Fullscreen HTML5Med Res (??MB)HTML4

Active Mains Soft Starter Part One by John Clarke High startup current appliances can be dangerous, damage your work, cause brownouts or trip out the circuit breaker when power is first applied. This Soft Starter prevents the high surge current, replacing it with a slow current build-up and reducing the ‘kick’ you get from many tools. H ave you ever used a power tool that rips out of your hand when power is first applied? Or do you have a bank of computers or audio equipment (or similar) that you want to power up together from a single power point? When you do so, sometimes the circuit breaker might trip, forcing you to go to the switchboard and reset it. Tools with a motor that are powered from the mains, such as circular saws, drop saws, hand grinders and routers can make a sudden movement Suitable for fixed or portable power tools rated up to 750W ➠ Switch on at GPO or equipment power switch, including triggers ➠ Relay contacts bypass soft start circuitry at completion for minimal power loss ➠ Six startup rate options from half a second to 10 seconds ➠ Indicators for power presence, soft starting and soft start end ➠ 10A continuous rating ➠ Uses trailing edge phase control ➠ Features & Specifications siliconchip.com.au inrush current can be caused by the appliance or appliances using a toroidal transformer or a switch-mode supply that rectifies the mains supply into a large capacitor or capacitor bank. The capacitance represents a near short circuit when power is first applied, causing a massive surge current. This new Active Mains Soft Starter significantly reduces startup current, solving that problem. It’s designed for devices that might be restarted frequently (like power tools), and its Active Soft Starter ➠ Also suits large amplifiers, computers or other equipment with a high inrush current as the torque from the motor startup rotates the tool. This can cause the tool to move dangerously. In the case of a saw, drill or router, it could move the cutting piece off position and possibly damage your work. You might also hear a nasty “splat” from the switch or plug when the equipment is powered up, indicating that it is being worn out by handling the high inrush current. All of that can be solved with a soft starter like this one. As well as large motors, a high Australia's electronics magazine February 2023  33 Fig.1: the mains waveform is a 50Hz sinewave with a positive voltage half the time & a negative voltage the rest of the time. effect will not diminish, nor will it overheat with multiple restarts if used with equipment within its ratings. You can use the Soft Starter with motorised tools up to 750W and appliances with substantial capacitance. Two ways to use it One way to use the Soft Starter is to have the appliance already plugged into the Soft Starter and switched on. You then switch on power at the GPO (general purpose outlet). That is ideal if you want to power up several appliances together from one power point. In this case, the soft start process begins at power-up (if an appliance is connected). Once the soft starting is completed, it supplies the full mains voltage until it is switched off at the power point. The second method of using the Soft Starter is to have it powered up via the power point (GPO), then switch the appliance on and off with its own switch. This method is ideal when using power tools. For both methods, the soft starter detects when the appliance is switched on and off by monitoring its load current. Soft starting only begins when current flow is detected. When the appliance is switched off, current flow ceases, and the power to the tool is also switched off, ready for another soft start. 