Core material for high power high frequency coil/solenoid

In summary, most ferrite cores designed for RF applications have a limited range of frequencies they are effective at. There are some core materials that can be used at very high frequencies, but they all have some saturation flux limit. It will be very hard to find a material that does not saturate at high fields.
  • #36
This is the fun part guys, I can understandably sense some irony @essenmein in your last post, the generator doesn't have any poles or pole pairs, it's not a synchronous or induction or any other type of commercial type of generator you would see in a power station.

I know this sounds lame but i thought some of you already knew what type of machine I'm building (trying to) based on all my threads here.
i'm using a conducting disc (Faraday disc) , due to Lorentz force the force felt by the electrons in the conducting disc is directly proportional to the rpm and B field strength so changing the B field strength at any time instant will also change the current/voltage amplitude in the disc, in other words the output directly follows the B field input which means that I'm getting out an amplified input signal, no poles no induction just Lorentz force.

Well induction would come in at other parts of the machine obviously since I'm using AC but the main generating part uses Lorentz force and Special relativity as it's theoretical working principle.
PS. back to the topic, so at Mhz frequencies my rotating surface should basically be very thin as using a thick one would be wasting metal as the B field would be stopped at the "skin depth" at the particular frequency ?

@Baluncore, maybe a stupid question but why do you think eddy currents would form because remember that my field is homogeneous and the conductor is fully immersed in it , so all rotating parts see the same field going in the same direction which should push electrons in a single perpendicular direction , current loops shouldn't form here , so that leaves me to the question would such a rotating conducting material in a homogeneous B field at 90 degree angle to it experience the so called skin depth or would the field be able to pass through , because the field at any given instant of time can be thought of as a DC B field if frozen in time while the disc spins perpendicularly to the field and due to Lorentz force sideways electron current arises.
well maybe I'm wrong just thinking..
Because as far as I know in order for there to be eddy currents opposing the applied B field the B field either needs to cut the same conductor in opposite directions on the same plane (as when dropping a magnet through a copper pipe) or the B field needs to only cut part of a conductor leaving other parts field free where return currents can form loops, but when the field only cuts the conductor once and everywhere is the same there shouldn't form loops of current instead there should be a single current direction driven by the lorentz force.
 
Last edited:
Engineering news on Phys.org
  • #37
artis said:
... but when the field only cuts the conductor once and everywhere is the same there shouldn't form loops of current instead there should be a single current direction driven by the lorentz force.
Is your applied B field AC or DC?
 
  • #38
AC , just that it is homogeneous , in other words covers the whole rotating area, so in theory all electrons being on the same radial distance from the center feel the same force and direction no matter where they are located on the disc, the flux of course is also everywhere in the same direction at any given time instant and when it reverses it reverses all together.as far as I'm aware the Lorentz force principle on a moving conductor cut by flux is the same both for static and time changing B fields.
 
  • #39
artis said:
AC , just that it is homogeneous ,
As I see it, the reversing B field induces a peripheral current in the disc.
The disc becomes a single shorted turn.
 
  • #40
I am not entirely sure what you mean by "reversing B field and peripheral current"
the disc is essentially a single turn , and given it's extremely low resistance we can consider it "shorted"
all that matters is that all around the disc there is at some point between the center of it and the circumference a field which is homogeneous so that electrons get accelerated from center to circumference or vice versa and so a potential develops between rim and center.So the question becomes in this scenario are there eddy currents ? To the best of my understanding there shouldn't be eddy currents not even with an AC B field , if this is true then what changes in the so called skin depth of the B field passing through the conductor in this case? Because as I understand this , this is different from an EM wave in a waveguide or wave reflecting on a sheet of foil where circulating currents form due to the field looping within the conductor, here we only have flux , B field lines cutting a conductor even though the field changes at high frequency but at any given time instant the flux through the disc is homogeneous and only in one direction, so electrons in the rotating disc get deflected altogether either one way or the other. Is this reasoning holding water?
 
  • #41
I believe what you are proposing is commonly called a hysteresis brake.
 
