Fullscreen HTML5Med Res (??MB)HTML4

Flavio Spedalieri’s Solid-State Flame Discharge Tesla Coil This relatively small and simple device generates extremely high voltages, enough to form a ‘flame discharge’ resembling a candle flame. It can also demonstrate wireless power transmission by lighting up neon globes and fluorescent lamps at some distance. T he inspiration for this project came from a YouTube video by Jay Bowles of Plasma Channel in January 2021 of a Plasma Flame Generator. I loved the simplicity of the circuit (tuning and operation is a challenge, though), its unique output, and the fact that the resulting device is relatively small. In this design, a solid-state oscillator drives a primary coil which excites the resonator (secondary) coil, producing a high-frequency, continuous-wave output. The discharge produced by the Coil is a very interesting “flame discharge” resembling a candle flame. The Coil can be used to demonstrate wireless power transmission by lighting up neon globes and fluorescent lamps. In the lead photo, you can see a matrix I made using 100 neon lamps, sections of which light up when placed 62 Silicon Chip in proximity to a strong electromagnetic (EM) field (such as generated by this Tesla coil). Depending on the panel’s orientation, it can display the amplitude of the EM field or the relative shape. I think this is a really interesting way to observe such fields. The first thing you might think of looking at photos of this device is: “is it safe?” Yes, and no. It generates about 150kV, and given its operating frequency of around 10MHz, it can cause RF burns. Clearly, you need to be meticulous in building, testing and operating such a device. But we won’t tell you “don’t try this at home”. Still, we don’t recommend that beginners assemble such a device. It is more suitable for someone who, for example, has built several mains-powered devices and is used to the safety precautions involved Australia's electronics magazine in working with 230V AC. That’s because such people normally have the required mindset of ‘hands off when power is applied’, double-­ checking everything before switching power on and thoroughly insulating all high-voltage conductors. So without further ado, let’s get into it. Tesla Coils This Tesla Coil is based on a Class-E RF power amplifier that’s tuned to oscillate at around 10MHz. It drives a tap on an auto-transformer; the transformer’s secondary is excited by the oscillator to produce a high-frequency, continuous-wave output. You might be used to seeing Tesla Coils with a doughnut-like metal toroid on top, from which the discharge emanates. This one is simpler, with a dome instead, but it’s still a siliconchip.com.au This device generates hazardous voltages! Although the unit operates from a low-voltage DC supply, its high-voltage output will cause RF burns if you come close to or contact the discharge terminal, even when no discharge is apparent. The flame produced is a plasma, which is extremely hot and capable of melting copper wire (not to mention flesh!). Without the brass/stainless steel breakout point, it can begin to melt the wire at the discharge point. Always ensure that you are nowhere near the breakout point when powering the unit up. Keep all parts of your body (or anyone else’s) clear of it until power has been switched off and the discharge stops. And remember that a high voltage can still be present even when no discharge is visible. The potentiometer specified has a plastic shaft; use caution if substituting a pot with a metal shaft. At a minimum, you would need to use a plastic knob and ensure that the knob fully covers the shaft. For added safety, the coils (L2 & L3) and the breakout point can be encased in a 150mm diameter transparent plastic film or Perspex surround, with an open top 50mm higher than the breakout point. Electromagnetic interference warning This Tesla Coil is an RF generator. The input power can be up to 240W (48V <at> 5A) and the Class-E amplifier is very efficient, converting a considerable amount of input power to RF energy. That said, when breakout is occurring, most of that energy is converted into light and heat. Be aware that it can cause RF interference when operating, mainly in the HF (3-30MHz) band. That includes shortwave radio, multiple amateur radio bands, aviation and maritime communications and CB radio. The operating frequency of this unit is very close to the amateur 40m band, so be careful, or you might make some radio hams very unhappy! Tesla Coil (we’ll describe a larger and somewhat more complicated Tesla Coil with a toroid in a later article). The Tesla Coil is a loosely coupled resonant transformer invented by