750W rating We tested the Soft Starter with various loads and power tools and found that it worked well for tools up to 750W. Some parts got uncomfortably hot when used with tools that draw more than that. Also, the ratings of some of the devices used are only sufficient up to that power level. This is less of a concern when switching equipment like computers and amplifiers, as their inrush current periods are short. In that case, you can comfortably connect up to 10A (2.3kW) of equipment to the output. For power tools above the 750W rating, consider building our Refined Full-Wave Motor Speed Controller that incorporates soft starting (April 2021; siliconchip.au/Article/14814). It is rated to handle 10A and therefore should handle any power tool that plugs into a standard GPO. You could leave it set to full speed all the time and just utilise its soft-starting feature. Presentation The Active Soft Starter is housed in a compact plastic case with an IEC mains input connector at one end and a GPO for the appliance. There are three neon indicators on the top. One shows when input power is applied; the second shows the slow voltage rise to the appliance, while the third lights when the soft start period has ended. The neons are very sensitive and light up with a minimal current applied, so they don’t show the full extent of the soft starting. However, they help to show what the device is doing. Soft starting methods The standard method to reduce the surge current is to add resistance in series with the mains supply, reducing the maximum current. We previously published two soft starters using that method, one in April 2012 (siliconchip.au/Article/705) and the other in July 2012 (siliconchip.au/ Article/601). Both utilised negative temperature coefficient (NTC) thermistors. These devices act as a resistor that reduces its resistance as it heats up from the current flow through it. As it starts cold, the resistance is high, so the current is restricted. Then as the thermistor heats up, the resistance drops and allows more current to flow. In both designs, after some time, the thermistor is bypassed by a relay to provide the full mains supply to the appliance. Bypassing the thermistor after the soft start prevents further heating of the thermistor, allowing it to cool down and be ready to provide another soft start when required. Still, if the appliance is powered up repeatedly at close intervals, the thermistor does not have time to cool between uses, so its resistance can be quite low on successive starts. This means that the soft starting is not as effective in such cases. Another consideration is whether the NTC thermistor can survive longterm use conducting current for an appliance that draws significant current at switch-on. If power is switched on at the maximum voltage point in the mains waveform, the initial current can be extremely high, especially if the thermistor is still hot. Over time, that can damage and possibly destroy the device. While our new Soft Starter does use a thermistor, it also includes phase control that initially applies a small portion of the mains waveform. The proportion of the mains waveform applied to the load increases slowly until the full mains cycle is applied. A relay contact then closes to bypass the soft start circuitry. In doing so, it causes very little heating in the thermistor, so repeated starts are not a problem, and the device is very reliable. Also, the phase control always starts at the beginning of the mains cycle, when the mains voltage is close to 0V. The control scheme used is called trailing-edge phase control and differs from the leading-edge phase control Fig.2: traditional leading-edge phase control varies the switch-on point during the mains cycle but always switches off at the zero crossing. So the earlier it switches on, the more power is applied to the load. …continued opposite 34 Silicon Chip Australia's electronics magazine siliconchip.com.au Warning: Mains Voltage The entire circuit of the Active Soft Starter floats at mains potential and could be lethal should you make contact with it. Don’t assume that because we use isolation between different parts of the circuit that some parts are safe to touch – they are not! The isolation between parts of the circuit is to allow for the differing voltage potentials in parts of the circuit rather than for safety. Fig.3: in the Soft Starter circuit, the N-channel Mosfet is connected to a diode bridge, so current always flows from its drain to its source. That way, its parasitic body diode is never forward-biased. The current paths are shown for when the Active conductor is more positive than the Neutral (i1) and for when the Active is negative with respect to Neutral (i2). method that is often used. Leading edge vs trailing edge Fig.1 shows the mains waveform, while Fig.2 shows these two types of power control. Our mains electricity supply (nominally 