Hi
i have searched the web for toroid designs for a year now.
and nobody ever says witch is best.
it turns out it is more art then science.
i personally don't believe in art so there must be a best way to wind a toroid.
now i made some examples of more common types of winding that i have seen on the net. randomly wound toroids are too much of an art so i made only symmetrical ones.
the examples aren't perfect so you will have to use some imagination.
all of them have windings using 2 wires(shoe laces for ease of winding),
the primary has total of 8 windings: that is 2 primary coils 4 windings each.
white color for one primary and white/red for the second
the toroid is used in a SMPS for a car amplifier, so it pumps 12V to whatever voltage necessary for the amp.
first question:
Should i keep the 1 and 2 primary wires close to each other or space them equally?
pic1:windings side-by-side
pic2:windings apart
second question:
Should i space the wires in each winding equally apart or side-by-side or should i twist them together(Litz wire)?
pic1:equally apart
pic2:side-by-side
pic3: litz wire
third question:
Should i start and finish the windings both on the outside (or inside) of the core, or can i start on the outside and finish on the inside?
pic1:starting and ending on the outide
pic2:starting outside, ending inside.
some examples that i made:
The last ones are wound one wire at a time and later you connect the wires to make windings. (say you have 4+4 primary windings, 5 wires for each you wind 10 wires 4 loops each end to end and then connect them to make 2 primary windings 5 wires each. the more wires the more ways you can connect them.)
any ideas and combinations of the ones i have brought out are welcome and if you have a different winding methods i'd like to see them as well.
also you can post how you have wound your toroids and bring out some pros and cons for that design.
i have searched the web for toroid designs for a year now.
and nobody ever says witch is best.
it turns out it is more art then science.
i personally don't believe in art so there must be a best way to wind a toroid.
now i made some examples of more common types of winding that i have seen on the net. randomly wound toroids are too much of an art so i made only symmetrical ones.
the examples aren't perfect so you will have to use some imagination.
all of them have windings using 2 wires(shoe laces for ease of winding),
the primary has total of 8 windings: that is 2 primary coils 4 windings each.
white color for one primary and white/red for the second
the toroid is used in a SMPS for a car amplifier, so it pumps 12V to whatever voltage necessary for the amp.
first question:
Should i keep the 1 and 2 primary wires close to each other or space them equally?
pic1:windings side-by-side
An externally hosted image should be here but it was not working when we last tested it.
pic2:windings apart
An externally hosted image should be here but it was not working when we last tested it.
second question:
Should i space the wires in each winding equally apart or side-by-side or should i twist them together(Litz wire)?
pic1:equally apart
An externally hosted image should be here but it was not working when we last tested it.
pic2:side-by-side
An externally hosted image should be here but it was not working when we last tested it.
pic3: litz wire
An externally hosted image should be here but it was not working when we last tested it.
third question:
Should i start and finish the windings both on the outside (or inside) of the core, or can i start on the outside and finish on the inside?
pic1:starting and ending on the outide
An externally hosted image should be here but it was not working when we last tested it.
pic2:starting outside, ending inside.
An externally hosted image should be here but it was not working when we last tested it.
some examples that i made:
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
The last ones are wound one wire at a time and later you connect the wires to make windings. (say you have 4+4 primary windings, 5 wires for each you wind 10 wires 4 loops each end to end and then connect them to make 2 primary windings 5 wires each. the more wires the more ways you can connect them.)
any ideas and combinations of the ones i have brought out are welcome and if you have a different winding methods i'd like to see them as well.
also you can post how you have wound your toroids and bring out some pros and cons for that design.
windings are a compromise between several competing/opposing optimizing "dimensions", leakage inductance vs parasitic C is one “opposing” pair - there is no universal optimum
core, circuit details are required to start prioritizing the trade offs
core, circuit details are required to start prioritizing the trade offs
I am new to electronics so i'd still need help.
i have looked for books and from the internet how to wind toroids. they all show how to chose the core and calculate the windings but none of them actually show how to wind them not to mention the different styles of winding and their pros and cons.