  • #42
The artis hypothetical model has excessive complexity in that not only is the B field alternating, but the disc is spinning. We can reduce that complexity since the B field alternates much faster than the disc rotates. What happens if the disc is stopped? It then comes down simply to the thickness of the disc compared to the skin effect depth at the B field frequency.

If the disc is thick, the B field will not be able to penetrate the conductive material because the induced ring current around the disc will cancel the incident field and so reflect the B field energy.

If the disc is thin, then induced currents during B field reversals, multiplied by the frequency of the reversals, will melt the disc.
 
  • #43
I feel that you have gotten this wrong, let me explain,
1) there aren't any disc rim/circumference currents around the edge of the disc because the whole edge is covered in the same flux going in the same direction as the middle of the disc, the return flux comes back outside the conducting material either loops through surrounding air or a higher permeability core material (depending on application/frequency etc)Secondly, I just set up a small experiment , I attached a brass disc with glue to an old brushless DC fan motor. I spun the disc , when i place a toroidal speaker magnet close to it there is no braking of the disc whatsoever, then i place a small neodymium magnet that covers only one small plane of the disc leaving the rest of the disc uncovered and now (due to eddy current forming loops creating an opposite field) the disc brakes down and spins much slower.
this is why the field homogeneity is key here, if the field is the same at each radial point in the disc all around it then the disc performs as if there would be no field at all, then the only braking effect can be made if the circumference is electrically connected to the center axis forming a current loop where Lorentz force can drive the electron current and this then becomes the Faraday disc. If the Faraday disc is placed within a proper homogeneous field there in theory should be no eddy currents forming in it.

So having this in mind , would the AC magnetic flux really stop at skin depth in this case or would it penetrate the copper as it normally would since the permeability of copper itself as metal is the same as that of air?
Because the way I see it the only reason why a changing flux can't penetrate metal with respect to static flux is because of the formation of eddy currents that oppose the incoming field, correct?
 
  • #44
artis said:
Because the way I see it the only reason why a changing flux can't penetrate metal with respect to static flux is because of the formation of eddy currents that oppose the incoming field, correct?
Yes, eddy currents oppose and delay changes in the magnetic field through conductors.

Mount your disc on a vertical axis that is parallel to the lines of the B field. Forget disc rotation. Now generate a 1 MHz current and use it to drive a single turn loop in the same plane as, but now just outside the disc. That creates your alternating B field. You clearly have an air-core transformer with the disc as the secondary, but that disc is a shorted turn.

Imagine lines of B, passing through a copper coin. As the B field increases each line will induce a current circulating around that line. The currents between adjacent B lines flow in opposite directions so they cancel. The only currents that do not cancel make up the peripheral ring current, your shorted turn.
 
Last edited:
  • #45
Ok , I visualized and now see your point, fair point.
Right I see that the disc (aka the shorted turn) against the B field is stationary in electrical terms even though it might rotate so it will get induced current both whether it stands still or rotates.

Ok one remedy might be to cut one or two thin and long slots in the radial direction so that they end at some minor radius close to center, this should make any large summed circumferential currents impossible and only radial Lorentz force currents possible, is this a sound reasoning?

but one question though, even without the slots with the coin analogy that you made, if the coin is fully within the flux so that even the circumference is covered , how can the circular current form because the field lines are everywhere the same and as you said yourself the small loop current that each line makes cancels due to the lines all having the same direction.?
If the middle would be empty like in a loop of wire the flux loops through the loop and each field line sums up to form current in the loop but here the disc is as you said shorted so the individual field lines can't make net effort as the individual small currents destroy one another as you said yourself ,
 
  • #46
artis said:
Ok one remedy might be to cut one or two thin and long slots in the radial direction so that they end at some minor radius close to center, this should make any large summed circumferential currents impossible and only radial Lorentz force currents possible, is this a sound reasoning?
If you break up the area of the disc with a tree of radial cuts then the field will propagate through the gaps at close to the speed of light. The field will reach the conductor faster from those cuts than it would without them. The disc can be much thicker if the thin cuts are separated by two skin depths. That is analogous to the way the laminations of a transformer function, they have a thin insulator between them that provides a path for the magnetic field from the windings to quickly reach the full surface area of the laminations. So that the entire volume of the core can be reached, the transformer laminations, or the particles in a ferrite are smaller than twice the skin depth at the operating frequency.

