Winding flyback secondary?

Quote: Originally posted by Foeskes

So after winding 1 layer you just bring the wire back to the begining and wind it again?

Oh the rabbit hole of practical transformer design really goes deep. There are quite a few tips, tricks and aspects about winding these things....a practical art on its own that really deserves a deeper plunge. Especially if we dwell into the high frequency domain of ferrite core transformers for voltage conversions.
For example the struggle to ensure maximum coupling between the primary and secondary of a transformer. In reality there is always a portion of the windings inductance that is not coupled into the energy transfer network of the transformer and it acts as a loose parasitic inductance in the switching circuit: leakage inductance be the term.
This portion of the inductance will generate voltage spikes that destroy switching transistors or interfere with resonance characteristics and just can be a major pain to deal with.
I ran into the issue when trying to wind transformers for DC-DC voltage converters and learned the hard way that transformer design is of utmost importance to be able to build a reliable converter.
Depending on core geometry there are different elaborate schemes using parallel primary windings that are divided amongst and between the layers of the secondary windings to ensure better coupling and less leakage inductance.

E.g. below one can observe a toroidal push pull type ferrite core transformer that I built for conversion of 12V to 800V at about 40kHz switching frequency. It has eight parallel sections of primary divided over the entire secondary to ensure minimal leakage inductance:




It is quite an elaborate design and I do not even want to mention about the actual pain of winding and building that thing by hand
But that was a very successful example that I managed to devise after a series of failed designs. But not really something I would want to do again.

Quote: Originally posted by Foeskes

I have 0.1 mm magnet wire on hand is that too small?

From what I can tell, this wire should only have a resistance of two or three ohms per meter.

For a high voltage, low current (as in, fractions of an ampere, maybe a few amperes at most), high frequency (compact core, and only a few 10,000's of turns at most) transformer, I think that would be fine. It might depend, a little, on exactly what you're using it for. A real fly back circuit* (the ones that work similar to an old-fashioned induction coil, except with high frequency solid-state switching) would almost certainly work well with .1mm wire, provided you didn't nick the insulation while winding. A ZVS/Mazzilli will probably be good with that wire, although I don't know how much current and voltage those put out once an arc has struck.


@Markx, that's a serious coil! Especially considering how tedious it is to wind a toroidal core. It's an accomplishment to be admired, but not envied How many turns is that? What current could it output? 12v-800v push-pull? That wouldn't have happened to be the power supply to this railgun? Would it?

*Just a silly observation: there are four classes of devices sometimes called "flyback transformers."

1, The transformers from a TV, driven with a waveform (primary current ramps slowly, then is shut off fast) similar to that in a TV.

2. All kinds of transformers, driven by waveforms similar to the TV one. An example is this puny homemade blocking oscillator voltage converter, with 6 turns on the primary and feedback windings, and 120 on the output. The transistor is in the middle of the toroid, and it's "potted" in hot glue to protect it, and make it ugly AF: It operates on 1-3v, and makes 100ish volts as an output, enough for a tiny neon indicator bulb.

3. TV flyback transformers driven with a ZVS, H-bridge, half bridge, push-pull, you name it.

4. Any kind of transformer, driven with any of the drivers in 3 (but not 50/60Hz mains), or in other words almost any high-frequency power converter with a primary and secondary that was designed or built at home.

Just thought I'd share that. Better to reflect on it here, then to just say "a real flyback circuit" without explaining the "unreal" ones.

Edit: Flyback converters (with TV-like waveforms) are used in other appliances besides old TVs, but they got their name from their usage in television. When the primary current was abruptly blocked, the electron beam would "fly back" to one side of the screen, and the secondary winding would emit its high-voltage pulse. But I somehow found the other explanation ("you'll fly back if you touch it") amusing, in a dark sort of way.


[Edited on 4-6-2018 by Vomaturge]

Quote: Originally posted by Vomaturge

@Markx, that's a serious coil! Especially considering how tedious it is to wind a toroidal core. It's an accomplishment to be admired, but not envied How many turns is that? What current could it output? 12v-800v push-pull? That wouldn't have happened to be the power supply to this railgun? Would it? [Edited on 4-6-2018 by Vomaturge]

Effin' aye! That was a serious pain to assemble this monstrossity. Fortunately there were not too many turns on the thing....if I remember correctly the secondary was about 160-200 turns or the like. With this kind of transformers the direct ratios turns/voltage division do not tend to apply quite exactly any more. So it is a bit of a trial and error type of work to get the correct output voltage. But really, winding the secondary was the easiest part on that build. Now getting the primaries divided equally and all the sections soldered in the correct sequence....that was a challenge. Not a venture to be repeated in short !

It was not for a rail gun though....but in essence still a charging circuit for a high voltage DC cap bank that was used as a reservoir for an impulse generator. Damn that still sounds suspiciously like a rail gun though

This little core (42mm) could transfer in excess of 400W of power without saturation when the switching circuitry was tuned into sweet spot. Quite impressive, but of course the duty was quite humble at these power transfer rates or overheating of the core would become a problem. The IRF3205 based switching circuitry became barely warm as the leakage inductance was low and no voltage spikes disturbed the clockwork.

I recently bought one of these https://www.ebay.co.uk/itm/15KV-High-Voltage-Inverter-Genera...
There are also completely assembled/soldered kits)


The core is easily dis-assembled, different air gaps can be tried,
the primary is easily removed and re-wound ...
(a new tie-wrap required per modification)
(the spec. is 15kV but I dipped my secondary in polyurethane varnish by habit)
It could be used to ionise noble gasses etc.
It does light a cigarette.
A fun kit that runs off one lithium cell ... a moderately safe introduction to high voltages.
(by changing the resistor up to 12 Vdc can be used but I've not yet tried)


(short runs only, after about 10 seconds the heat sink gets too hot for my liking)

P.S.with just a 0.2" gap/arc I could smell ozone.

[Edited on 10-6-2018 by Sulaiman 2]

Quote: Originally posted by Foeskes

Any tips on insulating the layers? I tested the old one I wound and it arced over, turns out 2 layers of klapton wasn't enough maybe wiping some epoxy on every layers.

Quote: Originally posted by A Fresh Lunatic

You may all be interested in this small booklet from 1951, "Small AC Transformers" by A C Avery, ive scanned in and turned into a PDF. It was the definitive pocket guide when I was a Radio Ham in my teens, back in the early 15th century........ The detail it goes into in terms of how to calculate windings for transformers of all sorts is superb. Please feel free to download and copy. https://app.box.com/file/303845307237 [Edited on 11-7-2018 by A Fresh Lunatic]