Nikola Tesla in 1899. It is capable of producing high-voltage, low-current, high-frequency alternating current. The voltages produced by Tesla Coils result from resonant voltage rise in the secondary and are not proportional to the turns ratio between primary and secondary windings as with traditional, tightly-coupled transformers. That allows exceptionally high voltages to be produced with a practical circuit; in some cases, over 1MV! The Tesla Coil comprises two L-C resonant tuned circuits. The primary tank circuit consists of the primary capacitor and a coil. The secondary coil (and often, high-voltage toroid) and the surrounding air form the secondary L-C circuit. The two circuits are connected in series and tuned to resonate at the same frequency for efficient energy transfer. The classical Tesla coil uses a spark gap arrangement to switch the energy stored in the primary capacitor into the primary coil. The energy in the primary circuit, moving back and forth between the capacitor and primary coil, transfers (couples) some of the energy to the secondary circuit. The voltage in the siliconchip.com.au The Tesla Coil when operating can produce a flame discharge which loosely resembles a candle. Care should be taken when operating the Coil as the flame produced is extremely hot and it produces very high voltages! Australia's electronics magazine February 2022  63 secondary continues to rise until the electrical field strength exceeds that of the insulating property of air surrounding the large surface areas of the top load and breaks out as an arc. Tesla coils can be scaled up to produce many millions of volts. Currently, the world’s largest Tesla coil is the “Electrum” designed by Eric Orr in New Zealand (see www.gibbsfarm. org.nz/orr.php) and built by Greg Leyh of Lightning on Demand (www. lod.org). Excitation methods The excitation methods for Tesla coils fall under three main types. Spark gap Tesla coil (SGTC) Includes static gap, triggered gap and rotary gap types. This type of excitation may also be referred to as “disruptive”. A high-voltage source is typically used. Solid-state Tesla coil (SSTC) Includes single resonant and dual resonant solid-state (DRSSTC) types. A DC power supply is used to charge the capacitor, with a power semiconductor such as a Mosfet or IGBT replacing the spark gap. Vacuum tube Tesla coil (VTTC) A similar topology to that used in radio transmitters. The main difference is that VTTCs operate in continuous-­ wave mode instead of the pulsed output of the previous excitation methods. The VTTC also requires a high-voltage supply such as specially configured microwave oven transformers. The Tesla Coil described in this article is interesting, as it falls within the solid-state coil (SSTC) category. However, it operates in continuous mode, not dissimilar to a VTTC, but at a much higher frequency of around 10MHz (rather than several hundred kHz to several MHz). We call this an HFSSTC. The main advantages of the HFSSTC are that it can be powered from a low-voltage DC supply, it doesn’t make much noise and you don’t need to deal with high-voltage primary power supplies. A continuous-wave coil operates at 100% duty cycle, resulting in silent operation. An interesting property of high-frequency, high-voltage output is its ability to produce a flame discharge, in which the ionised air (plasma) takes on the appearance of a candle flame. However, producing 64 Silicon Chip a stable flame is tricky and requires a fair bit of tuning. 5-10V signal at the gate of IRFP260N Mosfet Q1 to start the circuit oscillating. Feedback via capacitor C1 triggers and sustains the oscillation. The 4.7nF shunt capacitor and TVS diode provide some protection for the Mosfet; however, be prepared to lose a few Mosfets during testing and operation. ZD1 and TVS both aim to prevent the voltage at the gate from exceeding the gate-source voltage specification of the device, which is 20V. A 15V zener diode may also be used. L1 (10μH) is hand-wound with 24 turns of 0.5mm diameter enamelled copper wire on a cylindrical former. A 10μF capacitor is used for supply filtering, rated so that the circuit can be driven from a supply up to 63V (although 36-48V is sufficient). The primary coil (L2) consists of five turns of 1.32mm diameter enamelled copper wire wound on a 35mm high, 57mm diameter former. The resonator coil is installed inside the primary and is modular, so it can be easily removed. In my Coil, the 150pF and the primary inductance of 2.4μH gives a Circuit description As shown in Fig.1, the circuit uses a simple Class-E RF power amplifier to provide an RF drive current for the oscillator. This amplifier design dates back to the