230V AC) is a sinewave that repeats 50 times per second (ie, at 50Hz). For phase control, power is applied over a portion of each half of the mains cycle. The waveforms labelled “A” in Fig.2 show the situation when there is a small phase angle of the full sinewave applied to the load. In the left-hand waveform, the voltage is applied to the load from late in the waveform until the zero-crossing. However, on the right, the voltage is applied for a short period beginning from 0V, switching off a little while later. Both waveforms apply the same RMS voltage to the load and have the same area under the shaded portion of the sinewave curve. The difference is that one switches on at the end of the half cycle (leading-edge phase control), while the other switches on at the beginning of the cycle (trailing-edge phase control). Leading edge phase control has been used for around 50 years, mainly for dimming incandescent lamps. That is because it can be implemented using a simple circuit based on a Triac, a semiconductor device that switches on when its gate is driven. It can’t be switched off via the gate; instead, it switches itself off when the current flow through it drops to near zero. However, leading-edge phase control is unsuitable for providing soft starting to loads that charge a capacitor. If a voltage is suddenly applied to that type of circuit, it will create a high surge current, regardless of whether the phase angle it is on for is only a small portion of the mains waveform. The solution is to use a trailing edge phase control instead. The switching device now turns on at the mains zero-crossing where there is little or no potential difference between Active and Neutral. The voltage then rises relatively slowly, following the sinewave shape, to charge the capacitance. Current is drawn from the mains in much smaller and more tolerable pulses. Note that a typical circuit that charges capacitors includes a rectifier so that the capacitor is charged with DC voltage. For soft starting, we increase the duration of the waveform applied to the load over time, so the capacitor charges in small increments as the next cycle has a slightly greater phase length and hence a slightly higher peak voltage. The capacitor is ultimately charged, but at a slower rate than if the full supply were applied at power on. By the way, trailing edge control is also used for dimming LED lamps because they are usually powered by a capacitor-input switch-mode power supply (SMPS). If you are interested in learning more about this, Leo Simpson wrote about leading and trailing edge dimmers in July 2017 (siliconchip.au/ Article/10712). The disadvantage of trailing-edge phase control is that a Triac cannot be used. It needs a switching device that can be switched off at any part of the mains waveform. Fig.3 is a simplified version of how we implement trailing-edge phase control. We use a Metal-oxide Semiconductor Field Effect Transistor (Mosfet) and a rectifier bridge. The Mosfet is connected within the diode bridge, so current always flows from its drain terminal to its source. The current paths are shown for when the Active is more positive than the Neutral (i1) and for when the Active is negative with respect to Neutral (i2). The Mosfet circuit allows us to switch mains power to the load on or off at any point in the mains cycle. Results We measured the startup current for a bank of amplifiers that, when switched on normally, would trip the Trailing-edge phase control achieves a similar result, but instead, the load is switched on at the zero crossing and then switched off at some point later in the mains cycle. The later the switch-off, the more power is applied to the load. siliconchip.com.au Australia's electronics magazine February 2023  35 Scope 1: switching on a bank of amplifiers, the current peaks at 138A until the circuit breaker trips after 6ms. Scope 2: with the Soft Starter, the bank of amplifiers can be switched on without tripping the breaker. Scope 3: the 750W angle grinder draws 40A on the first mains cycle, dropping to 6A after half a second. Scope 4: with the Soft Starter, the angle grinder takes four times longer to spin up and no longer kicks. circuit breaker. We also tested it with a 750W angle grinder. For the amplifiers, the startup load is essentially a