artis said:
but one question though, even without the slots with the coin analogy that you made, if the coin is fully within the flux so that even the circumference is covered , how can the circular current form because the field lines are everywhere the same and as you said yourself the small loop current that each line makes cancels due to the lines all having the same direction.?
The circular eddy currents in the conductor only cancel if they are next to another line that penetrates the conductor. The line inside the edge has an eddy current, but the line just outside the disc does not, so there is an imbalance that makes the peripheral current.

artis said:
If the middle would be empty like in a loop of wire the flux loops through the loop and each field line sums up to form current in the loop but here the disc is as you said shorted so the individual field lines can't make net effort as the individual small currents destroy one another as you said yourself,
The best conductors make the best magnetic mirrors = magnetic insulators. It does not matter if the disc is replaced by a ring, the peripheral current will still be the same in the ring. If the flux that passes through the ring changes, it will induce a loop current in the ring. The middle of your disc is irrelevant.
 
  • #47
Even though this does not directly relate to my case let me ask, are you saying that if I have a wire loop through which a certain amount of flux passes through that only the flux closest to the physical wire induces current in it while the flux that is at the center (furthest away ) contributes nothing to the induced current?Ok so judging by your statements I can conclude that it's only the circumference part of the disc that has the potential to create the circumferential current in the disc, so my slots shouldn't run far down the radial direction just say for a 30cm diameter disc they would have to be some 1/2cm long measuring from the rim? Because deeper down all lines are in the same direction and any current loops should be canceled before they can form ?Two skin depths at Mhz frequencies sounds like an awfully small number , I don't think it's practically possible to cut the circumference or any part of the disc wth slots that close to one another or am I getting something wrong here?
 
  • #48
artis said:
Even though this does not directly relate to my case let me ask, are you saying that if I have a wire loop through which a certain amount of flux passes through that only the flux closest to the physical wire induces current in it while the flux that is at the center (furthest away ) contributes nothing to the induced current?
No. All flux that passes through a loop induces current in the loop. Eddy currents flowing in a sheet conductor follow the periphery of the sheet.

artis said:
Ok so judging by your statements I can conclude that it's only the circumference part of the disc that has the potential to create the circumferential current in the disc, so my slots shouldn't run far down the radial direction just say for a 30cm diameter disc they would have to be some 1/2cm long measuring from the rim? Because deeper down all lines are in the same direction and any current loops should be canceled before they can form ?
Cutting short slots just moves the peripheral current inwards slightly. One slot at least must run to the disc centre and form the trunk of a radial tree of slots that block the induced eddy currents.

artis said:
Two skin depths at Mhz frequencies sounds like an awfully small number , I don't think it's practically possible to cut the circumference or any part of the disc wth slots that close to one another or am I getting something wrong here?
A numerically controlled 'Wire EDM' could fill a 6” (150mm) THICK copper disc full of 0.020” (0.5mm) wide slits. A UV laser could do the same to a much thinner disc.

I don't think you should be worrying about construction until you have actually modeled the device and shown that it would work. It is cheaper and easier to optimise the design in a simulator than in a workshop.
 
  • #49
Well I would want if possible only one radial slit that goes halfway or more to the center of the disc because having more of those would disturb the Lorentz current created in the disc because the current is not straight radial instead it bends like a snake from the center somewhat like a cork screw until it reaches the rim so having too many long radial slits would disturb it and cause losses.

Well sure we have fancy CNC laser cutting etc but the thing that worries me here is that the disc has to be spinning and at rather high rpm, I'm afraid it will simply not hold together and break by the centrifugal forces when it will be cut into the slots.
Why such a thick disc as 150mm or is that just an example? and what you mean by saying the disc full of 0.5mm slits, so the 0.5mm is the slit width but what is the distance between each adjacent slit and how far do you think they should run say percentage wise since disc diameter can vary. As in percentage with respect to disc radius.

