Future Challenges for Magnetics in Power Electronics

Sound On. Power On. Your Power Electronics podcast, powered by PCIM Europe.

Marco Jung:

Marco Jung: Hello everybody and welcome to the fifth episode of Sound On. Power On. Your Power Electronics podcast, powered by PCIM Europe. My name is Marco Jung and I'm a professor for E-mobility and electrical infrastructure at the Bonn-Rhein-Sieg University of Applied Science at Saint Augustin, as well as head of Department of Converter and Electrical Drives at the Fraunhofer Institute for Energy Economics and Energy System Technology at Kassel. Both are located in Germany.

Marco Jung: Additionally, I'm the chairman of the IEEE Joint IES/IAS/PELS German Chapter, an active member of several national and international communities and committees. And today, our technical theme is future challenges for magnetics in power electronics. And this, I will discuss with Professor Dr. Peter Zacharias from the University of Kassel.

Marco Jung: Hello, Peter. How are you?

Dr. Peter Zacharias:

Dr. Peter Zacharias: Hello. I'm fine. Had a good start of this year and so, I'm really fine. I'm really relaxed. And so, we can start.

Marco Jung:

Marco Jung: Peter, our life has crossed 12 years ago. Since this time, I know you as a lecturer, supervisor of my diploma thesis, second supervisor of my dissertation and now, as a colleague. But I think our listeners are not interested in my personal view of you. I think then it is more important or more interesting to them to know a little bit more about you and your power electronics life. Please tell us more about your several stations in the past and share your experience with us.

Dr. Peter Zacharias:

Dr. Peter Zacharias: The power electronic life, that's a good point. My name is Peter Zacharias. I was born in Torgau. That's in Saxony in the eastern part of Germany. And I studied electrical engineering at the university in Magdeburg, Otto-von-Guericke University. And I started with several applications of electrical energy for machining of materials and treatment of materials, such as electric discharge machining and electrochemical machining and laser applications, treatment and so on. And then development of electrical equipment for lasers at the university scale and then in Lambda Physik in Gottingen and in Kassel.

Dr. Peter Zacharias: And from this, I changed to another part of my power electronic life. This was the Institute for Solar Energy and Technology in Kassel. And after this, I joined Infineon in Warstein. Or better to say, at the beginning, it was EUPEC company and then it became part of Infineon. And after this, I joined the University of Kassel. And since 2005, now a full professor here for electrical power systems, which is specialty of power electronics. And my personal interest is magnetic devices. So, that's very shortly.

Marco Jung:

Marco Jung: Great, Peter. There are really a lot of stations in your life with power electronics and inductors. And this leads in our topic and I think Peter's love, future challenges for magnetics in power electronics. I think, in the past, many innovations were done for inductors and transformers, core materials, winding concepts, isolation materials, as well as conductors and core geometries. Can you tell us the state of the art, Peter?

Dr. Peter Zacharias:

Dr. Peter Zacharias: To tell the state of the art, I will think there is not enough time to give a real overview about the state of the art but to introduce the others a little bit more. When I started as a power electronic engineer, I realized very soon that the topologies are quite interesting. But the main success very often can be achieved if you find special solutions concerning the magnetics. And magnetics, you can find in all the power electronic topologies therefore converting energy from one type of energy into another type of energy.

Dr. Peter Zacharias: So, this was a driving force for me to get more involved into development of magnetic devices. And so, I was quite interested in making designs for chokes and then the transformers and for different types of materials, which can be chosen for the design. And so, for this reason, I was more and more interested. So, the state of the art is that we have converters, which have already achieved very high converting frequencies.

Dr. Peter Zacharias: The switching frequency of the semiconductor switches are in the range of several 10 kilohertz up to a few megahertz. And if you go down to the power, to the converted power, you find even more than 10 megahertz, for the converting frequency. If you go up to the power, to the converted power, into the kilowatt or 100-kilowatt range, then the frequency comes down and you will find more and more bottleneck for the conversion. And the bottleneck for the power density, which can be achieved, they are based on some problems in magnetic devices or in inductive devices.