bank of capacitors that charges up at power-on. When discharged, they effectively form a short circuit, resulting in a huge current flow as power is first applied. This is shown in Scope 1, with each vertical division corresponding to 50A (10A = 1V here). The startup surge current (sometimes called the inrush current) peaks at about 138A before the circuit breaker trips. The time for the circuit breaker to trip is less than a mains half-cycle of 10ms (we measure 6ms to the small negative spike). Scope 2 shows the startup current for the same load with the Active Soft Starter connected, over a longer period (the timebase is now 50ms instead of 5ms). The is much more subdued, with only small peaks to a maximum of around 17A. The amplifier capacitor 36 Silicon Chip banks are fully charged after about 500ms, hence the drop-off in the current spikes. For the 750W angle grinder, the startup current (Scope 3) peaks at nearly 40A in the negative direction and then about 34A in the positive direction, tapering down to about 6A after 450ms. With the Soft Starter connected (Scope 4, again with a longer timebase), a small initial current rises to about 13A peak after 750ms and tapers to about 5A at the two-second mark. The fact that it takes considerably longer to spin up indicates that it has much less of a ‘kick’ to it. Block diagram Block diagram Fig.4 shows how the circuitry is arranged in the Active Soft Starter. Incoming mains Active (A) passes through a fuse and to the mains output for connection to the appliance Australia's electronics magazine while current transformer T1 monitors the current flow. The incoming Neutral (N) does not directly connect to the output, but instead, goes via the soft-start circuitry comprising Mosfet Q1 and bridge rectifier BR1. The relay bypasses this arrangement after the soft-start period. The Active mains wire passes through the centre of the current transformer T1 twice, forming its primary winding. The isolated secondary winding produces a voltage proportional to the Active current. This is rectified using a precision full-wave rectifier and low-pass filtered to give a smoother DC voltage, then fed to the AN1 analog input of microcontroller IC1. The current measurement is used for two purposes. One is to monitor when the appliance is switched on to initiate soft starting. The other is to determine when the appliance is switched siliconchip.com.au Fig.4: a simplified block diagram of the Active Soft Starter. The soft-start circuitry is connected between the incoming and outgoing Neutral; current flow is monitored in the Active wire so that it knows when to activate the soft-starting procedure. RLY1 bypasses the softstart circuitry once the full voltage has been applied to the load for maximum efficiency. off, to reset the circuitry, ready for the next power-on. Microcontroller IC1 controls all the soft starter functions. It monitors the appliance current, controls the gate of Mosfet Q1 and the coil of the relay, monitors the soft start rate setting potentiometer and also monitors the mains waveform zero-crossing timing. The gate drive for the Mosfet needs to be referenced to the negative terminal of the bridge rectifier, which is neither at Neutral nor Active potential. So For IC1 to drive the Mosfet, there needs to be electrical isolation between IC1 and Q1’s gate. This is achieved using an isolated power supply and an isolated gate driver. The isolated supply is produced via the GP4 digital output of IC1 that delivers a 1MHz, 5.5V square wave. That waveform is stepped up and isolated via transformer T2. After rectification and filtering, the result is a DC voltage suitable for driving the gate of Q1. The Mosfet gate is controlled via the GP0 digital output of IC1. This drives an opto-coupler (IC3) containing an infrared LED that is electrically isolated from the opto-coupler’s siliconchip.com.au optically switched transistor. That transistor controls the voltage at the gate of Mosfet Q1. The isolated drive for the relay coil is via an optically-coupled Triac driver (IC4) that connects the lower end of the coil to the output Neutral. The relay has a 230V AC coil with the top end connected to Active and the bottom end connecting to IC4. IC4 has an internal LED that optically triggers the output Triac. It is typically used to