Sure if it were just one vertical slit that goes through the radius (although that is also bad) but here , well maybe the best way to do it is to create multiple thinner discs each having two or three vertical slits that run say 2/3 radius length and then join the discs together in such a way that each next disc covers the previous disc in such a way as to join the slitted part so that no two discs have their slits lining up and creating a weak point. each disc could be laminated and then joined together by copper or other metal rivets under pressure, since all discs have the same radius the rivets could be conductive because the potential would be the same at that place on each disc. this is the best I could come up with thinking about it now.
 
  • #50
artis said:
Why such a thick disc as 150mm or is that just an example? and what you mean by saying the disc full of 0.5mm slits, so the 0.5mm is the slit width but what is the distance between each adjacent slit and how far do you think they should run say percentage wise since disc diameter can vary. As in percentage with respect to disc radius.
150mm is an example beyond your requirements.
Transformers have up to 10% insulation between laminations. You could have 10% cuts so no conductor was more than 5 cut widths away from air. Alternate cuts go deeper than others so as to cover the area with skin depth metal.

Many before you have discovered that their idea is impractical. There is a reason why you cannot buy one of these on the web. The problem here is not machining, but is to simulate a model design that will work, or to understand why it cannot.
 
  • #51
I'm not trying to blindly believe the idea is 100% practical, but I'm not willing to throw it away before I have fully understood the results of it, if not for anything else then at least for learning sake.

What do you mean by "cannot buy these on the web" , what is meant by "these" ?

Well laminating and cutting slits in copper sheet is indeed not a problem these days, in fact it's rather cheap.
On a second thought those rivets can't be made of conducting material because then they would electrically join the multiple isolated discs creating one large and so eddy currents would form again, but that can be solved.

So the skin depth at any frequency is determined by eddy current strength , so the weaker the currents the more into the material the flux can penetrate right? I wonder realistically how thick each sheet/disc could be made practically and how many of them can I stack together.
A stupid question maybe but say I stack one or two discs too many and the flux can't penetrate those last discs, where then does the flux go? Obviously the field lines must loop somewhere and enter back into the magnet that created them so if the area is blocked do they then go sideways parallel to the sheet of metal and then loop around or how?10% cuts for a disc that would be 360/10=36, so a slit after each 36 degrees or so, and one or two of them go deeper
 
  • #52
Ps. Since at high frequencies eddy currents form on the surface of the material then would stacking multiple isolated discs together be any more beneficial than simply taking one thicker disc and machining slits on the surface of the disc with some given depth while leaving the inner disc untouched?
 
  • #53
artis said:
What do you mean by "cannot buy these on the web" , what is meant by "these" ?
The device you are imagining does not now exist. That could be because it is impractical or impossible. You have not explained what you are really trying to do. I use the laws of physics to avoid wasted effort. You should study the field and do the same. Until then you are dreaming.
artis said:
Obviously the field lines must loop somewhere and enter back into the magnet that created them so if the area is blocked do they then go sideways parallel to the sheet of metal and then loop around or how?
The field lines initially run at the speed of light past the edge of the disc, or through the slots, then quickly pull into follow the entire surface. With time they diffuse into the conductor but only at about 100 m/sec. Any line that is still penetrating when the field reverses represents a waste of energy as the investment cannot be recovered. If material cannot be reached in the time available it represents a material, energy and a financial burden.
artis said:
10% cuts for a disc that would be 360/10=36, so a slit after each 36 degrees or so, and one or two of them go deeper
No. If the Wire EDM cuts were 0.25 mm wide then they would be separated by 0.25 mm * 10 = 2.5 mm. The maximum frequency is then determined by the penetration depth of 2.5 mm / 2 = 1.25 mm. For copper that will be about 5 kHz. Closer cuts reduce the skin depth and raise the possible operating frequency.

I don't think you understand the complexity of the field you are trying to enter. You are certainly not the first.
 