You guys should be the masters of this art and thats why i'm here, to learn from the best!
it would help if you show how you wind your toroids and comment why you did what and what are the pros and cons of this design.
also you could answer some questions and comment the different type of windings i have pictured in my first post.
i'm still learning electronics. i have made some simple projects before: mic-amps preamps, lin.power supplies. for those i had schematics, made changes to make it useful for me and made them. so i have learned to solder the parts, now i'd like to learn the science behind it..
for this purpose i decided to take on something more complicated.
i chose an old class E car amplifier. i asked the schematics from the manufacturer and started by learning what does what on there.
Here are the schematics:
http://kaur.planet.ee/savi/Picture 4.png
http://kaur.planet.ee/savi/Picture 5.png
i understand all of it but i got stuck with the transformer.
the supply is a SMPS that pumps out +-24 rail voltage
it uses SG2525A for the controller at around 70kHz.
ruff calculations say it has 2X4 primary windings and 2X11 secondary windings. now i haven't chosen the exact core so i haven't made more detailed calculations for it. before i'd like to know more about how they work.
so thats why im here.
Thanks!
i have looked for books and from the internet how to wind toroids. they all show how to chose the core and calculate the windings but none of them actually show how to wind them not to mention the different styles of winding and their pros and cons.
You guys should be the masters of this art and thats why i'm here, to learn from the best!
it would help if you show how you wind your toroids and comment why you did what and what are the pros and cons of this design.
also you could answer some questions and comment the different type of windings i have pictured in my first post.
i'm still learning electronics. i have made some simple projects before: mic-amps preamps, lin.power supplies. for those i had schematics, made changes to make it useful for me and made them. so i have learned to solder the parts, now i'd like to learn the science behind it..
for this purpose i decided to take on something more complicated.
i chose an old class E car amplifier. i asked the schematics from the manufacturer and started by learning what does what on there.
Here are the schematics:
http://kaur.planet.ee/savi/Picture 4.png
http://kaur.planet.ee/savi/Picture 5.png
i understand all of it but i got stuck with the transformer.
the supply is a SMPS that pumps out +-24 rail voltage
it uses SG2525A for the controller at around 70kHz.
ruff calculations say it has 2X4 primary windings and 2X11 secondary windings. now i haven't chosen the exact core so i haven't made more detailed calculations for it. before i'd like to know more about how they work.
so thats why im here.
Thanks!
I have to admit, never seen stranded wire used for this. Unfortunately, JCX is correct, "it depends". Transformers are one of the most complex devices.
Let me offer another approach: The 90% rule. Solve 90% of the problem at hand and move on to the next problem. This allows you to spend more time on the complete system instead of worrying about perfection in one stage. In other words, for your application, it probably does not matter. Do you get the voltage you require? If so, you are done. Move on to the filter stage. When you get all the first 90%-ers done, than go back and fix 90% of what you left on the first pass.
Let me offer another approach: The 90% rule. Solve 90% of the problem at hand and move on to the next problem. This allows you to spend more time on the complete system instead of worrying about perfection in one stage. In other words, for your application, it probably does not matter. Do you get the voltage you require? If so, you are done. Move on to the filter stage. When you get all the first 90%-ers done, than go back and fix 90% of what you left on the first pass.
stranded wire is used because of skin-effect that occurs at higher frequencies.
Skin effect - Wikipedia, the free encyclopedia quick overview of the problem.
i calculated for 8 amp transformer @ 70kHz idd have to use 10 d=0.5mm wires for one primary winding and 6 d=0.5mm wires for secondary windings.
at these high frequensies i can't use just a single d=1.5mm wire.
using mains 50 or 60 Hz it would not be a problem.
i am 90% sure all the rest works.. the inductors are the simple bit. as you said the transformer is a complex bit..
The transformer is the only thing that i have to learn from this project.
i study physics and electronics is just a hobby for me.
and yes i am a perfectionist. sure winding the the transformer randomly works, but thats not why i'm here.