Dr. Peter Zacharias: And so, it is very interesting to deal with these questions if you want to increase the power density of converters. It's very rough overview what are the problems there.

Marco Jung:

Marco Jung: Great. Can you give us a little bit more deeper insight where the problems come from magnetics?

Dr. Peter Zacharias:

Dr. Peter Zacharias: In the past, in all these discussions, the opinion was that the basic problem comes from the core, from the magnetic core materials, because if you increase at a certain flux density, if you increase then the frequency, then the loss density inside the core gets higher and higher. And so, sooner or later, you achieve a value that cannot be cooled. You cannot get out the heat of the core volume. And now, we have achieved so high switching frequencies that even the problems come not any more from the core material but from the conducting material.

Dr. Peter Zacharias: I think everyone knows the skin effect, that if you have high frequency current that will flow thinner and thinner skin at the surface of the conductor, if you increase the frequency. But this is not the only problem what you have. The conductor sees not only its own magnetic field but also, the magnetic field of the neighbor wires. And this causes also some lossy currents, eddy currents, and this is called the proximity effect.

Dr. Peter Zacharias: And if you have these conductors near some air gaps and you have very intense magnetic fields, which are crossing the conductors and also causing the eddy currents which are lossy and so, you have much more loss power density in the winding than you can handle. So, in principle, it is very, very difficult to exceed at high power switching frequencies of 100 kilohertz. Why? Because if you have high power, you have high current. And if you have high current, you need high cross or large cross-sections of the conductors. And if you have large cross-sections, then the skin effects and all this eddy current stuff plays a major role, what I explained before.

Marco Jung:

Marco Jung: Great. Now, I understand. We need new concepts especially in conductors. So, this means wires. Have you some ideas for us how they can look like and in which direction we must go?

Dr. Peter Zacharias:

Dr. Peter Zacharias: So, some ideas, I really do have. One idea is related to high power application because high power application require quite large core sizes. And if you have large core sizes, you have also large cross-sections. And if you have large cross-sections, that means every cross-section part with a certain length has its own capacity. So, all this electromagnetic stuff is linked or has linked problems between electric field and the magnetic field.

Dr. Peter Zacharias: In principle, every part of core represents itself also an oscillating or an oscillator circuit. And that means you have an inductive part, an inductance, which is coupled with a capacitor, and which have its own resonance frequency. And if the parts are quite large, then the resonance frequency is low. And so, you can observe the problem that with high sizes or big sizes of the core, the internal resonance frequencies, they decrease. And if this is near the operating frequency, you will have some problems.

Dr. Peter Zacharias: And so an idea could be to make some sections of the core, which are operated in parallel, so that every core part has its own high resonant frequency and so, you can easily handle this problem. This is a first answer to your question. And the second answer could be concerning the wire. If you want to operate at really high frequency, then the proximity effects and the skin effect can be lowered. Or this force you that only at the surface or near the surface, you can find the current lighter conductor.

Dr. Peter Zacharias: And so, why to use the whole cross-section of the wire, to use maybe Litz wire, which is quite common. But also, I would expect if you go to really high operating frequencies, like let's say more than 10 megahertz and maybe a combined exterior winding material which is made of some plastic core surrounded by a metallic surface, and could be an option to operate at first the device. And second, to have a very small weight because copper is quite heavy if you use it as conducting material.

Marco Jung:

Marco Jung: So, I understand that we need also a changing of the core or the core strategy. And in the past, I read some papers about the idea using coreless transformer and switch DC/DC converters for higher power applications. Do you see some possibilities in this case?