mid-1960s. Unlike a typical RF amplifier which drives a 50W resistive load, the Tesla Coil (secondary resonator) is a high-Q filter network. This type of circuit can achieve highly efficient switching using a Mosfet with zero-current switching (ZCS). This high efficiency is required to produce enough output power for a sustained discharge. ZCS means that the Mosfet is switched when the current flowing through it is at a minimum. The heart of the circuit is the LC oscillator formed by L2 (2.4μH) and C1 (150pF). The values of these components determine the oscillator’s frequency. In this case, around 10MHz (give or take). The voltage divider formed by VR1 and its 1kW series resistor generates a DOME COIL WINDING DETAILS L1: 24 TURNS OF 0.5mm DIAM. ECW ON A 22mm DIAM. FORMER L2: 5 TURNS OF 1.32mm DIAM. ECW ON A 57mm DIAM. FORMER L3 SECONDARY L3: 150 TURNS OF 0.5mm DIAM. ECW ON A 27mm DIAM. FORMER IRFP260N ZD1 A G K F1 12–63V DC (4A LIMITED) + – 10A D D PTC1 150pF q RXE250 10 m F 80V 4kV (C1) 1kW 2W ZD1 12V K A SC L2 2.4mH PRIMARY L1 1 0 m H 1W Ó2022 S D 1k W 2 W VR1 10kW G 4.7nF 2kV 0.5W 15V TVS Q1 IRFP260N S HF SOLID STATE TESLA COIL Fig.1: the circuit of the Solid-state Tesla coil is simple and elegant, with 150pF feedback capacitor C1 causing Mosfet Q1 to drive C1 and L2 at resonance. The inductances are chosen so that C1/L2 resonate at the same frequency as L3 and the stray capacitances around it (including the breakout point at its top). This results in extremely high voltages being efficiently generated at the top of L3, creating a flame discharge. Australia's electronics magazine siliconchip.com.au theoretical primary resonator frequency of approximately 8.34MHz. However, the interconnecting wires will increase inductance. The measured frequency of my oscillator is 7.42MHz, dropping slightly when the discharge is ignited, to 7.37MHz. The voltage rating on the 150pF capacitor needs to be a minimum of 4kV, so four 2kV capacitors are used in a series/parallel arrangement to double the voltage rating while maintaining the same capacitance. Mosfets have a fair bit of parasitic capacitance and non-zero switching time, and therefore ‘dislike’ operating at high frequencies. However, the use of zero-current switching (ZCS) operation helps in this respect. Secondary resonator The second resonant circuit is based around the secondary coil, L3. This develops a high voltage at the top of the Coil when it is excited at the same resonant frequency. The secondary comprises approximately 150 turns of 0.5mm diameter enamelled copper wire wound on a 25mm (ID) x 106mm tall PVC pipe former. An M4 x 12mm stainless steel bolt and a brass acorn nut is used as the breakout point or “top load”; it also influences the overall resonant frequency of the Coil. Another important reason for having this sort of discharge point is that the temperature produced by the discharge is enough to melt copper wire! Before constructing the secondary coil, I modelled the coil parameters in a Tesla Coil design software tool, “JavaTC” (shown below). This calculated the resonant frequency of the Coil and allowed me to make adjustments as required. Tuning Dealing with such a high frequency, it is surprising how minimal changes can affect the operation of the Coil. A slight tweak may mean that it doesn’t work at all, produces more of a corona discharge (rather than a flame) or blows the Mosfet. Tuning the Coil properly is therefore critical. I was fortunate enough that after I built my Coil, I managed to get it operating in the desired manner. But this was not without its challenges. Initially, I was cooking inductor L1. I was originally using a 12V SLA battery. I later learned that at a particular setting of the control potentiometer, there was a momentary current surge of more than 20A, which turned L1 into a fuse and it took the Mosfet with it. Therefore, I recommended using a current-limited supply to run the Coil. In case you still want to use a battery, I have added a PTC thermistor and fuse at the input of the final circuit, which will hopefully prevent damage under these conditions. Still, it’s best to use some form of supply current limiting if possible. In a pinch, this can be done with a wirewound series resistor of a few ohms, although that will reduce the overall efficiency of the circuit. Once you have achieved stable operation, tuning can be accomplished by adjusting the number of turns of the primary coil (L2), the interwinding spacing and its overall position (height) with reference to the secondary coil. The most significant effect that I found was the use of the stainless-steel bolt and acorn nut. This “top load” lowers the Coil’s resonant frequency, and