum power, the overall light emission amount would be very low so that the viewer is not disturbed.

[0024] Continuing with the example of fig. 3b, if an AC power of less than 10% of the maximum power shall be output, the triac device is to be switched ON at a phase angle between 90° and 180°. This may or may not be admissible, depending on the definition of switching mode (b). For example, if the predetermined first portion of the semi-cycle, which is available for performing a switching operation (triac switching adjustment zone), is defined to span a phase range from 0° to about 90°, as shown in fig. 3c, switching at a phase angle beyond 90° would not be admissible, and the AC output power could not be adjusted within range of 0% to 10% of the maximum power. As another example, if the predetermined first portion of the semi-cycle spans a phase range from 0° to about 135°, the AC output power could be adjusted down to a smaller fraction of the maximum power, i.e. to about 4% of the maximum power. In the latter case, the adjustment resolution is improved, but switching occurs more closely to the zero crossing point so that there is a higher risk of erroneously triggering the switching operation at a time when the next semi-cycle has started. If this happens, the triac would be switched ON during the entire next semi-cycle. - -

[0025] In the further description of the example, it shall be assumed that the predetermined first portion is defined as a phase range of 0° to 90° for each positive semi-cycle and 180° to 270° for each negative semi-cycle, assuming that each positive semi-cycle spans a phase range of 0° to 180° and each negative semi-cycle spans the phase range of 180° to 360°. However, the predetermined first portion may be defined differently.

[0026] As also illustrated in fig. 3c, during the predetermined first portion of the semi-cycle, i.e. in the triac switching adjustment zone, the switching phase angle can be selected between 0° and 90° (or 180° to 270°, depending on whether a positive semi-cycle or a negative semi- cycle is selected) to adjust the AC output power between 10% and 20% of the maximum output power. If the switching phase angle is selected to be 0° (or 180°), AC output power would be switched ON during one entire semi-cycle, which corresponds to an output power of 20% of the maximum output power and which would also correspond to switching mode (a).

[0027] If the AC output power is to be increased to a value between 20% and 40% of the maximum power, the example described selects an approach where two periods (semi-cycles) during which the triac is switched ON are distributed over the sequence of five semi-cycles; see example of fig. 3d. More specifically, when an AC output power of 20% to 40% of the maximum power is desired, the switching sequence of this example includes two times switching mode (b) wherein the respective semi-cycles during which the triac is activated, are distributed over the sequence of five semi-cycles, as evenly as possible. In the specific example described, switching mode (b) is selected for the second and the fifth semi-cycles, wherein the triac is deactivated (switching mode (c)) during the first, third and fourth semi-cycles. Alternatively, switching mode (b) could be selected for the first and third semi-cycles or first and fourth semi-cycles, or second and fourth semi-cycles, for example. Clustering of triac activation periods should be avoided. Using the triac switching adjustment zones in the second and fifth semi-cycles, the AC output power can be adjusted between 20% and 40% of the maximum power. Due to the relatively even distribution of the semi-cycles during which the triac is activated, generation of harmonics and flicker can be kept at a minimum.

[0028] If the AC output power is to be adjusted from 40% to 60% of the maximum output power, the example of fig. 3 selects an approach where one semi-cycle of switching mode (a) is combined with two semi-cycles of switching mode (b), as shown in fig. 3e. Again, the semi-cycles during which the triac is activated are distributed evenly over the switching sequence of five semi-cycles. Switching mode (a) is selected for the first semi-cycle, and - - switching mode (b) is selected for the fourth and fifth semi-cycle, in the example depicted in fig. 3e. Alternatively, switching mode (b) could be selected for the third and fourth semi- cycles, for example. Further, if switching mode (a) was selected for another one of the semi- cycles, switching mode (b) could be shifted to other semi-cycles accordingly, so as to evenly distribute the time periods during which the triac is activated. Using the triac switching adjustment zones, the AC output power can be adjusted between 40% and 60% of the maximum power.

[0029] If the AC output power is to be adjusted from 60% to 80% of the maximum output power, the example of fig. 3 selects an approach where two semi-cycles of switching mode (a) are combined with two semi-cycles of switching mode (b), as shown in fig. 3f. Again, the semi-cycles during which the triac is activated are distributed evenly over the switching sequence of five semi-cycles. Switching mode (a) is selected for the second and fifth semi- cycles, and switching mode (b) is selected for the first and third semi-cycle, in the example depicted in fig. 3. Alternatively, switching mode (b) could be selected for the first and fourth semi-cycles, for example. Further, if switching mode (a) was selected for another combination of semi-cycles, switching mode (b) could be shifted also to another combination of semi- cycles, so as to evenly distribute the time periods during which the triac is activated. Clustering of triac activation periods should be avoided. Using the triac switching adjustment zones, the AC output power can be adjusted between 40% and 60% of the maximum.

