Hybrid Tesla-Flyback Coils - Develop 10X+ More Voltage Than Predicted, Driven With ZVS

When I first started making these coils, I had not intended to be making a hybrid Tesla-Flyback coil system. I started winding a coil intended to go on a typical flyback core. I realized about 3 weeks ago that what I have been winding are multi-layer Tesla Coils with cores.

I saw a YouTube video that said I should wind a 600-wind secondary coil. Because I wanted higher voltage, I wound a 900-wind coil. And instead of randomly winding wire, I wound 3 layers of 6 mil LDPE (0.457 mm total thickness of LDPE) to insulate the core from arc-over, then I wound 150-winds of a single layer of 30 AWG enameled copper wire. I then wrapped another 3 layers of 6 mil LDPE and another layer of wire. After 6 layers of LDPE and wire, I put a final 18 mil of LDPE to insulate the outermost wire layer. When switching to a new layer, the wire will depart from the windings out to the edge of the LDPE, then back in to start the next row of windings. Keep track of where you put these, because you are going put high-voltage rubber tape over these so you don't get arcing into the core. I space them about 15 mm apart to keep them from arcing with each other. PVC tape will not work for this purpose. I will provide links, descriptions, and photos below for the necessary items in the parts section. I leave 17-25 mm of LDPE with no wire windings at the ends of each row of wire windings, to prevent arcing around the edges of the LDPE. I learned about using the LDPE from a YouTube video about making home-made high-voltage capacitors. 2 mil LDPE did not break down until ~20 kV. If you want a center tap or other taps on this coil, then simply turn your crossover points at the ends of each row, into a tap by making a larger loop of wire there. After I finish winding the coil, I solder a length of 18 AWG solid copper wire the the taps so that the finer wire of the coil is not subjected to mechanical damage from connecting to and unconnecting it from the screw terminal blocks.

I have built ~20 of these coils so far. I am presenting the best accumulation of the knowledge and experiences with these coils that I have up to this point in time. At the end of this Instructable, I included a photo-capture from a video I made of a coil made using PVC tape for insulation, arcing through to the core, then arcing from the core to the lower voltage output.

I have some large UU flyback transformer cores with 30 mm diameter cylinders where the coils are installed. For the coil formers, I used cardboard tubes out of the center of a wax-paper or aluminum foil rolls. Their ID is ~34 mm. This is ideal because you are going to have high-voltage tape folded over the edge from the outside of the final coil cylinder to the inside of the coil cylinder, and the tape is 0.762 mm thick and you will be using this tape in several places around the top and bottom of the coil cylinder. Do not use super-glue or any spray-on enamel, or any other spray polymer, with these coils. They will damage the LDPE, and, they are not necessary.

I use a coil winder I bought from Amazon. I really like this one. When I first used it, it was consistently miscounting winds. I wrote a message to the seller and they sent back a short series of button presses for me to make. That fixed the problem. I think it was an adjustment of a memory register of an internal gear-ratio that needed adjusting. I like having exact counts of the winds of wire because I want good resonance with these coils. The coil winder does not come with any rubber cones to mount the coil-former cylinders. I used refrigeration equipment mounts for these particular coils. For other rubber cone sizes, you can buy rubber stoppers for carboys or chemistry equipment on Amazon. I use stainless steel, M8 threaded rods, couplers, nuts, and washers to fit the coil-former cylinders to the coil winder.

Does anyone else wind coils like this? I have never seen any presented or described. Let me know in the comments section if you know of other coils like this.

Also, if you like this project, give it a heart. If you have a suggestion to make the project better, the presentation better, and/or if you find errors, leave a comment. Thanks.

I buy all my equipment and supplies from Amazon and eBay. Out of the hundreds of electronic components I have bought, only 1 item was defective (it was a pack of 50 gate drivers ordered from China on eBay), and the vendor refunded my money right away and I tossed the components in the trash. So it all works out to my satisfaction, better than 99% success rate.

1. UY1658 UU Shape Power Transformer ferrite core MnZn
2. Remington Industries 30SNS 30 AWG Magnet Wire, Enameled Copper Wire
3. Shaxon SO18-100RD Solid Copper Wire on Spool, 100-Feet, Red I like the white insulated wire better, the PVC insulation is more flexible than either the red or the green.
4. 69 kV Rubber Tape
5. Automatic coil winding machine 500W Industrial Winding Machine Digital Control Coil Winder
6. Clear Plastic Greenhouse Film, 6 mil Polyethylene
7. Screw Terminal Blocks, 600V 15A Dual Row
8. Goupchn High Voltage Oscilloscope Probe P4250 100:1 2KV 250MHz
9. HANMATEK DOS1104 Digital Oscilloscope with 4 Channels
10. Chanzon TVS Diodes 1.5KE200CA 1500W 200V Bidirectional
11. A set of high voltage (20 kV) capacitors 100 pF, 1 nF, 10 nF. These are the blue ones and are all available on eBay. I keep 20-50 of these on hand. I have burned up a few, and blown up a few.
12. DC12-30V ZVS Flyback Driver I really like this ZVS. It has lasted for over a year, and is still going.
13. Items for coil winder: Rubber Studs Shock Absorber M8 40x26mm, M8 60 mm coupler, M8 50 mm coupler, M8 40 mm coupler, M8 30 mm coupler, M8 lock nuts, M8 washers, M8 threaded rods. If you are going to wrap 2 inch coils on PVC pipe, you can order these #11 stoppers to mount your coil former.
14. Switches, 15A: SPST toggle, SPDT momentary on. I am going to change my power switch to momentary ON because if I get shocked, I want the switch to turn my device OFF automatically.
15. Cooling fan items: Fan for single MOSFET heat sink, Switch.
16. 12VDC power supply Of course, you can use a 12V car battery instead of a switched power supply.