Skin effect - Wikipedia, the free encyclopedia quick overview of the problem.
i calculated for 8 amp transformer @ 70kHz idd have to use 10 d=0.5mm wires for one primary winding and 6 d=0.5mm wires for secondary windings.
at these high frequensies i can't use just a single d=1.5mm wire.
using mains 50 or 60 Hz it would not be a problem.
i am 90% sure all the rest works.. the inductors are the simple bit. as you said the transformer is a complex bit..
The transformer is the only thing that i have to learn from this project.
i study physics and electronics is just a hobby for me.
and yes i am a perfectionist. sure winding the the transformer randomly works, but thats not why i'm here.
if that is a push pull converter. use your very first pic (side by side) to make sure core is magnetized symmetrically preventing core saturation. I always do it that way.
At 70K skin effect should not be enough of an issue to justify the cost. You do know that each strand needs to be individuality insulated to have the benefit you are seeking. It would be interesting to measure the difference in efficiency between the two. If you do, pleas publish.
As dj suggests, I would wind all the turns evenly around the core. Your middle picture I would not pick. Just checked the Mouser and Newark catalogs to see if there was a hint to be had from those in the business making converters. No help. They do not list HF transformers at all and from the pictures, most OEM switchers seem to be using e-cores.
I applaud you endeavor. I too like too understand the details. I have been at it long enough to know when enough is enough.
As dj suggests, I would wind all the turns evenly around the core. Your middle picture I would not pick. Just checked the Mouser and Newark catalogs to see if there was a hint to be had from those in the business making converters. No help. They do not list HF transformers at all and from the pictures, most OEM switchers seem to be using e-cores.
I applaud you endeavor. I too like too understand the details. I have been at it long enough to know when enough is enough.
Big Thanks to all who have taken the time to answer.
at 70kHz the skin-effect should be about 0.25 deep so using wire with a diameter over 0.5 is a little pointless..
made some quick calculation:
yes you are correct, 0.5mm diameter vs d=1mm wire, the difference is small. then again, the thinner wires react faster to current change, so in theory it should make a bit more power, but yeah.. that is not measurable i think..
cheapest magnet wire i can buy around here is sold by the kg. so it makes no difference in price. or i can buy twisted litz wire, now thats is really expensive stuff. it is made of d=~0.1 mm wires or smaller. all insulated and also sold by the kg..
as i can see nobody has ever come around with this question?
looks like i have to challenge you old timers and just try them all?
now that will get costly and also i don't have the necessary measuring equipment so it would not be this informative in other words great loss of money and time (and fingers).
i'm still interested, if nobody knows the answer to my questions, how do you make your own toroids?
?
at 70kHz the skin-effect should be about 0.25 deep so using wire with a diameter over 0.5 is a little pointless..
made some quick calculation:
An externally hosted image should be here but it was not working when we last tested it.
yes you are correct, 0.5mm diameter vs d=1mm wire, the difference is small. then again, the thinner wires react faster to current change, so in theory it should make a bit more power, but yeah.. that is not measurable i think..
cheapest magnet wire i can buy around here is sold by the kg. so it makes no difference in price. or i can buy twisted litz wire, now thats is really expensive stuff. it is made of d=~0.1 mm wires or smaller. all insulated and also sold by the kg..
as i can see nobody has ever come around with this question?
looks like i have to challenge you old timers and just try them all?
now that will get costly and also i don't have the necessary measuring equipment so it would not be this informative in other words great loss of money and time (and fingers).
i'm still interested, if nobody knows the answer to my questions, how do you make your own toroids?
?
Last edited:
I think you have discovered why there are no basic guidelines to your question, and why the SPICE model is lacking. It's too darn complicated!
I only made one set of torrid transformers. They were a moving coil step up. 1000 turns of 38 ga wire, by hand. Once was enough!
I only made one set of torrid transformers. They were a moving coil step up. 1000 turns of 38 ga wire, by hand. Once was enough!
- Status
- Not open for further replies.