[0030] If the AC output is to be adjusted from 80% to 100% of the maximum output power, the example of fig. 3 selects an approach where three semi-cycles of switching mode (a) are combined with two semi-cycles of switching mode (b), as shown in fig. 3g. Again, the semi- cycles during which the triac is activated in the respective switching modes are distributed evenly over the switching sequence of five semi-cycles. Switching mode (a) is selected for the first, third and fourth semi-cycles, and switching mode (b) is selected for the second and fifth semi-cycle, in the example depicted in fig. 3. If switching mode (a) was selected for another combination of semi-cycles, switching mode (b) would be shifted accordingly to another combination of remaining semi-cycles, so as to evenly distribute the time periods during which the triac is activated either fully or partially. Using the triac switching adjustment zones, the AC output power can be adjusted between 80% and 100% of the maximum output power. [0031] Even if the same output power shall be generated during subsequent switching sequences, each switching sequence comprising a predetermined number of semi-cycles, the order of switching modes within the respective sequences can be but does not have to be identical but can be varied to evenly distribute the time periods during which the triac is activated, not just within one switching sequence but also among a number of switching sequences.

[0032] As a result of the switching approach described above, even if switching mode (b) is selected two times or less, the AC output power can be adjusted continuously over the entire range of 0% to 100% of the maximum output power, except at the very low power level where there is a jump from zero (0) to the first fixed power, such as 10% of the maximum power in the example described above.

[0033] In the selection of switching modes for each of the semi-cycles in a sequence of five semi-cycles according to various examples, switching mode (b) can be selected two times or less in order to reduce generation of harmonics which may be caused by sharp rising voltage edges created when the triac device is switched ON during one semi-cycle, rather than at zero crossing. Further, in the selection of switching modes for each of the semi-cycles in a sequence according to various examples, switching modes (a), (b) and (c) can be distributed as evenly as possible, in order to avoid clustering of time periods during which the triac is switched on and time periods during which the triac is switched off. This helps to reduce or avoid flicker.

[0034] Other sequences of semi-cycles can be selected and the switching modes can be adapted and distributed accordingly. For example, if a sequence comprises three semi-cycles, and assuming that the predetermined first portion of the respective semi-cycles according to switching mode (b) spans a phase range from 0° to 90° (or 180° to 270°), the minimum AC power output greater than zero (0) would be about 17% of the maximum output power, corresponding to the scenario depicted in fig. 3b. If an AC output power of less than 17% of the maximum power is needed, it should be possible to switch ON the triac device at a phase angle between 90° and 180°. This may or may not be admissible, depending on the definition of switching mode (b). Reference is made to the description of fig. 3 above.

[0035] Continuing with the example of a sequence of three semi-cycles, determining and distributing switching modes for adjusting the AC output power in steps of about 17% to 33%, 33% to 67%, and 67% to 100% of the maximum power can be performed in a manner analogous to determining and distributing the switching modes for the five semi-cycles sequence - - described above. The range of 17% to 33% of the maximum power corresponds to the switching sequence of fig. 3 c, except that two semi-cycles where the triac is not activated are missing. The range of 33% to 67% of the maximum power corresponds to the switching sequence of fig. 3d, except that there is only one switching mode (c) between two switching modes (b). The range of 67% to 100% of the maximum power corresponds to the switching sequences of fig. 3e, except that there is no switching mode (c).

[0036] Similarly, for a sequence of four or six semi-cycles, switching modes and their distribution across the sequence can be determined accordingly.

