Insanely Overdriven Flyback Transformer

In this build, I tested and eventually found resonance of this flyback transformer. This was achieved by achieved by placing the output of electrodes of the AC flyback far enough apart that all I could see was corona discharge coming off these wires. I tuned the variable frequency signal generator to achieve maximal corona discharge from these electrodes which were kept fixed distance while doing this. I then went beyond this point and the corona discharge started to die down. I then went the other way till it maximized again. I chose this point as the resonant frequency of the flyback transformer. This turned out to be 13.5 kHz which is in the audible range. The noise isn’t too bad and barely noticeable as the whole thing sits under oil. Initially when I ran this with a variable frequency generator the duty cycle was fixed at 50%. The problem was too much current was being drawn and was damaging the variable transformer. Now I’ve cut the duty cycle back to about 30%. I’m able to push a lot more through it in terms of voltage. Here I’m running around 45 V AC input, which is then rectified and stored in the bus capacitor. KEY Construction detail: The secondary former is made of snap together discs which basically have long slots and short slots cut on the outer edge of each PLA disc. The wire enters a long slot on the first disc and winds from the center out then it enters into a short slot on the second disc which is 180° opposite. Then it winds half a turn in the gap between the dics. It enters the long slot of the third disc which is 180° opposite after making this turn. Then it winds from center to outer to the outer edge and then it enters the short slot on the 4 th disc and into the second gap space then half a turn etc. There are a total of 14 discs that are printed so that they snap into each other. I also added epoxy glue to give it extra strength at the snap points. Basically I’m driving a full bridge of SKM 100 GB 123D Half Bridge modules using IXDN and IXDI non inverting and inverting gate driver chips with a push pull set up. I’m using a gate drive transformer for electronic isolation from the high voltage side of the circuit. I’m using a variable frequency variable pulse width signal generator at 5 V to drive the IXDN which then drives the full bridge. I’m using a 2 uF1000 V blocking capacitor on one of the legs of the full bridge outputs to the primary coil. The primary coil has 11 turns home made litz wire. The native resonant frequency of the flyback with its inherent capacitance is about 13.5 kHz which enabled me to get this kind of output. This was only a 60 V run and I think I could go beyond it which is gonna be my next demo. You are slightly more specific details on the IXDN driver chips and the gate, Dr. transformer: I’m driving full Bridge SKM 100 modules. These modules are great because they have a very low gate capacitance. The driver chips I’m using are low side inverting and non-inverting gate driver chips - they can drive 16 A up to 35 V. I’m feeding them 20 V. They are TTL compliant so I’m putting in 5 V signal into these. They’ll put out a good 20 V at 16 A. I’m using a custom-made board so that these chips can just be plugged into the board. I can get you the Gerber files for these custom-made boards if you are interested. I had them made through JRCPCB. I am using the push pull output of the driver chips to drive a gate drive transformer, which is wound with cat five cable. The 4 lighter colored wires in the cable are all paralleled up and function as the primary. The four solid colored wires function as secondary coils to drive the diagonals of the full bridge with one side on and one side off, just by inverting 2 of the wires. This seems to work very well. It also provides electronic isolation. The great thing about these heat sinked driver chips is that they can be bolted directly on to the same heatsink where the full bridge is located without the need for thermal pads as both driver chips share a common ground. The chips you want to get to do this are the IXDN614CI and the IXDI614CI chips, which are respectively non-inverting and inverting.