drive the gate of a larger Triac, but for our circuit, we are just using it to power the relay coil. The power supply for IC1 is not shown in Fig.4; its supply is derived via a mains-rated capacitor that acts as a current limiter to a zener diode clamp, resulting in the 5.5V supply voltage. The positive side of this supply is referenced to mains Active. Potentiometer VR1 is used for the soft start rate adjustment. It is connected across that 5.5V supply, producing a varying voltage at the microcontroller’s AN2 analog input. Neon indicators NEON1 lights when there is mains Australia's electronics magazine power at the input. NEON2 is connected across the mains output, so it starts dim and reaches full brightness when the soft start period ends. We call this the “run” indicator. Finally, NEON3 lights when the relay is on after the soft start period completes. This is called the soft start “end” indicator. Circuit details The entire circuit is shown in Fig.5. A lot of the circuitry has already been explained by the block diagram. However, several parts of the circuit haven’t been described in any detail. As mentioned earlier, Triac-­output opto-coupler IC4 drives the relay coil. We are using the MOC3042 with zero voltage crossing detection, so its Triac always switches on when the mains supply is at zero voltage. That is not strictly necessary for our circuit, but it does not hurt. Its internal Triac between pins 4 and 6 is guaranteed to trigger, provided there is at least 10mA through the internal infrared LED between pins 1 and 2. We also include a snubber across the Triac terminals, comprising a 22nF February 2023  37 Fig.5: IC1 is the controlling PIC while generating an isolated Mosfet gate voltage supply by feeding a high-frequency square wave into transformer T2. It controls the Mosfet gate across that isolation barrier using opto-coupler IC3, and it monitors the output of the current sense transformer via the full-wave precision rectifier formed by dual op amp IC2. Two transient voltage suppressors and a zener diode protect Mosfet Q1 from voltage spikes. X2-rated mains capacitor and a 150W resistor, connected in series between its pins 4 and 6. This limits the voltage rise time so that the Triac will not switch itself on when power is first applied to the circuit. The 1MW resistor just discharges the capacitor when power is off for safety. The snubber limits sudden voltage rises across the Triac by charging 38 Silicon Chip over time via the 150W resistor. This prevents the voltage from rising faster than 1000V/μs, which is the maximum dV/dt rating for the Triac in IC4, below which it is guaranteed not to switch on by itself. Another precaution against that is connecting pin 4 of the Triac to the Neutral output of the soft-start circuitry rather than directly to the Australia's electronics magazine incoming Neutral. So when power is first applied, there is no voltage across the Triac. As the soft start process begins, the voltage across it rises at a controlled rate. Protecting Mosfet Q1 As well as a snubber for IC4, there is a 220nF/470W snubber across the AC terminals of BR1 to reduce the siliconchip.com.au magnitude of voltage spikes seen by Mosfet Q1. This also has a 1MW bleeder resistor for safety. This snubber also provides a small current flow when an appliance is switched on before the soft starting process has activated. This is enough current to detect and initiate the soft start. Q1 is also protected against over-­ voltage conditions that could destroy the device; it has a 500V maximum drain-source rating. Two transient voltage suppressors (TVS) are used to prevent the voltage from going over that limit. TVS2 is connected directly between the Mosfet’s drain and source and conducts to shunt voltage at the TVS clamp voltage of 400V (255V AC rectified gives ~360V DC). However, this TVS can be damaged if the over-­ voltage spike has too much energy, so a second line of defence is used. A second TVS, TVS3, is connected in series with a 100W resistor between the Mosfet drain and gate. If the drain voltage rises too high, TVS3 conducts and causes the Mosfet gate voltage to rise, so the Mosfet starts to conduct, shunting the voltage spike itself. Zener diode ZD3 prevents the gate voltage from going over 15V in this case, which could otherwise damage