adjusting its position has a significant effect. In my case, the final resonant frequency of the secondary is 8.12MHz. The calculated inductance for L3 is 168mH, which in theory should give a resonant frequency very close to 10MHz. It’s likely 20% lower than this due to stray capacitance. Input current limiting As mentioned earlier, I added the PTC ‘fuse’ (PTC1) because I found that it is possible to make the circuit draw so much power that it blows up the Mosfet and burns out L1. PTC1 goes high resistance if it conducts more than about 5A. Once you switch power off and let it cool, it should then work normally the next time. I have also added a 10A fast-blow fuse in case the PTC can’t act fast enough. There’s no guarantee that it will save the other components, but it’s cheap insurance. Neither of these components should do much other than provide peace of mind if you are using a 3.5A to 5A current-­limited supply as suggested. But I expect many people will not Output from the software JavaTC, which is used for designing Tesla coils. siliconchip.com.au Australia's electronics magazine February 2022  65 have such a supply. In theory, with this final circuit, you can power it from something like a battery that can supply many amps, and it should hopefully survive. The secondary coil was wound with the assistance of a hand drill, but it can be done by hand. Construction The first task is to prepare and wind the secondary resonator coil. The former is made from standard 25mm inner diameter PVC pipe available from Bunnings or any plumbing supply store. I cut it to a length of 106mm, which was based on my calculated winding data from JavaTC and allowed for extra material at each end for mounting. The outer diameter of the PVC tube is 26.9mm, and the winding itself is 82.2mm high. I gave the surface a light sanding, followed by a light coating with electrical-­grade varnish; however, this is not critical. As mentioned earlier, the secondary coil is wound using 0.5mm diameter enamelled copper wire, available from Jaycar, Cat WW4016 or Altronics, Cat W0405. The secondary coil can be wound by hand or with the assistance of a hand drill. Once finished, apply several coats of clear polyurethane varnish to seal the coil. Another option is “Ultimeg” electrical varnish, which I have used; it is available from Blackburn Electric Wires in Kingsgrove, NSW (see www.bew.com.au/varnish and also www.bew.com.au/wire). I built the base of the unit around a large heatsink, Jaycar Cat SY4085. As well as cooling the Mosfet, it’s heavy enough that the Coil won’t fall over easily. The central channel provides a space to mount the driving electronics. Also, it has flanges to act as feet, with holes to attach spacers for holding the upper structure. The base plate supporting the primary and secondary coils is made from an off-cut of 3mm FR-4 substrate (basically a PCB without copper). I obtained this from a transformer manufacturer in Wollongong, NSW but it can be purchased through Blackburn Electric Wires (see links above). Alternatively, you can also use an acrylic (Plexiglas/ PMMA) sheet. The heatsink needs holes to be drilled and tapped for the mounting points, as well as the Mosfet. I mounted the driving components on a cut piece of unclad, punched laminate, 56mm x 107mm. Silicon Chip has produced a PCB design to make assembly easier. I cut the board so that it fit snugly inside the heatsink channel. Our driver PCB is coded 26102221 and also measures 56 x 107mm. Mount the parts on that now, using the overlay diagram (Fig.2) as a guide to see which parts go where. The control potentiometer is mounted on a PCB measuring 56 x 30mm. This is mounted at 90° on the end of the main PCB using tinned copper wire braces to produce a robust mechanical support. L1 is a 10μH inductor. In my design, this is 24 turns of 0.5mm diameter enamelled copper wire on a length of 20mm diameter PVC pipe. However, I had to rewind this three times during initial testing due to it burning up. 0.5mm wire will not handle 20A, which I discovered during troubleshooting. However, after moving to a current-limited power supply, I have not had any problems with it. If doing it all over again, I would consider using larger diameter wire. To connect the base of the secondary back to the driver, I used a 2mm banana plug and socket so that I could remove and disconnect the secondary to work on the device. The connections to the Mosfet are terminated on the underside of the board (the solder side). The wires pass through holes drilled in the heatsink and are terminated to a three-pole pluggable screw terminal. The Mosfet is connected via the plug. I highly recommend this approach, as it is reasonably likely that you will blow up a Mosfet at some point during testing. Fig.2: we designed this driver board based on Flavio’s, which he made on a piece of unclad, punched FR4 fibreglass insulation. It’s pretty straightforward as there aren’t that many components, but we have kept the tracks well spaced apart to prevent arcing. 