How to build a flyback transformer

[David lopez]

I am trying to build a flyback transformer. I have a ton of questions. For example can I use a
Nail as the core? If not I don't know
Why not. So why not? I have already studied the design on the internet. I still have tons of questions.

 

[DaveE]

Yes, you can use a nail as a transformer core. But it won't be a great design. Better, IMO, to buy a core which will work better in many ways.
It will be worth doing some research first if you aren't very familiar with this sort of design. It's hard to get it right by trial and error.
The first thing to know is that a flyback transformer works more like an inductor with multiple windings than like a conventional transformer. The difference is that the flyback stores lots of energy, while a normal transformer mostly transfers the energy immediately to the other windings.
You can search for "flyback converter design" to learn more.

edit: use a nail with iron it it, aluminum won't work.

 

[David lopez]

What kind of core would be better and
Why?

 

[NascentOxygen]

Once you have a workable transformer, what circuit will you try it in?

 

[berkeman]

I am trying to build a flyback transformer.

"Flyback Transformer" has a couple of different meanings. There is the flyback topology DC-DC converter circuit that is used in some power supplies. And there is the high-voltage generating flyback circuit used in CRT displays and other high-voltage applications. As @NascentOxygen asks, what is your end application?

I have already studied the design on the internet. I still have tons of questions.

Switching power supply design and the associated transformer design is an interesting and involved subject. When I get back to work on Monday, I can post links to a few of the books that I learned it from. You can probably learn a lot of it online now, but if you are reading how to try to use a nail for the core, you are reading the wrong websites, IMO.

 

[David lopez]

It's a dc-dc converter. And I hope to use it in a jacob's ladder.

 

[DaveE]

It's a dc-dc converter. And I hope to use it in a jacob's ladder.

A Jacobs ladder requires very high voltage and also reasonably high power. The transformer design for this HV really requires expert knowledge, even if you are using a voltage multiplier. You are likely to have a lot of component failures and not much success until you know more. The people that do know how to do this wouldn't consider using a nail for the transformer core.
Also not killing yourself in the process is a significant concern, and I, for one, am not going to help you do that.
Try making a low voltage flyback converter first.

 

[David lopez]

How do I make a low voltage flyback
Converter?

 

[DaveE]

How do I make a low voltage flyback
Converter?

Start by researching it on the web. Try searching for "flyback converter design". There is a lot of material available. However, you may need to get familiar with basic electronics first (depending on your background, of course). If you want to understand it and do it well, it's not simple. I would start with learning all about the difference between resistors, inductors, and capacitors first. Then I would learn about buck converters next. Then flyback converters.
This text is a great treatment of the subject, but it is a thorough college level text, so it may take a while to get through it.
https://eleccompengineering.files.wordpress.com/2015/01/fundamentals-of-power-electronics_2nd_erickson_full.pdf

 

[NascentOxygen]

I am trying to build a flyback transformer. I have a ton of questions.

Have you read through the Related Threads listed at the foot of this page? ⇩⇩

 

[arydberg]

First a transformer core is in a loop. a nail is not. Also the ferrates used on flyback transformers. are optimized for the frequencies involved. The steel in a nail is not the best material. I would suggest a transformer from eBay. and a circuit like. this one.
https://circuitswiring.com/10-30kv-tv-flyback-driver-with-2n3055/
Also having been shocked from color TV's it was not lethal but MUST be respected.

Also the Jacobs ladders i have seen were usual made from an old fashioned oil burner transformers. This output is probably lethal. Much care is required.

 

[DaveE]

Also having been shocked from color TV's it was not lethal but MUST be respected.

What I think you mean is that it didn't kill you that time.
I wouldn't be so cavalier about telling people that their HV circuit is safe; particularly when you don't know either their background knowledge or their specific implementation.
I also have been shocked a few times by circuits that might have killed me. Each time it was stupid and not something to dismiss lightly. My take away in each case was to learn not to ever do that again and to be more careful in general around this stuff.

 

[Shane Kennedy]

What kind of core would be better and
Why?

A flyback transformer is a high-frequency device. Iron, even iron powder, is not good for high-frequency. You need a gapped ferrite core if you want decent efficiency. They are off-the-shelf parts from many suppliers. It may be counter intuitive, but the energy is actually stored in the air-gap rather than the ferrite.

 

[Dr_Zinj]

What I think you mean is that it didn't kill you that time.
I wouldn't be so cavalier about telling people that their HV circuit is safe; particularly when you don't know either their background knowledge or their specific implementation.
I also have been shocked a few times by circuits that might have killed me. Each time it was stupid and not something to dismiss lightly. My take away in each case was to learn not to ever do that again and to be more careful in general around this stuff.