[0037] Fig. 4a shows an example of a switching sequence of five semi-cycles used for switching the triac device 18 of the system of fig. 1 or another switching device. In this specific example, switching mode (a) is selected for the second cycle, switching mode (b) is selected for the third and fifth cycles, and switching mode (a) is selected for the first and fourth cycles. In the third cycle, the triac is activated at a phase angle of about 60°, and in the fifth cycle, the triac is activated at a phase angle of about 70°. The resulting AC output power is about 50% of the maximum output power. More specifically, in the example depicted in fig. 4a, the maximum current amplitude of the switched AC output power is I_max = 60A and the root mean square current Ijrms of the switching sequence of fig. 4a is I(rl)_rms = 31.2A. In a comparative example, shown in fig. 4b, the same the root mean square current Ijrms,

I(r2)_rms = 31.2A, and hence the same switched AC output power can be obtained by switching ON the triac in the middle of each cycle, i.e. at 90° in the positive semi-cycle and 270° in the negative semi-cycle. These two scenarios have been compared by simulation and by calculating the respective Fourier components at the operating frequency of 50 Hz and its harmonics, up to the 7 ^th harmonic. The results are reproduced in the two tables below.

. _

[0038] Table I: Fourier components of I(rl):

[0039] Table II: Fourier components of I(r2):

[0040] As illustrated above, the AC power control method using defined switching sequences of selected switching modes according to the various examples can significantly reduce higher order harmonics which are generated by the switching process. Focusing on the "worst" harmonics, namely the third order, fifth order and seventh order harmonics, and taking into account the normalized Fourier components, it can be observed that there is substantial improvement in harmonic suppression, i.e. from 0.54, 0.18, and 0.18 down to 0.18, 0.08, and 0.02, i.e. up to almost a factor of 10. Harmonics hence can be reduced substantially when compared to a fully periodical switching strategy. - -

[0041] Fig. 5 shows a flow diagram of a method of controlling AC power to drive a load according to an example. The method starts at block 48 where a power level to be applied to the load is determined. The power level can be determined relative to a maximum AC power, such as a power rate of 0% to 100% of the maximum AC power. The power level can be determined in a closed or open control loop, e.g. for controlling the power/temperature of the heating device or another load.

[0042] In block 50 a predetermined number of semi-cycles of an AC power signal are identified to define a sequence. The sequence may comprise any suitable number of semi-cycles, such as three, four, five, or six semi-cycles. The sequence also can be defined in terms of entire cycles or another suitable entity.

[0043] Once a sequence has been defined, a switching mode is selected for each semi-cycle in the sequence; see block 52. The switching modes can comprise: (a) switching ON the AC power at the start of the semi-cycle, (b) switching ON the AC power during a predetermmed first portion of the semi-cycle; and (c) not switching ON the AC power during the semi-cycle. The switching modes can be selected in such a way that a desired AC output power is achieved wherein the predetermined first portion of the semi-cycle can be used as a switching adjustment zone. The switching modes can be selected in such a way that periods during which AC power is supplied to the load, e.g. because a switching device, such as a triac, is turned ON, are distributed evenly, as far as possible, over the semi-cycles of the sequence. This minimizes generation of flicker. If switching mode (b) is used two times or less, generation of harmonics can be minimized and still the output power can be adjusted continuously over almost the entire range from 0% to 100% of the maximum output power, except at the very low power level with there is a jump from zero (0) to the first fixed power, such as 10% of the maximum power in the example described above. The decision which sequence of switching modes to select and at which points within a switching adjustment zone (predetermined first portion of semi-cycle according to switching mode (b)) to switch the AC power can be easily managed by a local microcontroller, such as controller 16 shown in fig. 1. The switching modes for each cycle are selected according to the overall power level to be achieved, as determined in block 48.

[0044] In block 54, and an AC power signal applied to a switching circuit, such as the circuit 18 shown in fig. 1, is switched in each semi-cycle according to the selected switching mode. - -

At the end of each semi-cycle, the AC power can be switched off; see block 56. If a triac is used as a switching device, AC power would be switched off automatically at the end of each semi-cycle (next zero crossing after the triac has been switched ON) unless the switching device is triggered again to switch on AC power. In block 58, the switched AC power is applied to a load, such as load 30 shown in fig. 1, so as to output the power level as determined in block 48. This load can be a heating device or a group of heaters in a 3D printer, for example.

[0045] If the system and method described in this disclosure are applied to a heating lamp or a group of heating lamps in a 3D printer, the heating lamp or lamps can be switched on and off as needed to generate a desired heating temperature or heating temperature profile for fusing a 3D built material. If there are several heating lamps or several groups of heating lamps which are to be controlled to generate different heating temperatures or heating temperature profiles, a respective sequence of switching modes can be determined for each heating lamp or group of heating lamps in such a way that the AC power consumption across the several sequences is equally distributed or approximately equally distributed. Additionally or alternatively, the several sequences of switching modes can be randomized to the extent that the overall switching strategy still reduces harmonics and flicker, applying the principles explained above.