it, while the 100W resistor limits the zener current to a safe level. Current detection Current transformer T1 produces an output current from its secondary winding that’s proportional to the current flow through the Active mains wire. The 10kW loading resistor gives about 4V AC output with a current flow of 1A and one turn of the Active mains wire through the current transformer core. We use two turns through the core, giving about 4V AC with 500mA current through the primary. While the input current to output voltage conversion is not very linear using a 10kW loading resistance, we use the high value to improve sensitivity. A 100W loading resistor would be used instead for this current transformer to measure current accurately. That would provide a more linear relationship but only gives 1V AC for a 10A primary current. Current sense voltage rectification Another transient voltage suppressor (TVS1) clamps the output voltage siliconchip.com.au from transformer T1. This limits the current into the following op amp inputs to a safe level. The output from T1 needs to be rectified to give a DC voltage suitable for monitoring by microcontroller IC1. A precision full-wave rectifier is used, made from dual op amp IC2 and associated resistors; note the lack of diodes. The gain of this precision rectifier is 1.5 times. While it may appear impossible to rectify the incoming AC voltage without diodes, it is possible, provided that the op amp has specific characteristics. The op amp needs to be able to operate with an input below its negative supply rail, and the op amp must be able to pull its output close to that negative supply rail. Here, we are using an MCP6272 dual op amp (IC2). One stage (IC2b) is connected as a unity gain buffer, while the other (IC2a) provides the 1.5 times gain. To understand how the rectification works, refer to Fig.6, where A to E correspond to the waveforms at the identically labelled parts of the circuit in Fig.5. That is assuming that our example waveform is present at point A. Sample waveform A is a 2V peak-topeak sinewave. For the negative half of the cycle, the signal applied to the non-inverting pin 5 input of IC2b via the 15kW resistor will cause the voltage at that pin (point B) to be clamped at around -0.3V due to IC2’s internal input protection diode. The output of IC2b (point C) therefore sits at 0V during negative portions of the cycle, since its negative supply rail is at 0V, and it cannot pull its output lower than that. IC2a adjusts its output (point E) so that the voltage at its inverting input pin 2 (point D) matches the voltage at non-inverting input pin 3 (point C). Since the 10kW resistor from point D to ground has no voltage across it, it plays no part in the circuit during the negative portions of the cycle. With the 10kW resistor essentially out of the circuit, IC2a operates as a standard inverting amplifier with both inputs (points C and D) at 0V. Its gain is therefore -30kW divided by 20kW, which equals -1.5 times. So the -1V peak of the waveform is amplified and inverted to produce +1.5V at point E. The way it works for a positive voltage at the input (point A) is more Australia's electronics magazine complicated. Firstly, the voltage at pin 5 (point B) is reduced compared to the 1V peak at the input. This is because of the divider formed by the 15kW and 18kW resistors, so the voltage becomes 0.5454V (1V × 15kW ÷ [15kW + 18kW]). Point C will also peak at 0.5454V since IC2b is working as a unity-gain buffer producing the same voltage at its output as its non-inverting input. Once again, op amp IC2a adjusts the output voltage (point E) so that the voltage at the inverting input at pin 2 (point D) matches the voltage at the non-inverting input, pin 3 (point C). To determine the resulting voltage, we must calculate the currents through the three resistors connecting to the inverting input of IC2a at point D. 1. The current through the 10kW resistor is the waveform D voltage divided by 10kW. This peaks at 54.54μA (0.5454V ÷ 10kW). 