How to wind a toroid transformer?

[gary350]

I have a book with formula for winding EI type transformers, is there a book, information, formulas for winding toroid transformers?

I notice a lot of electronics these days have toroid transformers with a primary and secondry coil. I want to learn how this is done. For an EI transformer the magnetic goss, cross sectional area of the laminations, voltage on the primary determines inductive reactance and how many turns will be on the primary coil.

What determines how many turns are on a toroid transformers? What about formulas?

I opened up a converter the other day. After breaking open the plastic box that plugs into the wall I find a toroid with a primary and secondary winding. Several turns of wire for the primay coil connect to the 120 vac plug. The other coil on the toroid was the 9 volt coil. There did not appear to be a lot of wire turns on the toroid compaired to what you find on an EI transformer. The toioid transformers seem to be very small compaired to EI transformers.

I need to learn more. I might want to wind my own toroid transformer.

 

[Nigel Goodwin]

Almost all PSU transformers are switch-mode these days (and have been for a LONG time), for a toroidal 50/50Hz transformer the number of turns and ratios are similar to EI transformers. However, switch-mode ones are MUCH, MUCH smaller as they operate at much higher frequencies.

For normal 50/60Hz transformers there's not a lot of size difference from a toroid either, think of it as an EI with the corners chopped off

 

[camerart]

Hi,

Not to do with formulae, but more to to with winding. I've always been puzzled how Toroids can be wound by a machine?

Camerart.

 

[atferrari]

I've always been puzzled how Toroids can be wound by a machine? Camerart.

Me too.

 

[camerart]

Me too.

View attachment 91863

That's clever. I will have to study it more to see why the spool winds one way while the ring go the other way.

Camerart

 

[atferrari]

That's clever. I will have to study it more to see why the spool winds one way while the ring go the other way.

Camerart

To be honest I never managed to understand it 100% and I always wonder how is it used for mass production. I guess that setting the whole thing for the the next toroid should be not simple. Just guessing...

 

[camerart]

Digging the dusty corners of my memory.

I think the toroid could be split in half and joined later. Toroids are more efficient I think.

I think as with other transformers, the thickness of the primary coil and secondary wire needs to be calculated, then wind the calculated number of coils. The size of the toroid is probably from the need to be able to take the number of windings.

The toroid material also as an affect on efficiency etc.

Gary, you mention 120V, be careful! As mentioned, the low coil count could be due to it being switch mode and not actually 120V. There should be many coils if it is 'normal'

Camerart.

 

[MrAl]

Hello there,


I've only seen one real life toroid winding machine in my life. It's quite a complicated machine.
It winds the turns around the core as it rotates the core around, and the bumpers that hold the core are cushioned and they turn to make the core rotate. It's quite amazing to watch, maybe there is a video somewhere on the web.

To do it by hand, a 'shuttle' is used. First the total length of the wire is computed, knowing the core surface circumference and taking into account how that changes as the turns are added. This length of wire is then wound onto a small bobbin that can fit through the center of the toroid. The turns are then applied by pushing the bobbin through the center of the core repeatedly until the correct number of turns are applied.
A layer of tape would then be applied, then the next winding.
Bifilar windings are done the same way except two wire lengths are cut and wound onto the bobbin.
The bobbin is usually referred to as the "shuttle".

For sine waves the formula for the flux density due to the primary winding is:
B=E*10^8/(4.44*F*A*N)

and for square waves it is:
B=E*10^8/(4*F*A*N)

which is the same as for EI laminated cores.

Here B is flux density in Gauss, E is RMS voltage in volts, F is frequency in Hertz, A is core cross sectional area in square centimeters, N is the number of turns. Solve that for N to get the number of turns.

 

[BobW]

Lots of Youtube videos showing how they work:

 

[ChrisP58]

Micrometals has a design tool for their cores.

**broken link removed**

 

[atferrari]

Amidon sent me information, quite complete, let me say, printed in paper, by mail, some 15 months ago.

Everything about hteir products was there. Learnt a lot just by reading. Pity I cannot find it in my archives. Typical...