Uploaded August 2025 | Updated November 2026, 3 hours ago
Going over a couple more of these builds for reference but the general foundation has remained the same of using the secondary current for feedback.
I can say that using the IRFP250N's I have seems to work decently enough to consider it worth building, but I'd rather spend time trying to tune the one that uses the 3205's. Getting them for like a dollar each makes it so I can risk blowing them up a few times or more but I doubt it will happen more than once. I noticed that in the circuit drawing it says to apply about 4v to the gates but I don't think that was actually the case with that circuit, probably more like 400mV or higher. At least, that's about what the voltage would read across the gates when the circuit is powered but not oscillating. You don't need to know a specific voltage though, you just start from the lowest value and slowly increase it til you have feedback or just a noticeable bump in current draw.
I need to figure out how to set my phone to record in a way where it picks up the arcs properly but you can see the general trend in the outputs. I'm relying on mainly a primary impedance and coupling to tune the circuit here where I'll probably do that first minus a series primary cap. When I think I've found a good tuning where it seems to provide solid feedback through interrupted and CW, where I can also pull any kind of arcs and I don't hear/see any feedback issues then I might try to start playing around with series cap value. I would basically try to set this value to where it would help me tune the output to a maximum at 30V, but find my way there through a gradual process of bringing the voltage higher and higher from about 15V at first. If I reach a point where I can't encourage better arcs at 24V, there's no point assuming it's going to get much better at 30V. That is to say, it might be better at 30V until I start trying to draw arcs to the hand for example. Often times I'll be able to find a ballpark tuning where I can do all the minor tweaks in that regard by just moving the primary coupling up or down slightly. If the fets seem to survive but start getting too hot at a certain tuning, I might have to start scoping it and assume they are seeing too many high voltage peaks and are on the edge of dying at higher input voltage. For these 55v rated fets I think in the past I tried to keep them below about a 75V max swing.
These are the fets I'll be using in the future for maybe something like 35V max input builds.
mouser.com/ProductDetail/Vishay-Siliconix/SUM110N10-09-E3?qs=xrZAeqbNnhgQVYiPAGvF3Q%3D%3D
Going over a couple more of these builds for reference but the general foundation has remained the same of using the secondary current for feedback.
I can say that using the IRFP250N's I have seems to work decently enough to consider it worth building, but I'd rather spend time trying to tune the one that uses the 3205's. Getting them for like a dollar each makes it so I can risk blowing them up a few times or more but I doubt it will happen more than once. I noticed that in the circuit drawing it says to apply about 4v to the gates but I don't think that was actually the case with that circuit, probably more like 400mV or higher. At least, that's about what the voltage would read across the gates when the circuit is powered but not oscillating. You don't need to know a specific voltage though, you just start from the lowest value and slowly increase it til you have feedback or just a noticeable bump in current draw.
I need to figure out how to set my phone to record in a way where it picks up the arcs properly but you can see the general trend in the outputs. I'm relying on mainly a primary impedance and coupling to tune the circuit here where I'll probably do that first minus a series primary cap. When I think I've found a good tuning where it seems to provide solid feedback through interrupted and CW, where I can also pull any kind of arcs and I don't hear/see any feedback issues then I might try to start playing around with series cap value. I would basically try to set this value to where it would help me tune the output to a maximum at 30V, but find my way there through a gradual process of bringing the voltage higher and higher from about 15V at first. If I reach a point where I can't encourage better arcs at 24V, there's no point assuming it's going to get much better at 30V. That is to say, it might be better at 30V until I start trying to draw arcs to the hand for example. Often times I'll be able to find a ballpark tuning where I can do all the minor tweaks in that regard by just moving the primary coupling up or down slightly. If the fets seem to survive but start getting too hot at a certain tuning, I might have to start scoping it and assume they are seeing too many high voltage peaks and are on the edge of dying at higher input voltage. For these 55v rated fets I think in the past I tried to keep them below about a 75V max swing.
These are the fets I'll be using in the future for maybe something like 35V max input builds.
mouser.com/ProductDetail/Vishay-Siliconix/SUM110N10-09-E3?qs=xrZAeqbNnhgQVYiPAGvF3Q%3D%3D