2. The current through the 20kW resistor; with 1V peak at the input (point A), there will be 22.73μA ([1V[A] − 0.54V[D]] ÷ 20kW). So we have 22.73μA flowing into the node at point D via the 20kW resistor and 54.54μA flowing away from that node via the 10kW resistor. The extra Fig.6: these waveforms demonstrate how the active precision rectifier used for current monitoring works. They correspond to the expected waveforms at the points marked A-E on the circuit for the condition where there is a 2V peak-to-peak sinewave at point A, corresponding to a resistive load drawing about 88mA RMS. February 2023  39 input (GP3) is filtered with a 4.7nF capacitor, providing a near-zero voltage when the mains voltage is at zero. IC1’s pin 4 input detects when this voltage changes from being positive to zero or negative and vice versa. The voltage at pin 4 is clamped by the internal protection diode to -0.3V during the negative part of the cycle. For positive excursions of the mains waveform, diode D2 clamps the voltage to about 0.6V above the 5.5V supply or close to 6V. This diode is required since the pin 4 input is not protected with a diode to the positive supply. That’s so this input can be used for programming the microcontroller, where the voltage at this pin needs to go above the supply voltage. A sneak peek at the assembled PCB for the Active Mains Soft Starter, with construction details coming next month. current to balance currents at node D needs to come via the 30kW resistor. This is 31.81μA (54.54μA − 22.73μA). Remembering that voltage at point D peaks at 0.54V, the required voltage at point E is 1.5V (31.81μA × 30kW + 0.54V). So the circuit operates as a full-wave rectifier with a gain of 1.5. The degree of precision depends on the op amp parameters and resistor tolerances. The lower the offset voltage of the op amp and the lower the op amp input bias current, the more accurate the fullwave rectification will be, particularly at low signal levels. Fortunately, we are not overly concerned with absolute accuracy here. We just need full-wave rectification of the incoming AC signal from the current transformer. Scope 5 shows the 1V peak sinewave at the input to the full wave rectifier (point A) on channel 1, shown in yellow. Below that is the full-wave rectified waveform at point E, shown in cyan. A 2.2kW resistor and 10μF capacitor filter the rectified waveform to produce a smoothed DC voltage suitable for the IC1 to monitor via its AN1 analog input and internal analog-to-digital converter (ADC). Mains zero-crossing detection IC1 monitors the mains waveform at the mains Neutral via a 330kW 1W resistor. The voltage at its pin 4 digital Scope 5: the input to the active rectifier at point A and the output below (point E). Note the gain. 40 Silicon Chip Mosfet gate drive To drive the Mosfet gate, we need an isolated DC supply and a method of connecting and disconnecting that supply to the gate. As mentioned previously, these voltages need to be galvanically isolated from IC1. The isolated DC supply is generated by applying a 1MHz square wave to the primary winding of high-­ frequency transformer T2 from IC1’s clock output at GP4 (pin 3). This is ¼ the frequency of its internal 4MHz oscillator. The primary has 10 turns, while the secondary has 48, giving a 4.8:1 voltage ratio. Since the primary is a 5.5V peak-topeak square wave, we can expect the secondary to deliver a 26.4V (5.5V × 4.8) peak-to-peak square wave. After half-wave rectification by diode D3, we obtain a 13.2V DC output that is filtered by a 1μF capacitor. 15V zener Scope 6: the isolated Mosftet gate drive signal. It switches on faster than it switches off due to the isolation scheme. Australia's electronics magazine siliconchip.com.au diode ZD2 limits the voltage to a safe level for the Mosfet gate. The Opto-coupled output transistor of IC3 switches the Mosfet gate on or off. It is driven by the pin 7 digital output (GP0) of IC1. When this is high (at 5.5V), it drives the internal infrared LED of IC3 via a 1.5kW current-limiting resistor. The LED then lights and switches on the output transistor within IC3 that connects the 13.2V DC supply to the gate of Q1 via a 47W resistor. When the GP0 output of IC1 goes low (to 0V), IC3’s LED switches off, so the gate of Q1 is pulled to 0V by the 22kW resistor, switching the Mosfet off. Scope 6 shows the gate drive to the Mosfet when driven for 5ms on and 5ms off at 100Hz. When switched on, the gate voltage is initially 14.3V, drooping to 12.3V over the 5ms period. The voltage droop is due to the 1μF capacitor being loaded by the 22kW gate-source