Dr. Peter Zacharias:

Dr. Peter Zacharias: I expect these coreless transformers only in the lower power range because in the high-power range, you need quite large cross-sections. And if you have such coreless transformers, the magnetizing inductance is not that big. So, you have always some stray field, which is crossing the windings and also due to any currents, you will observe some heated-up winding. So, this is not really that important if you go to lower power. And so, I think this will be up to several 10 watts an option in the low-power range.

Dr. Peter Zacharias: But it is true if you go to that high frequencies, as I mentioned like more than 10 megahertz, of course, the maximum operating flux density for the alternating flux is that small. That arise the question why should I use the core, which is a concentrating part for the magnetic field? If I have to go down to that small values for the flux current, it's much more convenient, I guess, to drop the core and to use only coreless transformers or even coreless chokes.

Dr. Peter Zacharias: But in these cases, you have very strong magnetic field which is crossing the winding and causes losses. You have no core losses anymore, but you have additional losses in the winding. So, there's really no universal solution. And this is a thing, which interests me and this challenges me to find new solutions. And so, I am dealing with all these problems since more than 40 years, and there are still new things, which you are challenged with. And so, it's quite exciting.

Marco Jung:

Marco Jung: So, that means for the really high switching frequency, it may be a good idea to use coreless transformers, but we have some increase in losses in the winding. But I think if we have a look in, let's say, higher power applications and then with higher switching frequency using silicon carbide, so that means more than 100 kilohertz. So, I think we need an improvement in the core materials too. So, can we expect a revolution in core materials?

Marco Jung: I think one key issue would be improving the possible magnetic flux density and reduction of the core losses. I think material scientists and electrical engineers must learn to speak the same language more and more. My experience is they are still speaking different languages in the development part of new core materials. What do you think?

Dr. Peter Zacharias:

Dr. Peter Zacharias: I support your opinion that the language of power electronic developers and material developers are different. And so, the understanding can be really improved. So, I am very lucky that we have also since many years a lot of projects together with industrial companies where we combine all these knowledges and where we present the results of the different working groups to each other, so everybody can learn from each other.

Marco Jung:

Marco Jung: Can we expect a revolution in core materials?

Dr. Peter Zacharias:

Dr. Peter Zacharias: You can, every time, believe in some revolution but maybe there is coming up something. But I would expect in material design, in combinations of different materials, I would expect some very interesting changes in properties of those magnetic materials, let's say, another mixture of existing materials. For example, we used different core materials operating in parallel to extend the frequency, the operating frequency of magnetizing core in a transformer. This is especially in transformers, one can use it.

Dr. Peter Zacharias: If you, for example, combine manganium zinc ferrites with nickel zinc ferrites, you can, for example, extend the operating frequency from two, three megahertz, which is really the maximum which can be used for the most of the power ferrites based on manganium zinc. And you can extend this to 10 or 15 or even to 20 megahertz. So, this is an approach which require not new materials but a new design of the cores.

Marco Jung:

Marco Jung: So, we had spoken a lot of, let's say, the material area, the inductor area, but let's come back to the system level. I think for DC/DC converters concept ideas with magnetic wires allows new material reduction possibilities. Is it correct? What are your experience?

Dr. Peter Zacharias:

Dr. Peter Zacharias: Yes, we are making experiments not only with different designs of magnetic devices or inductive devices, but also with different topologies. And one of our emphases is at the moment the possibility to separate, for example, the DC flux from AC flux. And if this is possible, then you can optimize the different devices, which come out of that concept. For the special application, one which has mostly DC flux and the other which has mostly AC flux seems to be very successful. And this might be possible or gives us the possibility to decrease the size of the single devices, which are components of the whole system.

Dr. Peter Zacharias: That is a concept what we have in one direction concerning changing the system for a special application. And this can be used for DC/DC converters but also, for DC/AC converters, if you have in mind that the AC converters are dedicated for the feed end of electric power into the public grid. And 50 hertz is almost, if you look on the magnetic devices of area devices, if you look on these, 50 hertz are almost like DC for them. And so, it is in principle for a high frequency combination with low frequency.