66 Silicon Chip Australia's electronics magazine siliconchip.com.au I also recommend purchasing a bulk quantity (eg, 10 pieces) to ensure you can continue to experiment. I glued the primary coil (L2) former and secondary (L3) plastic coupling to the FR-4 fibreglass base using two-part epoxy. I began to use the Loctite brand (see parts list) over Araldite and have not looked back. It works very well and is also cheaper. Mosfet choice I recommend using the IRFP260N Mosfet but I have also tested the IRFP460N. This is a 500V, 20A device (compared to 200V, 50A for the 260N). So far, it has been working well. In total, I have blown up three IRFP260N and two IRFP460N Mosfets and burnt out L1 twice in the process of building and experimenting with this device. Testing Before proceeding, make sure to keep your body away from the secondary coil at all times, especially the exposed metal at the top. This sort of voltage at such a high frequency can cause severe RF burns. Always power the unit up with the potentiometer would fully anti-clockwise. As mentioned earlier, the recommended power supply is a current-­ limited power supply delivering around 32V DC. 3.0-3.5A should be sufficient. You can test the unit initially without the secondary coil. Place a small neon lamp near the primary (not connected electrically) and power up the circuit. The electromagnetic field will cause the neon to light up if it is siliconchip.com.au While this Tesla Coil prototype was built on a veroboard, a manufactured PCB will be available from the Silicon Chip Online Shop. The finished board is then mounted comfortably inside the heatsink. The adjacent photo shows the mounting arrangement for the Mosfet, which is located on the other side of the heatsink underneath the main board. Australia's electronics magazine February 2022  67 Parts List – Tesla Coil 1 double-sided PCB coded 26102221, 56 x 107mm 1 double-sided PCB coded 26102222, 56 x 25.5mm 1 12-60V DC 3-8A current-limited supply 1 5A trip PTC thermistor (PTC1) [eg, RXE250] 2 M205 fuse clips (F1) 1 10A fast-blow ceramic fuse (F1) 1 heatsink with flanges [Jaycar SY4085 recommended] 1 plastic knob to suit potentiometer VR1 [Jaycar HK7010] 1 pair of red & black cables with inline bullet connectors [Jaycar WC6018] 1 2-way screw terminal with 5.08mm spacing (CON1) [Jaycar HM3172] 1 3-way vertical pluggable header [Jaycar HM3113, Altronics P2533] 1 3-way pluggable terminal block and vertical socket [Jaycar HM3113+HM3123, Altronics P2533+P2513] 1 120 x 100 x 3mm sheet of unclad PCB material (FR-4) or acrylic sheet (for coil base) 1 25mm length of 20mm inner diameter PVC pipe (former for L1) 1 35mm length of 55mm inner diameter PVC pipe (former for L2) 1 106mm length of 25mm inner diameter PVC pipe (former for L3) 1 25mm PVC coupling (to mount L3) 4 6mm-long untapped Nylon Spacers [Jaycar HP0930] 4 32mm-long untapped Nylon spacers (tap with M4 threads) [Jaycar HP0988] 4 M4 x 10mm Nylon machine screws [Jaycar HP0160] 4 4mm ID Nylon washers [Jaycar HP0166] 4 M4 x 10mm panhead machine screws 1 M3 x 10mm panhead machine screw and flat washer 1 M4 x 12mm stainless steel machine screw (for breakout point) 1 M4 brass acorn nut (for breakout point) 1 15m length of 0.5mm diameter enamelled copper wire (for winding L1 & L3) [Jaycar WW4016, Altronics W0405] 1 1m length of 1.3mm ◉ diameter enamelled copper wire (for winding L2) 1 150mm length of cable tie (for mounting L1) various lengths and colours of insulated hookup wire epoxy glue (Loctite brand recommended, available from Bunnings 1210127) clear polyurethane varnish (for coating the secondary coil) nail & paddle pop sticks (to make breakout starting tool) ◉ 1.25mm diameter ECW could be used, but some adjustments might need to be made to the design [Jaycar WW4024, Altronics W0409] Semiconductors 1 IRFP260N ▣ 200V 50A N-Channel Mosfet, TO-427AC (Q1) [Digi-Key IRFP260NPBF-ND, Mouser 942-IRFP260NPBF] 1 12V 1W zener diode (ZD1) [Jaycar ZR1412, Altronics Z0632, Digi-Key 1727-1946-1-ND, Mouser 512-1N4742A] 1 1.5KE15CA 15V 1500W transient voltage suppressor (TVS) [Digi-Key 1.5KE15CALFCT-ND, Mouser 603-1.5KE15CA/B] ▣ It’s a good idea to buy a few, so you have spares in case they fail during testing, the IRFP460N rated at 500V, 20A also works Capacitors 1 10μF 80V+ electrolytic [Jaycar RE6078, Digi-Key 493-4781-1-ND, Mouser 647-UCA2W100MHD1TO] 