resistor. The switch-on rise time is around 43μs and the fall time is 324μs. The fall time is longer due to the 22kW discharge resistor having a higher resistance than the opto-coupler output transistor and 47W resistor that charges the gate up. Power supply Power for microcontroller IC1 and op amp IC2 is derived directly from the mains using a 470nF X2 mainsrated safety capacitor. The circuit operates by transferring charge to a 470μF capacitor via zener diode ZD1 and diode D1. For one polarity of the mains waveform, D1 is reverse-biased and ZD1 is forward-biased, so the charge from the 470nF capacitor is transferred to the 470μF supply filter capacitor. During the other half of the mains waveform, diode D1 is forward-biased and the zener diode clamps to 6.2V between the +5.5V supply rail and the cathode (K) of D1. Since the forward voltage of diode D1 is about 0.7V, the overall voltage across the 470μF capacitor is limited to 5.5V (6.2V − 0.7V). Next month The follow-up article next month will have all the construction details for the Active Mains Soft Starter, along with the testing procedure and usage instructions. SC siliconchip.com.au Parts List – Active Soft Starter 1 double-sided, plated-through PCB coded 10110221, 159 × 109mm ● 1 171 × 121 × 55mm polycarbonate or ABS enclosure [Altronics H0478, Jaycar HB6218] 1 153 × 107mm panel label 1 10A IEC panel-mount mains input socket with integral fuse holder [Altronics P8324, Jaycar PP4004] 1 10A IEC mains power lead 1 mains GPO socket [Altronics P8241, Jaycar PS4094] 1 Talema AX1000 or AC1010 10A current transformer (T1) ● 1 Hongfa HF105F-4/240A1HSTF 30A 240VAC chassis mount relay, 240V AC coil (RLY1) ● 1 SL32 10015 15A 265V AC NTC thermistor (NTC1) ● 3 plastic-bodied mains neon indicators (NEON1-NEON3; optional) [Altronics S4016, Jaycar SL2630] 1 10A M205 fast-blow fuse (F1) 4 2-way 15A 300V screw barrier terminals (CON1-CON4) [Altronics P2101] 1 100kW linear PCB-mount potentiometer (VR1) [Altronics R1948] 1 8-pin DIL IC socket (for IC1) 1 18 × 10 × 6mm ferrite toroid (for T2) [Jaycar LO1230] Hardware & wire 1 1.25m length of 0.25mm diameter enamelled copper winding wire (for T2) 2 4.8mm insulated female spade crimp lugs 1 350mm length of (blue & brown) 7.5A mains-rated wire 1 200mm length of blue 10A mains-rated wire 1 250mm length of brown 10A mains-rated wire 1 150mm length of green/yellow striped 10A mains-rated wire 1 75mm length of 10mm diameter heatshrink tubing 1 20mm length of 5mm diameter (blue, red & green) heatshrink tubing  1 20mm length of 3mm diameter (blue & red) heatshrink tubing  1 20 × 15mm piece of thermal transfer tape [Altronics H7240, Jaycar NM2790] 2 M3 × 10mm Nylon countersunk machine screws 2 M3 × 15mm panhead machine screws 4 M3 × 6mm panhead machine screws 4 M3 hex nuts 17 100mm cable ties  black tubing can be used instead, if preferred. Semiconductors 1 PIC12F617-I/P 8-bit microcontroller programmed with 1011022A.hex, DIP-8 (IC1) ● 1 MCP6272T-E/SN dual rail-to-rail op amp, SOIC-8 (IC2) ● 1 4N28 or 4N25 opto-coupler, DIP-6 (IC3) ● 1 MOC3042M or MOC3043M zero-crossing triggered Triac driver, DIP-6 (IC4) ● 1 SIHS36N50D-GE3 36A 500V N-channel Mosfet, TO-247 (Q1) ● 1 PB5006 45A 600V bridge rectifier (BR1) ● 1 6.2V 1W zener diode (ZD1) [1N4735] ● 2 15V 1W zener diodes (ZD2, ZD3) [1N4742] ● 1 4KE15CA bidirectional TVS, 400W, 12.8V standoff (TVS1) [Jaycar ZR1160] ● 1 1.5KE400CA bidirectional TVS, 1500W, 342V standoff (TVS2) [Jaycar ZR1180] ● 1 4KE400CA bidirectional TVS, 400W, 342V standoff (TVS3) [Jaycar ZR1164] ● 1 1N4004 400V 1A diode (D1) ● 2 1N4148 75V 200mA diodes (D2, D3) ● Capacitors 1 470μF 16V PC electrolytic 1 220nF X2-rated metallised polypropylene (PP) 2 10μF 16V PC electrolytic 4 100nF 63V or 100V MKT polyester 1 1μF 50V multi-layer ceramic 1 22nF X2-rated metallised polypropylene (PP) 1 470nF X2-rated metallised PP 1 4.7nF 63V or 100V MKT polyester Resistors (all 1/2W metal film ±1% unless noted) 3 1MW 1W ±5% 1 15kW 1 330W 1 330kW 1W ±5% 2 10kW 1 150W 1W ±5% 1 30kW 1 2.2kW 1 100W 1 22kW 1 1.5kW 2 47W 1 20kW 1 1kW 5W ±5% wirewound 1 18kW 1 470W 1W ±5% ● these parts are available as part of a set from the Silicon Chip Online Shop, Cat SC6575, for $100 + P&P. All the other parts are available from Jaycar or Altronics. Australia's electronics magazine February 2023  41