  • #54
trust me @Baluncore I do understand the complexity, just that I'm not so good at maths but nowhere am I blind to the multiple setbacks and dead-ends, I just like to figure things out, I have spent quite some time thinking this one through, right now the biggest problems are the field magnets and this one.

Those slits then are very close even at 5Khz , the disc rim would turn into a saw blade rather than a rim, I wonder how long radially those cuts would have to be measuring from the rim.

judging by what you said if I ever wanted to get above 1 Mhz and have low eddy's, my disc would have to be so thin and so slotted that it would resemble a piece of paper shot up by a shotgun , because conductive metals are not among the strongest ones and making them so thin and then slotting them would essentially render their mechanical properties useless I guess.
 
  • #55
I found a simulation here
https://www.cst.com/academia/examples/eddy-currents-on-copper-disc

it seems that this simulation shows accurately how a disc within a changing homogeneous B field would look like from the standpoint of eddy currents?

a stupid question maybe but still , how about instead of slotting the disc letting it have a larger radius so that the return flux from the magnet partly crosses the very outer part (circumference of the disc) since these return flux lines would have opposite direction through the disc they should impose eddy's that run in opposite direction within the circumference and the two currents running in opposite directions should cancel at least partly cancel one another?
Well we can have fewer slots in order to greatly save the mechanical integrity of the device and let the radius be longer so the return flux induces opposite eddy currents that cancel the original eddy current in each part or segment.

What do you think about this?
 
  • #56
That simulation is for a 4mm diameter Cu disc at 50 Hz, so the current has time to enter the material.
artis said:
What do you think about this?
Until I know what you are trying to do I can't really help you.
 
  • #57
eddy current question.png


I have a question, see this is a slotted disc but the circumference is solid (shown by the green) the B field lines penetrate only the slotted part of the disc, (shown by the blue dots) if the flux does not go through the small circumferential solid circle only through the larger portion of the disc that is slotted would then there be circumferential currents in the circumference?
I ask this because it would make the disc structurally better if I could keep the disc solid at center and circumference and only make slots in the main part of the disc where the flux will pass through,
 
  • #58
artis said:
... if the flux does not go through the small circumferential solid circle only through the larger portion of the disc that is slotted would then there be circumferential currents in the circumference?
Yes, a circumferential current will flow whenever there is a change to the magnetic flux passing through the shorted turn.
 
  • #59
Ok I see, no matter the geometry as long as there is a closed loop around a bunch of field lines there will be induced currents. I guess I could use this slotted geometry with a circumferential solid ending but I would then need atleast one break in the circumference also in order to destroy the loop.

Ok another question then, since the B field always goes both directions i wonder if the flux cuts the disc in one direction I could make the disc diameter larger and at least half of the return flux would cut the outer part of the disc in the opposite direction , now this should then induce eddy's in opposite directions canceling each other and the net result would be no or very little circumferential current, is this reasoning sound?
 
  • #60
artis said:
... now this should then induce eddy's in opposite directions canceling each other and the net result would be no or very little circumferential current, ...
Correct; but why have the disc if the lines loop back so effectively they do not pass through the disc.
 
  • #61
I'm not sure I understood your remark, well the field lines are going to loop back anyway whether there is a disc or isn't one I was just thinking that maybe I can put them to good use, although on a second thought if they experience another disc in their path that would in total lessen the strength of the field and I'm not sure whether the effect of cancelling circumferential currents would outweigh the extra mechanical complexity and other factors, probably not.
If I make a slotted disc I would have to use some low permeability (the same as air) composite material disc on which to attach the multiple thin and slotted copper fragments that join at the center for more structural integrity, this way I could sandwich the thin copper plate between the non-conductive discs to make one larger rotating structure.
 
  • #62
You need to specify exactly what you are trying to do. Many things are provably impossible. The design of magnetic machines is well defined. You need to shorten the magnetic field lines and you need to minimise the air gaps. You cannot get away with breaking the laws of physics by negotiating with the devil, or by getting a more expensive lawyer. Are you dreaming or engineering?
 
  • Like
Likes berkeman
Back
Top