DaveE is correct. So much depends on the path of the electricity through your body; and it's never the same way twice. Next time it could cross the heart and disrupt, or even destroy, the nerves regulating your heartbeat. Either case usually causes death, either delayed, or instant.

If it doesn't kill you instantly, then get yourself to a hospital ASAP, even if you feel fine, and tell them you shocked yourself so they can monitor your heart and either confirm there wasn't any damage, or get treatment to fix it. Also, you might be really surprised to find out where a burn can pop up.

 

[Shane Kennedy]

The output voltage of a flyback transformer depends on four things 1) The turns ratio (assuming that you HAVE a primary and a secondary winding). 2) The breakdown voltage of the switching transistor. 3) The breakdown voltage of the windings. 4) The nature of the load.
Basically, the output voltage will rise until current flows. If the load is a rectifier and battery, it is the battery voltage which is the limiting factor. If the output is open circuit, the voltage will rise until something breaks down, be it the transistor, or the winding insulation, or sparkover at the closest point between the output terminals. Sometimes, capacitance will be enough to limit the output.

 

[Tom Rauji]

You are not going to be able to use a flyback style system to power a Jacob's ladder. The Jacob's ladder requires an almost continuous arc voltage to generate a steady arc-plasma that climbs with air convection from the plasma heating surrounding air.

A flyback or induction system is a "recoil" system that charges a magnetic core with flux from a current source, abruptly stopping current. Because an inductance tries to main current at a steady value, voltage across the flyback (or inductor) increases. It does this as the magnetic field collapses and the core tries to return energy to something in the system.

The core materials can be any type of magnetically "soft" iron material. It can be solid, laminated, powdered, ferrite, or whatever. It can be an open or closed magnetic path. The exact construction and materials depend on the pulse repetition rate, duration, discharge time, and the energy levels. Flyback or induction systems range from high voltage supplies including vehicle ignition systems to switch mode power supplies.

None of this matters, because what you probably want is a simple very high voltage transformer with limited current, so an arc is struck but then current is limited. This will make a square wave, with only a brief interruption at zero crossing. The plasma will stay during that brief zero crossing. I suppose you could build one with direct current, but it certainly would be far more complex.

It really would be best to just find a workable design using standard parts and copy it. I do this stuff for a living and I wouldn't try to build a suitable transformer. It would take me a hundred hours to design and build a transformer that could be purchased for a few dollars. A neon sign transformer or furnace igntion transformer would be a good place to start!

 

[Shane Kennedy]

Unless you actually want to build the transformer yourself, you can buy HV generators on eBay for a few $
As an aside, generally, as you increase the rep rate, the per Watt size of the transformer comes down. I am in the process of building an (3) isolated and interleaved flyback circuit, and unless I have seriously misinterpreted the specs, a 500W transformer can be less that a cubic inch, whereas a 500W 50/60Hz iron cored trafo would be something like 8inches cubed.

 

[DaveE]

You are not going to be able to use a flyback style system to power a Jacob's ladder. The Jacob's ladder requires an almost continuous arc voltage to generate a steady arc-plasma that climbs with air convection from the plasma heating surrounding air.

A flyback or induction system is a "recoil" system that charges a magnetic core with flux from a current source, abruptly stopping current.

Which is why almost every flyback smps has a low pass filter after the transformer. OK, wait, let's say almost every smps of any type has a LPF at the output. A constant output is a pretty common requirement; things like microprocessors and audio amplifiers want that too.

 

[berkeman]

I am in the process of building an (3) isolated and interleaved flyback circuit, and unless I have seriously misinterpreted the specs, a 500W transformer can be less that a cubic inch,

That's pretty impressive, if it works. I don't think I've seen that energy density in an SMPS transformer before. Even running around 1MHz, I'd expect a larger transformer to be needed. Can you publish your test results here when you get it working?

Here is a picture of an SMPS that advertises 500W total output:

https://www.aliexpress.com/item/32832089254.html
https://ae01.alicdn.com/kf/HTB1XMYR...-board-Dual-voltage-PSU-board.jpg_640x640.jpg

 

[DaveE]

Unless you actually want to build the transformer yourself, you can buy HV generators on eBay for a few $
As an aside, generally, as you increase the rep rate, the per Watt size of the transformer comes down. I am in the process of building an (3) isolated and interleaved flyback circuit, and unless I have seriously misinterpreted the specs, a 500W transformer can be less that a cubic inch, whereas a 500W 50/60Hz iron cored trafo would be something like 8inches cubed.

