Maker Pro
Maker Pro

Best way to wind flyback core?

Clarkdale44

Sep 27, 2014
4
Joined
Sep 27, 2014
Messages
4
Hello

I am building a fluorescent driver circuit that will be powered by 2n3055, i don't have a ferrite toroid but i have two flyback transformer's core(ferrite). I just need to know what is the best way to wind them?
I have previously winded small E type core transformer (Plastic bobbin and E type ferrite core outside), i winded it like first primary then over it feedback then over it secondary.
CFL_Driver_a.gif
The circuit above is just for the basic idea of how to wind the ferrite core..It may or may not be the circuit i will build.
IMG00020-20110228-1607.jpg
The core i have is like the one on the right side of the above image.
As i have got so far, there are two ways to wind a flyback core.

1. Wind it like normal regular transformer, coil above coil on the one side.
2. HV windings on the one side and LV on the other.
Are there any other ways to wind them?
I am quite new to electronics.


Thank you
 

Harald Kapp

Moderator
Moderator
Nov 17, 2011
13,700
Joined
Nov 17, 2011
Messages
13,700
Welcome to our forum.

I suggest:
1) Do not wind directly onto the core. The ferrite has rough edges which can cut into the insulation of the wire, thus creating short circuits. Use a suitable coil former. If you can't find one for your core, use at least a layer of insulating tape between core and windings.

2) Winding the coils on top of each other minimizes magnetic losses due to tight coupling. On the downside it increases capacitive coupling.
Winding coils on separate legs of the core minimizes capacitive coupling. On the downside magnetic coupling is not so good.
The choice depends imho on your preferences and possibly circumstances unknown to us.
For a good high-voltage separation I'd recommend winding on separate legs - distance is the best insulation. A small coupling capacitance also minimizes potentially troublesome EMC effects at high frequencies. The additional losses due to stray magnetic fields should be rather small and tolerable.
If you need very tight coupling, wind the coils above each other. Use a generous amount of insulation between the coils.

Note that you need to keep an area free from wire at the sides of the coils to keep creepage and air gap in check. See this schematic:
trafo-png.16034


Beware of dangerous voltages when working with this kind of circuit.
 

Attachments

  • Trafo.png
    Trafo.png
    10.5 KB · Views: 267

KrisBlueNZ

Sadly passed away in 2015
Nov 28, 2011
8,393
Joined
Nov 28, 2011
Messages
8,393
That blocking oscillator circuit relies on the saturation characteristics of the core. (Google blocking oscillator or look it up on Wikipedia for details.) Do you know the saturation characteristics of the cores you have? Do you know what grade of ferrite they are made from? Do you know the AL value of the core?
 

Clarkdale44

Sep 27, 2014
4
Joined
Sep 27, 2014
Messages
4
That blocking oscillator circuit relies on the saturation characteristics of the core. (Google blocking oscillator or look it up on Wikipedia for details.) Do you know the saturation characteristics of the cores you have? Do you know what grade of ferrite they are made from? Do you know the AL value of the core?


Well that's the issue, i don't know a thing about the core.. All i know is that its made of ferrite
 

Clarkdale44

Sep 27, 2014
4
Joined
Sep 27, 2014
Messages
4
Welcome to our forum.

I suggest:
1) Do not wind directly onto the core. The ferrite has rough edges which can cut into the insulation of the wire, thus creating short circuits. Use a suitable coil former. If you can't find one for your core, use at least a layer of insulating tape between core and windings.

2) Winding the coils on top of each other minimizes magnetic losses due to tight coupling. On the downside it increases capacitive coupling.
Winding coils on separate legs of the core minimizes capacitive coupling. On the downside magnetic coupling is not so good.
The choice depends imho on your preferences and possibly circumstances unknown to us.
For a good high-voltage separation I'd recommend winding on separate legs - distance is the best insulation. A small coupling capacitance also minimizes potentially troublesome EMC effects at high frequencies. The additional losses due to stray magnetic fields should be rather small and tolerable.
If you need very tight coupling, wind the coils above each other. Use a generous amount of insulation between the coils.

Note that you need to keep an area free from wire at the sides of the coils to keep creepage and air gap in check. See this schematic:
trafo-png.16034


Beware of dangerous voltages when working with this kind of circuit.


Thanks i will try winding the coils on top of each other. Also one more thing i wanted to ask, is this necessary that i wind primary and feedback beneath secondary? Can it be ok to wind it above secondary? Or does this has anything to do with which coil should be closer to the core??
Please explain..
I appreciate
 

KrisBlueNZ

Sadly passed away in 2015
Nov 28, 2011
8,393
Joined
Nov 28, 2011
Messages
8,393
If it's from a flyback transformer from a TV set, it's not intended to saturate. You can MAKE it saturate, but if it doesn't saturate sharply, you will probably waste a lot of energy during the period from when the core starts to saturate until the feedback voltage drops far enough for the transistor to star to turn off.

My suggestion is to drive the transistor with a signal generated by a separate oscillator - a 555 for example - instead of using the blocking oscillator with the feedback winding. Watch the primary current with an oscilloscope to see how long it takes before the core starts to saturate, and set the oscillator's ON time to somewhat less than that time, if you want maximum output power.

I'm also worried about the choice of transistor. The 2N3055 is not a very fast device, and it's only rated for 60V collector-emitter voltage. With a 1:10 turns ratio, that corresponds to an output voltage of 600V. How much voltage are you expecting to get out of this circuit? There are much more appropriate devices available. A MOSFET may also be a better solution than a bipolar transistor.

Re winding the primary first, I think there are several reasons for that. A good one is that you want to keep the high voltages physically a long way away from the core, I suppose. Also you want to minimise DC resistance in the primary so it's better if the primary turns are shorter.

This is not my specialty either. We have some other users here who have been experimenting with flyback converters; they maybe able to offer some better advice. And Steve seems to know about everything!
 

Clarkdale44

Sep 27, 2014
4
Joined
Sep 27, 2014
Messages
4
If it's from a flyback transformer from a TV set, it's not intended to saturate. You can MAKE it saturate, but if it doesn't saturate sharply, you will probably waste a lot of energy during the period from when the core starts to saturate until the feedback voltage drops far enough for the transistor to star to turn off.

My suggestion is to drive the transistor with a signal generated by a separate oscillator - a 555 for example - instead of using the blocking oscillator with the feedback winding. Watch the primary current with an oscilloscope to see how long it takes before the core starts to saturate, and set the oscillator's ON time to somewhat less than that time, if you want maximum output power.

I'm also worried about the choice of transistor. The 2N3055 is not a very fast device, and it's only rated for 60V collector-emitter voltage. With a 1:10 turns ratio, that corresponds to an output voltage of 600V. How much voltage are you expecting to get out of this circuit? There are much more appropriate devices available. A MOSFET may also be a better solution than a bipolar transistor.

Re winding the primary first, I think there are several reasons for that. A good one is that you want to keep the high voltages physically a long way away from the core, I suppose. Also you want to minimise DC resistance in the primary so it's better if the primary turns are shorter.

This is not my specialty either. We have some other users here who have been experimenting with flyback converters; they maybe able to offer some better advice. And Steve seems to know about everything!

Thanks for your explanation, i will consider your suggestion of using a 555 for driving the transistor. The circuit i have attracted here is not the one i will be building. I don't have oscilloscope. Anyway thanks for all your help. I think i found the answer what i was looking for.
 
Top