Uzzors2k Banner

High Voltage Transformer Winding
Roll your own!

Using flyback transformers in ways they were never intended to be used eventually leads to failure of the transformer. Fortunately the core almost always remains intact, and can be salvaged to make a new transformer. I've tried several times to make my own flybacks, but each attempt has been unsuccessful until now.  I finally seem to have developed a relatively easy method, which creates good HV resistant transformers. This is what must be kept in mind:

1) Insulate the layers with something stiff and HV resistant. I use overhead transparencies, which can stand some 14kV per sheet IIRC. Use good insulating tape to hold things in place, but do not overdo. It's important to leave enough space between the top and bottom of layers to prevent arc-overs between layers.
2) When terminating a layer, bring the wire back to the start of the winding before continuing on the next layer. Insulate as required, both before crossing back and after, before starting on the next layer. This reduces interwinding capacitance.
3) Submerge in oil or vacuum pot. This is to prevent corona from eating away at the insulation and eventually destroying the transformer. Oil is the easiest for a hobbyist to use, as it just takes a container and some oil. Cooking oils are supposed to be pretty good, and are actually used in some power transformers. I use mineral oil however, so I don't need to worry about it going rancid. Standard hydraulic oil has worked well for me.

Winding Jig  Trannie under oil

Use single strand wire for the primary, or you'll have problems with oil seeping out of the wire by capillary action!

That should have the most important points covered. As for the number of turns, base it on the size of the core, drive frequency and desired output voltage. The core size and frequency will dictate a minimum amount of primary turns, and the supply- and desired output voltage a turns ratio. The math behind this is the same as when winding GDTs, here's a link to the calculator. An important note: I used 500 secondary turns in my transformer and 14 primary turns. Driven from my half-bridge this should give 5kV out, but it gave 10kV. This is due self-resonance and parasitic L and C components causing ringing on the output. So you can almost count on twice the output voltage that turns ratio alone would predict. For my transformer I used a wimpy AC flyback core from a xfrmr which had previously arced-over. Given the overall size of the new secondary winding, a standard DC flyback core would have worked just as well. When picking a driving frequency it's important to note that the primary inductance is low (20-40µH), which results in a good portion of magnetizing current being switched by the inverter. Increasing the drive frequency reduces the magnetizing current. This can also be countered by using an external air core ballast inductor of some 20µH, and tuning for the parallel resonant frequency of the transformer. The parasitic capacitance on the secondary side is squared when seen at the primary, resulting in considerable parallel capacitance. The exact resonant frequency can be found by placing a small lightbulb in series with the primary, and tuning the frequency until a minimum in lightbulb brightness is found. Without a ballast inductor, the square-wave voltage-source inverter like any half- or full-bridge would suffer huge losses. (Remember that square waves consist of several sine wave harmonics, and each harmonic above the resonant frequency sees an increasingly smaller impedance.) With the ballast inductor however each rising harmonic sees an increasing inductive impedance, sparing the inverter.

Acr 1 Arc 2

Driven from my Multipurpose Inverter. Arcs start at 1cm.

Youtube Video!

The Big Mofo HV Transformer

I won this core for an excellent price on ebay, and originally I was planning on using it in a SLR inverter for charging large high voltage capacitor banks. However due to failure in getting SLR to work properly I have been forced to hard-switch the transformer for the time being. That doesn't prevent awesomeness however, and this transformer can supply some real power despite being run well below it's capabilities. It's specs are 15.2 cm tall x 9.3 cm long x 3.0cm wide, core material N27 and 880mm^2 of cross-sectional area. Quite the mammoth compared to the "little" flyback cores I'm used to seeing. Currently driven by the Multipurpose Inverter Mega (which is much too weak) at 120 - 145kHz and 100% duty. During initial tests with this transformer capacitive coupling to the core were causing large amounts of corona and eating away at the PVC secondary former. This puzzled me as the output voltage should only have been 10kV on each leg and the clearance seemed to be enough. I later learned that the open circuit voltage is up to FOUR times greater than predicted, due to unknown effects of stray reactive components and presumably parallel resonance. The design goal of this transformer became to power my CW Tower to create a 80kV DC supply, at substantial power. After some testing and many failed configurations I determined that a turns ratio of only 1:15 would provide roughly 20kV of output voltage. I opted for 15 primary turns and 225 secondary turns. The transformer was purposely wound to maximize leakage inductance which helps limit current when drawing arcs, so each winding was wound on separate limbs of the core. Given the tendency of corona formation I found myself forced to submerge the entire assembly under oil.

Litchenburg figures on secondary former  Setup

Lichtenberg figures on the secondary coils from corona damage. This was before oil submersion.

Transformer internals  Transformer in oil container

Core assembly. The 8 turns shown in the picture are from the testing phase. Unlike half-bridge drive where the voltage is twice what is expected, I found the voltage to be 4 times the expected value. The current configuration uses 15 primary turns, which fills the winding window better anyway.

Youtube Video! (Video was taken during 8-pri turns configuration)


Creative Commons License This work is licensed under a Creative Commons Attribution-Noncommercial-Share Alike 3.0 Unported License.

Disclaimer: I do not take responsibility for any injury, death, hurt ego, or other forms of personal damage which may result from recreating these experiments. Projects are merely presented as a source of inspiration, and should only be conducted by responsible individuals, or under the supervision of responsible individuals. It is your own life, so proceed at your own risk! All projects are for noncommercial use only.

807797 unique visitors since 28th July 2009.

Website layout by Elinor, and Eirik Taylor, 2016.