Once you have wound your secondary coil, you need to insulate the crossovers and/or taps with high-voltage tape. I use 3/4 inch wide 3M Scotch Linerless Rubber Splicing Tape 130C. It is good to 69 kV. For the taps on this coil, I measured out 55 mm of tape, 20 mm goes inside the coil-former cylinder, 35 mm outside, with the tap in the center of the tape's width. Then I cut another ~35 mm piece of tape and lay it sticky-side to sticky-side to make a sandwich of wire in between the two pieces of tape. For the crossovers that are not taps, I put ~45 mm of tape over them, 20 mm on the inside of the coil former and 20 mm on the outside. The extra 5 mm covers the thickness of the copper windings and the LDPE insulation. I have done this procedure with PVC tape, and the coil wire arcs through it immediately (see last photo below).

Assemble the two pieces of the UU core with your new 900-wind coil in place. Put 1 layer of 24 mm wide masking tape around the free cylinder part of the transformer core. This will keep the two core pieces in place while you wrap your primary wire. I use 18 AWG solid copper wire. Start your winding from the center of the wire, your center tap. I make a total of 12 winds. Make all the winds in the same direction. I typically wind CCW. I hold the primary coil together with several zip ties. After the primary coil is complete, I put a zip tie around the entire core to keep every thing in place. I put zip ties at the top and bottom of the primary coil to keep the coil tight and in place.

Using hot-melt glue, I place 2 cardboard feet on the bottom of the transformer core. The cardboard is not corrugated. It came from a take-out cake plate that I cut into smaller pieces with my Dremel saw.

Finally, make a label for your secondary coil telling how many winds of what size wire, how many rows of wire, how much insulation you have between the rows, the ID, OD, and length of your coil-former tube, and the date you completed your coil construction. You will see an error in the label in my photos. The secondary coil has 6 rows, not 4.

I use a 15A switch for the +12 VDC line into the ZVS. I also put two fans on the heat sinks for the MOSFETs on the ZVS. The fans are not usually necessary if everything is tuned optimally. They MOSFETs in the ZVS are IRFP250N, 200V Drain to Source voltage. You can build your own ZVS, which I have done a couple dozens of times, or you can buy them off Amazon for ~$14 each. The way I run my current coil set is with 12 VDC from a 12 VDC power supply connected to the center tap of the primary coil, and each of the end taps of the primary coil are connected to the ZVS outputs. You can connect the ZVS using a ground wire connected to the center tap of the primary coil and let the ZVS MOSFETs supply the +12 VDC current to the primary coil. A third way you can connect the primary coil to the ZVS is by connecting one end of the primary coil to one ZVS output and the other end of the primary coil to the other ZVS output. I have a preference for the first way mentioned above.

You will need to tune the coil set with one or more capacitors across the inputs of the primary coil. I use trial and error (or trial and success) to find the best wave form for the coil set. You will probably need to connect a load, like a 20 W halogen bulb, across the secondary outputs, using both end taps of the secondary coil. On my current coil set, I have 30 nF of capacitance across the primary inputs for the best tuning. I use an oscilloscope to see the wave forms. The oscilloscope probe has a 22 AWG piece of copper wire connected to it, acting as an antenna, and I set it near the coil set. I don't ever connect my oscilloscope probes to the coil set, the voltage is way too high, and I have 2000V probes. Depending on how you use your coil set, you will probably have to adjust the capacitance across the ZVS inputs, somewhat.

Finally for the ZVS, I have two bidirectional TVS (Transient Voltage Suppressor) diodes connected to the ZVS outputs and each of those go to ground. These protect the ZVS from any transient voltage spikes. Ideally you should use 200 V TVS diodes. I currently have two 1.5KE200C TVS diodes in place. An even better choice is a set of 1.5KE200CA because they have a 5% tolerance. The suffix letters CA have meanings as follows: C means bidirectional, the A means 5% tolerance.

At present,I am very conservative and careful with my equipment. I have burned out a channel on my older oscilloscope. To measure voltage coming out of the secondary coil, I used taps at winds 750 and 900. That is 1/6 of the winds. I connected in series, 12X 10 Mohm resistors. I put the oscilloscope probe across 1 of the 12 resistors. I measured 264 VAC, peak to trough. The voltage coming in the primary is 12 V up and 12 V down, so 24 V over 12 winds, 2 V per wind. The expected voltage out should be (900 winds / 12 winds) (24 V) = 1800 VAC. Calculating the actual voltage out is (264 VAC / 1 resistor) (12 resistors) (6 rows) = 19.01 kV. That is 10.6X of what should be coming out of the coil set.

For me, this coil system has some distinct advantages over a standard Tesla air-core coil. They are:

1. Slower oscillation than coreless coil system - These hybrid coils oscillate between 3 and 24 kHz. I have had Tesla air-core coils oscillating at 4-25 MHz.
2. Magnetic field concentrated and conserved in core. Air-core coils lose their magnetic field as soon as the current stops flowing. There is no storage of the created magnetic field. The stored magnetic field in the core metals gets used when the current starts flowing the opposite direction.
3. Less cantankerous than a coreless coil, no constant tuning and troubleshooting.
4. Can use a ZVS instead of Slayer Exciter. Uses AC rather than pulsed DC.
5. Form-factor is much smaller than a comparable Tesla air-core coil.