I was thinking of instead of just going back and forth while winding. (for example a normal transformer)
when you fill up the initial first layer of winding. instead of starting the next layer of winding where you left off.
you jump back to the top where the first winding was started. then continue on again.
therefore MORE than quadrupling the amount of angled magnetic field potential
since all the layers of wire are parallel
but not going across themselfs on the next layer diagonally.
Wouldn't that improve the amount of efficiency?
For the points where you jump straight back. you go 1 further ahead for each layer.
so it sort of forms a pattern going all the way around it. as to not go completely on top of each other and lose efficiency.
I feel like the angle/direction the wire is coiled around the core has an enormous impact on efficiency and strength of the magnetic field.
is there such a winding method like this? or something similar?
I think toroids are the easiest to accomplish this because its native to their design.
when you fill up the initial first layer of winding. instead of starting the next layer of winding where you left off.
you jump back to the top where the first winding was started. then continue on again.
therefore MORE than quadrupling the amount of angled magnetic field potential
since all the layers of wire are parallel
but not going across themselfs on the next layer diagonally.
Wouldn't that improve the amount of efficiency?
For the points where you jump straight back. you go 1 further ahead for each layer.
so it sort of forms a pattern going all the way around it. as to not go completely on top of each other and lose efficiency.
I feel like the angle/direction the wire is coiled around the core has an enormous impact on efficiency and strength of the magnetic field.
is there such a winding method like this? or something similar?
I think toroids are the easiest to accomplish this because its native to their design.
I don't understand what you are suggesting, but you can be reasonably certain that it won't work. People have been winding transformers for a long time, and they know what they are doing.
If you go back to the start of the first row you have at the start of each row a single winding at a different angle to the rest.
Z winding sounds close to what i'm talking about.
basically when you finish coiling the core to the other end of the transformer.
you move back to the start of the original winding position immediately.
instead of coiling back towards the other end of the core.
basically when you finish coiling the core to the other end of the transformer.
you move back to the start of the original winding position immediately.
instead of coiling back towards the other end of the core.
Won't that little bit of perpendicular wire crossing the coil look like a shorted turn added to the coil? What's to be gained?
i dont know. but each time you repeat it. you wouldn't cross at the same path over again.
if you use a toroid can you get around the wires crossing and just loop it around many times going the same direction basically doing the same thing?
that would mean a toroid transformer would be massively more efficient.. Well at least 50% more efficient. depends on the shape and number of layers of insulation on the coil and toroid. and thickness of wire.
would it be better to use many parallel windings of wire? or thicker wire?
if your making an inductor i'd think many turns of many thin wire in parallel. but thats not a transformer and off-topic
if you use a toroid can you get around the wires crossing and just loop it around many times going the same direction basically doing the same thing?
that would mean a toroid transformer would be massively more efficient.. Well at least 50% more efficient. depends on the shape and number of layers of insulation on the coil and toroid. and thickness of wire.
would it be better to use many parallel windings of wire? or thicker wire?
if your making an inductor i'd think many turns of many thin wire in parallel. but thats not a transformer and off-topic
Last edited:
A transformer core really doesn't know about this aspect of 'coil alignment' - the only thing that matters re. the core flux is the number of turns, not which part of the core that these turns' flux is impressed upon.
But, there is a capacitance between windings that is developed when you wrap a set of windings over another set of windings. This type of capacitance, within the same coil, ends up being a shunt capacitance across the coil as a whole, and it should affect the coil's self resonant frequency.
So, by winding from one side of a bobbin to the other side, and then moving back up to the top, will create one sort of shunt capacitance, whereas winding down to the bottom then skipping back to the top will end up with a different shunt capacitance.
By going 'down then back', the 'back' part of the winding will travel over wire that will have a more similar voltage than if you skipped back to the top of the bobbin where these windings will have more voltage drop than at the bobbin end. Since the signal voltage between windings is less, then the capacitance should be less, making the shunt capacitance lower if you just wind from one end of the bobbin to the other and go back and forth.
So, I guess skipping back to the top of the bobbin will affect the inter-winding capacitance, but in a way that should increase it - generally not a good thing.
If you're concerned about this, another idea you should contemplate is the idea of a 'multifilar' winding - what you mentioned when you talked about parallel windings. A multifilar winding uses several wires wound together on a bobbin to create several coupled windings, brought out of the transformer, and then connected externally to create one or more series windings, and even separate primaries and secondaries.