1 4.7nF 2kV plastic film [Digi-Key 399-12555-ND, Mouser 80-R73UN14704000J] 4 150pF 2kV plastic film [Digi-Key 1928-1172-ND, Mouser 505-FKP1150/2000/10] Resistors 2 1kΩ 2W * 5% [Digi-Key A138277CT-ND, Mouser 279-RR02J1K0TB] 1 10kΩ 24mm ½W potentiometer with plastic shaft (VR1) [Digi-Key 450D103-3-ND] * Increase the power rating for supply voltages greater than 48V 68 Silicon Chip Australia's electronics magazine oscillating correctly, as shown in the lead photo. Remember that you will need to wind the potentiometer clockwise a bit before anything happens. Power it down and place the secondary inside the primary. When powered back up, you may be able to observe a discharge. If you have a compact fluorescent lamp (CFL), bringing it near the secondary should cause it to light up, again due to the EM field. Operation I have found my Tesla Coil to have relatively stable performance. I am driving my Coil from a dedicated 48V 5A Mean Well switchmode power supply. To start the Coil, you slowly rotate the control pot until the circuit starts to pull current, then tap the acorn nut with an insulated metal tip. The Coil will not establish the discharge on its own; the arc must be established using a small metal tip quickly tapped on the acorn nut. I made a simple little tool from paddle-pop sticks and a nail for this purpose. The tool is simply made by sandwiching a nail between two paddle-­pop sticks, with the assembly held together by epoxy glue. For a nice touch, cover the sticks with heatshrink tubing. Start the breakout by turning the control pot to about halfway and tap the breakout point with the tool. One advantage of this approach is that it minimises the loading on the Coil, siliconchip.com.au which can cause the arc to go out. I was able to get a ‘flame’ just over 5cm long by supplying 32V DC at 3A (96W). If you have an oscilloscope, you can carefully probe the gate of the Mosfet to check the oscillation frequency. It should be around 7MHz. Scope 1 shows what you can expect to see when probing the Mosfet gate (in this case, during discharge). Note how the waveform is not a square wave or a sinewave. You might expect it to be a square wave, but there are all sorts of resonances plus parasitic capacitances and inductances in the system that conspire to make it look a bit messy. At this sort of frequency, Mosfet switch-on/off waveforms generally have edges that look like ramps with a step in them due to capacitive feedback within the Mosfet. So, a waveform like that shown in Scope 1 is not unusual for high-frequency switching. It is possible to run the Coil at higher voltages and power levels, up to 60V/8A. I recommend you experiment with care as it’s pretty easy to blow it up at high power levels. Experimentation One interesting experiment that can be performed is placing a tiny amount of elemental salt on the electrode. This will cause the flame to burn with vivid colours. I found that the best salt is simply a tiny amount of common sodium The coupling arrangement for the two inductors (L2 & L3) as viewed from the top of the Coil. Adding some sodium bicarbonate makes an especially interesting looking flame. bicarbonate (baking powder). This generates a very aggressive flame that is very yellow (Sodium-D lines). Finally, I would like to thank the engineers at Coast Electric Industries (http://coastelectrical.com.au) and Illawarra Transformers in Wollongong. They have helped me immensely with this and other related projects. You can download a copy from www. classictesla.com/java/javatc/javatc. html The theory of tuning a Tesla coil is covered at www.hvtesla.com/tuning. html (more so for classic coils, but it’s still relevant to measuring the secondary resonant frequency in this design). My website is www.nightlase. com.au and the page for this specific project is www.nightlase.com.au/ ?pg=hfsstc A video of my Tesla Coil working can be downloaded at: www.nightlase.com. SC au/?pg=hfsstc#HFSSTC-Videos References For more reading about Tesla coils, see https://w.wiki/4Mt6 JavaTC is an excellent and free piece of software used in Tesla Coil Design. A front view of the mounting arrangement of the Coil’s main circuit board gives a better perspective of how snug a fit it is in the heatsink. Scope 1: the waveform measured at the gate of Mosfet Q1 relative to ground. This is during discharge, and you can see the resonant frequency in this condition is 7.37MHz. The gate waveform is roughly trapezoidal; parasitic circuit capacitances (and especially those within Mosfet Q1) are pretty significant at this sort of frequency, so you can’t expect a clean-looking waveform. siliconchip.com.au Australia's electronics magazine February 2022  69