I've always liked the interleaved flyback designs for high power (mine was a 1kW, 1kV capacitor charger, back in the day). The input and output caps have much less stress and the distribution of current, heat, and mechanical loads on the PCB is nice at high power, compared to a single flyback. There comes a point when more stuff in a small space (xfmrs, MOSFETs, etc.) just doesn't pay off. We could also have N+1 redundancy without too much extra cost.
Flybacks are nice for capacitor charging because the high impedance energy conversion is a nice buffer between the low impedance of a voltage source and the low impedance load.
However, most high power designs opt for topologies with more efficient magnetics (i.e. bipolar flux, less energy storage) and lower stresses on the switching elements. Interleaving helps mitigate these high stresses by spreading the circuit out.
If you really want small in high power, then I like the full bridge phase-shift topology, where leakage inductance is your friend. These you wouldn't interleave or you'll end up with way too many switches.

 

[essenmein]

I always figured flyback were not really practical above a few hundred W, not because you can't make it work, but because as power comes up device losses catches up on you fast, double power is double current (assuming same voltage) means conducted losses are 4x. This is compounded by the large voltage stress it puts on the primary switch compared to other topologies, now you have poor rdson and higher current...

At some point switching to half or full bridge type topologies is better even given the extra complexity due to the voltages impressed on the device being lower and importantly clamped...

Maybe FETs getting better every other hour moves that point around a fair bit.

 

[DaveE]

I always figured flyback were not really practical above a few hundred W, not because you can't make it work, but because as power comes up device losses catches up on you fast, double power is double current (assuming same voltage) means conducted losses are 4x. This is compounded by the large voltage stress it puts on the primary switch compared to other topologies, now you have poor rdson and higher current...

At some point switching to half or full bridge type topologies is better even given the extra complexity due to the voltages impressed on the device being lower and importantly clamped...

Maybe FETs getting better every other hour moves that point around a fair bit.

Yes, I agree. That is the conventional wisdom. One of the biggest advantages of flybacks is they are cheap, but only at low power. Fewer parts under (relatively) higher stress is a good choice at low power. At high power the stresses matter more than component count.

 

[Tom Rauji]

This is going all over the place. He wants to build a Jacob's Ladder. His application requires many kilovolts (depending on spacing) and current limiting (dozens of milliamperes) for that voltage, and works much better with low frequency AC.

What works at low voltage high current dc isn't going to work at all in his application.

He needs to generate a plasma pretty much equally between two parallel conductors. Starting the arc will take somewhere up in the 20 kV range. Sustaining the arc without burning something up will require limiting current to a few dozen mA. The load is an almost infinite impedance until the plasma forms, and then once started becomes just hundreds or thousands of ohms.

He needs a source more like a furnace igniter or neon sign transformer. Something like this would not only be a better fit, he would also be less likely to disrupt broadcast or communications systems for miles.

...and good luck on that 1 cubic inch 500W SMPS transformer if it has any appreciable average power or duty cycle!

 

[DaveE]

What works at low voltage high current dc isn't going to work at all in his application.

As I said in post #7:
"Also not killing yourself in the process is a significant concern, and I, for one, am not going to help you do that.
Try making a low voltage flyback converter first."
Note that I didn't suggest a Jacob's Ladder. I opted for education where he might be successful.
There are a lot of people here who could build a Jacobs Ladder with one component purchased at a surplus store, but that doesn't mean we're going to help someone we've never met do it.

 

[Tom Rauji]

Any supply he builds for a Jacob's ladder is going to be a big risk. That's the nature of the beast.
It has to start an arc, generate plasma, and sustain that plasma.

 

[essenmein]

This is going all over the place. He wants to build a Jacob's Ladder. His application requires many kilovolts (depending on spacing) and current limiting (dozens of milliamperes) for that voltage, and works much better with low frequency AC.

What works at low voltage high current dc isn't going to work at all in his application.

He needs to generate a plasma pretty much equally between two parallel conductors. Starting the arc will take somewhere up in the 20 kV range. Sustaining the arc without burning something up will require limiting current to a few dozen mA. The load is an almost infinite impedance until the plasma forms, and then once started becomes just hundreds or thousands of ohms.

He needs a source more like a furnace igniter or neon sign transformer. Something like this would not only be a better fit, he would also be less likely to disrupt broadcast or communications systems for miles.

...and good luck on that 1 cubic inch 500W SMPS transformer if it has any appreciable average power or duty cycle!

Flyback is basically ideal for this application especially if they want to learn a little bit as the play (although a nail will never make a good transformer lol), flyback use to be used in CRT (waaaaaay back in the day) to generate all the crazy voltages needed to aim charged particles in vacuum!
https://en.wikipedia.org/wiki/Flyback_transformer

 

[sysprog]

I concur with the others who have said that the project you have said you are contemplating is way too dangerous, and quite possibly severely injurious and possibly lethal. It's also possibly illegal, as it's difficult to construct a Faraday cage to contain the RF interference effects, and the FCC takes a dim of that. If you haven't done so before, constructing a PA amplifier with an internal 10 watt full-wave rectified power supply might be instructive for you. To construct the device you've mentioned, you should at least be fully proficient, if not a full-blown expert.