This winding idea was used for a number of famous output transformers, and while it creates a lot of capacitance between the primary and secondary, it provides tight coupling. There's also no reason why a primary could not be wound multifilar, with a completely separate secondary, either with an electrostatic shield to separate them, or by using two separate bobbins to provide I/O isolation.
One advantage of a multifilar winding is that by using two wires wound together, it's easy to connect those two windings as a center tapped coil, since they will have traversed the same part of the bobbin as they were wound, and will thus have very tight balance.
I'm not an expert, but generally, people did not wind a layer along a bobbin and then 'hiccup' back to the top of the bobbin to continue the next layer. The problem as I see it is that short winding jumping to the top could cause odd problems, by partially shorting the core. Simply traveling back up the bobbin to the top again was what people tended to do, and it actually has some electrical benefits re. inter-winding capacitance.
But, there is a capacitance between windings that is developed when you wrap a set of windings over another set of windings. This type of capacitance, within the same coil, ends up being a shunt capacitance across the coil as a whole, and it should affect the coil's self resonant frequency.
So, by winding from one side of a bobbin to the other side, and then moving back up to the top, will create one sort of shunt capacitance, whereas winding down to the bottom then skipping back to the top will end up with a different shunt capacitance.
By going 'down then back', the 'back' part of the winding will travel over wire that will have a more similar voltage than if you skipped back to the top of the bobbin where these windings will have more voltage drop than at the bobbin end. Since the signal voltage between windings is less, then the capacitance should be less, making the shunt capacitance lower if you just wind from one end of the bobbin to the other and go back and forth.
So, I guess skipping back to the top of the bobbin will affect the inter-winding capacitance, but in a way that should increase it - generally not a good thing.
If you're concerned about this, another idea you should contemplate is the idea of a 'multifilar' winding - what you mentioned when you talked about parallel windings. A multifilar winding uses several wires wound together on a bobbin to create several coupled windings, brought out of the transformer, and then connected externally to create one or more series windings, and even separate primaries and secondaries.
This winding idea was used for a number of famous output transformers, and while it creates a lot of capacitance between the primary and secondary, it provides tight coupling. There's also no reason why a primary could not be wound multifilar, with a completely separate secondary, either with an electrostatic shield to separate them, or by using two separate bobbins to provide I/O isolation.
One advantage of a multifilar winding is that by using two wires wound together, it's easy to connect those two windings as a center tapped coil, since they will have traversed the same part of the bobbin as they were wound, and will thus have very tight balance.
I'm not an expert, but generally, people did not wind a layer along a bobbin and then 'hiccup' back to the top of the bobbin to continue the next layer. The problem as I see it is that short winding jumping to the top could cause odd problems, by partially shorting the core. Simply traveling back up the bobbin to the top again was what people tended to do, and it actually has some electrical benefits re. inter-winding capacitance.
well heres a different idea. you could use extremely thin micro nano size wire.
and then parallel many many of them.
paralleling many sets of them on top of each other. since the wire is so thin. it can be very long except still in parallel.
so the impedance isnt too low. but it is very compact and has all the windings angled to the same direction.
That would work but require lots of work and would be expensive. if someone doesnt mind it being expensive and wants it anyways then would that work??
and then parallel many many of them.
paralleling many sets of them on top of each other. since the wire is so thin. it can be very long except still in parallel.
so the impedance isnt too low. but it is very compact and has all the windings angled to the same direction.
That would work but require lots of work and would be expensive. if someone doesnt mind it being expensive and wants it anyways then would that work??
Something hints to me that there is something about the way that transformers (and inductors) work which has escaped you. As I still don't understand what you are suggesting I am not in a position to diagnose your issue, but if you read post 9 very carefully it may help.
i worked for a power supply house where we had our own transformer division in house for over 25 years. I can tell you that very thin wire (# 40 0r higher) is EXTREMELY EXPENSIVE, Litz wire is even more expensive. It is prohibitively expensive to do this without a really good reason.
Transformer houses do what they do for really good reasons - second guess them at your peril.
Well extremely expensive wouldn't be a problem even if it was 20,000$ per transformer
its just a theory that it would be better because it would be more efficient with less losses to heat.
its just a theory that it would be better because it would be more efficient with less losses to heat.
No, a thinner wire will have a higher DC resistance per turn, and that is a direct energy loss with the same amount of flux. Your best bet is to use square or rectangular cross section wire to maximize the packing density, and use a tape wound toroidal core if you want efficiency. I'd imagine that whatever they use for modern electrical substation transformers would be the sweet spot of efficiency and cost - moving kilowatts all the time will make tiny efficiencies worthwhile. For the rest of us, not so much.
- Status
- Not open for further replies.