Marco Jung:

Marco Jung: And you said, okay, it's possible to have higher power density. It is a possibility to increase efficiency too?

Dr. Peter Zacharias:

Dr. Peter Zacharias: This is very often combined each other. If you want to have a high-power density, the power density is mostly not limited by the flux density, or this is only at very low frequency the case. But the loss density is not a real limiting factor. The limiting factor is the temperature, which can be achieved in operation. And if you can avoid losses, then the temperature, which can be achieved, will be lower. And so, this way, you can lower the size when you can increase the loss density to a certain level based on some influences you are take advantage of. That is a way how to increase the power density of such converter application.

Marco Jung:

Marco Jung: Do you know the limits of such technology? Because I was thinking if I, let's say, shrink the volumes the whole time the thermal resistance increases, so I need maybe additional cooling concepts for higher power applications. Can you tell us a little bit more about the limits?

Dr. Peter Zacharias:

Dr. Peter Zacharias: We have to make a difference between cooling methods. One cooling method, which is quite common, is to cool the magnetic inductive devices by air, which is blown by a fan to the chokes or to the transformers. In this case, you have an internal thermal resistance, which limits the transportation of the heat from the heat source to the surface. And you have another thermal resistance, which gives you the transfer of the heat to the air, which is flowing at the surface.

Dr. Peter Zacharias: And if you decrease the surface, that is what you mean, if you decrease the surface, then of course, you have a higher thermal resistance because of the formula of the thermal resistance. So, there is a limit of what you can get out of thermal power, if you go like this. But if you combine the heat transportation out of the core and of the winding to a water-cooled system, we have another cooling problem because you have, in this case, not convection, you have only heat conduction through a solid body.

Dr. Peter Zacharias: And the solid body, let's say, if there is a heat sink made of aluminum, then you have a very high thermal conductivity of aluminum and get out a high-power density. But then you have still some limits inside the core because the thermal conductivity of the core is not really high. You have values about four watts per meter in Kelvin, and this is really not high conductivity, thermal conductivity. If you compare this to copper, this has almost 400 watt per meter in Kelvin or aluminum, this is about 200 watt per meter in Kelvin.

Dr. Peter Zacharias: And so, you have a self-limiting effect of the core itself. You cannot make legs of the core material, which are really long to get out the heat from the heat source, which is internally of the material to the surface. So, this is limited, and this brings us to low profile cores. If you have low profile cores, then the power density, the loss power density is generated also in the center or the site leg. And to bring it to the surface, which is cooled, you need a certain distance which should not be too large, because in that case, the thermal conductivity of the material limits itself and causes high temperatures inside the material.

Marco Jung :

Marco Jung : Thank you. I think we have spoken a lot of time, but one question I have, what we will see from Peter Zacharias in the future?

Dr. Peter Zacharias:

Dr. Peter Zacharias: As you know, by the end of September, I will retire. And so, in that case, usually, one ends the professional life. But in my case, I hope to be in very close contact to the ECPE and I want to engage myself, my person, into activities of the ECPE, to give some advice to others, and also to be involved in some projects as an senior advisor or something like that. I definitely will not stop my activities at all, but I will slow down.

Marco Jung:

Marco Jung: Step by step, I think so. Well, I'm sure sometimes you will increase it. I know you.

Marco Jung: Okay, Peter, thanks for your statements. It was really a pleasure for me. Thank you.

Marco Jung: To all the listeners, wherever you might be, thank you very much for listening. We hope you have enjoyed today's episode and gained some valuable insights. Make sure to subscribe to our podcast, which is available on all major podcast platforms. If you would like to share your feedback with us, please do so via an email to podcast-pcim@messago.com. You can write to us also if there is a specific topic you would like to be covered in the future, or if you have a particular guest on your mind. Until then, have a great time.

Marco Jung: We hope you enjoyed this edition of Sound on. Power on. Powered by PCIM Europe. Do subscribe and share.