How to Calculate Induced Voltage on Four-Coil Toroidal Transformer | DIY Guide

[Jdo300]

Hello All,

I was wondering if someone could help me figure out how to calculate the induced voltage across each of the four windings on a toroidal transformer that I am going to make for a project (see attached sketch).

The two black coils are the power in and out coils, and the two red coils are sense and feedback coils. Basically, I'm making a switching circuit that will pulse square waves into the primary side and step them up to higher voltage spikes on the secondary side. From there the spikes will travel through a series of coils and then back to the red feedback coil. I'm trying to make a circuit that will self-tune as it is running. The top trigger coil will pick up the feedback spikes and switch on the pulse input to send another pulse through the system. I am going to be winding this transformer myself around a ferrite core that is about 2.25" ID x 3.125" OD x 0.375" thick. I'll be using the magnet wire that you can get from RadioShack (comes in 22, 26, and 30 gauge) to wind the coils with. But I want to get a ballpark idea of how many turns to make for each of the four coils. In case this helps, I'll be pulsing 25V square waves into the primary and I want just enough voltage induced on the trigger coil to switch on the gate of the MOSFET that I'm using to pulse the primary (4V).

Could someone point me to some equations that can be used to figure this out?

Thank you,
Jason O

 

[berkeman]

I believe that what you are asking about is pulse transformers. They are optimized to transmit the wide range of frequencies involved in the input pulse waveform. This is as opposed to narrow-band transformers (like for datacomm) and power transformers (like for DC-DC converters).

I have a pretty good book, "Ferromagnetic Core Design & Application Handbook" by Doug DeMaw. Maybe see if your local technical library has a copy, or if there are similar books there about transformer design.

I also googled pulse transformer tutorial, and got a number of useful hits. Here's a short page with a few tips that you will need to keep in mind:

http://www.rhombus-ind.com/app-note/circuit.pdf

For example, you may want to wind your primary and secondary bifilar, to lower the leakage inductance that is going to limit your pulse transfer fidelity. To get an idea of the number of windings, the magnetizing inductance of the primary will need to be high enough so as not to load the input source too much. So the higher the frequencies involved, the fewer turns you will need, and the opposite is true as well. What is the output impedance of your source? What frequencies are you going to be using?

Another consideration will be saturation of the toroidal core. Since there are no gaps (like you would have in a pot core or EI, etc.), you are more at risk for saturating the core if you start pushing too much power through it with a lot of turns. Do you have the full magnetic specs on the core? Is it ferrite or powdered iron? Is this more of a power application, or a signal integrity application?

Oh, BTW, I'm not sure that your polarity dots are correct in your sketch. Remember that if you have one long coil with a dot at the start, and cut it in the middle, you've formed two coils, both with the dots on the beginning end. The dot for the first half is at the original start location, and the dot for the 2nd half is in the middle of the two coils, at the new start end for the 2nd half. Think about that mental image as you place the dots on your coils above. And it will probably be easier just to wind the coils all in the same direction anyway. You can just wind from the same shuttle, and cut the wire after each correct number of turns and keep going with the next winding.

 

[Jdo300]

HI Berkeman,

Thanks for all the great info about the transformer. One thing you mentioned really peaked my interest. When you spoke about using bifilar windings on the transformer to cut down the inductance on the coil and heighten the pulse fidelity. If your bifilar wind the coil, wouldn't that mostly cancel the magnetic field in the coil? So if that is the case, how do you transfer the energy from the primary to the secondary?

Thanks,
Jason O

 

[berkeman]

No, bifilar winding does not hurt the magnetic transfer of energy via the core flux. It just lowers the leakage inductance (and raises the primary-to-secondary winding-to-winding capacitance, which can be a bad thing).

Think about it -- whether you wind the two coils on separate parts of the core, or wind them together on top of each other, the core flux will be passing through them, and transfer energy. The bifilar part is just that you twist the two wires together first, and then wind the twisted pair around the core for the "bifilar" wind. The wires still go the same directions, but they are now more tightly coupled, so there is less leakage flux and hence lower leakage inductance. The higher winding-to-winding capacitance, and lower primary-to-secondary breakdown voltage are the typical reasons that bifilar windings are not used a lot. But in the case of the pulse transformer, as long as the lower breakdown voltage is not an issue, (and the extra cost of bifilar windings is not an issue), then it is probably a good design choice.