All of your video's about an intuïtive aproach are always the most usefull, at least for me because I love it when i can get that insight picture clear
Interestingly I was taught ONLY this method for inductor/flyback core design years ago and then learnt the Ap method from you, I don't know who came up with this idea originally though but it seems very intuitive and easy to understand.
Hello, firstly thank you for the knowledge you give us. I have a question can you tell reasonable air gap height limit. For an example for transformer i design with that ferrite E55/28/21-3C94 i need air gap 30 mm, is that to much considering to EMI.
@@sambenyaakov Perhaps It is a real challenge to design LLC converter. The limiting factor i see is delta B. Could you suggest core material with low losses. (I need to go to 0.4 T). Although i redesigned the resonant stage and now i need 6.5 mm air gap for the same core , but still seems to much for me.
Sam, it is a very tricky thing assuming ur ~ 3000 - 4000. Ferrites or other materials with that high permeability are not generally usable due to low frequency range, low stability, high dependence on external magnetic field and low accuracy. If you go to realistic values of 1000 - 2000 you find out that a substantial portion of energy IS NOT STORED IN THE GAP but rather stored also in the magnetic core.
Dear Gregory, Thanks for comment. Permit me to disagree. Below is a table from EPCOS. Up to 100kHz the mur are around 3000. The ratio between energy stored in gap to energy stored in ferrite is =mur*(volume of gap)/(volume of ferrite) . Asssume 1mm long gap and 100mm ferrite length and mur of ferrite 1000, the energy stored in gap is 10 times that in the ferrite. I would say that this is substantial. Regards sam __________________________ Power Transformer (low loss) f1 MHz PC200 µi = 800 frequency up to 4 MHz
Very interesting , once I read some PhD notes by Harold aspden written at Southampton university on his research into magnetic phenomena . It lightly showed that the gap in a laminated mains transformer had a larger energy density in the gap which he cut out. He was into many types of discourses into research at an astronomical level. Recommended refreshment here!
It looks like in the discussion all participants periodically forget that any device that stores magnetic energy should store it in some volume that is proportional to the power to store for a while in order to transfer it to the load later. In order to store energy in some substance you have to be sure the substance is CAPABLE of: 1. Magnetizing and de-magnetizing, i.e. changing orientation of the magnetic dipoles it consists of (it may not consist of any) under the influence of the external magnetic field and 2. Providing a mode when ALL the dipoles change the orientation. If some of the dipoles cannot change their orientation in the changing magnetic field, they are saturated, and no magnetic field transfer is possible by these dipoles. The more intense magnetic field is applied to the substance displaying magnetic properties (H), the more sensitive the dipoles are to the external magnetic field (B) and the larger physical volume they take (Vol), the higher power they can transfer through the magnetization / de-magnetization process. This is why we should not forget about the air gap VOLUME or distributed gap magnetic core VOLUME while discussing about the gapped inductor design. We can define the air gap in the ferrite core but the cross-sectional area should always be present since we store energy in the VOLUME. Otherwise, the inductor will fail to transfer energy / power.
Gregory, I take issue with your assessment: "In order to store energy in some substance you have to be sure the substance is CAPABLE of: 1. Magnetizing and de-magnetizing, i.e. changing orientation of the magnetic dipoles it consists of (it may not consist of any......." This scenario is possible but not a necessary condition. Think about an air core inductor. Furthermore we know from fundamental principles that the energy stored in any volume is the 'Volume Integral' of 1/2(B*H)dV, no magnetic dipoles are required.
Sam, you noticed that I was using a term "substance" to designate ANY material for the inductor core, air included. The air core inductor will definitely store energy but its efficiency would be very low due to the energy distribution "everywhere" since there would be no substance that concentrates the stored energy. And the fringing effect that kills inductors' efficiency would be ubiquitous. This is why we use ferromagnetic cores, which are capable of the magnetic field lines concentrating in a bunch and reducing stray parameters. Therefore, for real inductors the scenario I described is a necessary condition. By the way, in your lectures you describe inductors having ferromagnetic cores for the same reason, which is absolutely correct.Another issue is your statement that "...we know from fundamental principles that the energy stored in any volume is the 'Volume Integral' of 1/2(B*H)dV, no magnetic dipoles are required." is absolutely incorrect because the magnetic dipoles are hidden in the magnetic flux density B and they are required by the relative permeability factor that is a substantial part of the B value. Moreover, the magnetic dipoles make the vector B deviate from the vector H thus delivering some interesting effects to the magnetics but this is a different story.
Thanks for the video . At 6:50 you come to the conclusion that the air gap is inversely proportional to the cross sectional area of the core by equating the volume of the air gap (Area x length) with an expression which contains B . This is not a valid move because B actually depends on the cross sectional Area of the core (B=flux/Area = N.I/Area) with the consequence that the Area of the core has not been isolated but rather is present on both sides of the equation . Applying the above in your equation we then have Area x Lg ~ K x Area squared (because B is squared) ( ~ stands for proportionality ) K= 2Muo.E/(N.I)squared Dividing through by the Area gives Lg ~ Area . i.e. the air gap is directly proportional to the cross sectional Area . The same result is obtained by substituting for B=flux/Area = N.I/Area in your original equation .
The entry point of an inductor design is an already specified B, based on saturation or losses (see relevant videos in my RUclips channel). The design needs to comply with the specified B and not the other way around.
@@sambenyaakov Thanks for the reply . However, you are avoiding my assertion that there is a wrong argebraic move at 6:50 which leads to an incorrect proportionality relationship .
Hi, sorry but I do not agree. 1. As I have written earlier B is a fixed number in this derivation 2. (B=flux/Area = N.I/Area) is incorrect B=mu*H = mu*N*I/le , le magnetic path length 3. The fact that the gap volume is proportional to the energy is well known and hence for a fixed energy lg is proportional to 1/Ae
All of your video's about an intuïtive aproach are always the most usefull, at least for me because I love it when i can get that insight picture clear
Thanks Rob. Intuitive understanding is to me a prerequisite for a good design.
@@sambenyaakov yes indeed, that is something só true!
Interestingly I was taught ONLY this method for inductor/flyback core design years ago and then learnt the Ap method from you, I don't know who came up with this idea originally though but it seems very intuitive and easy to understand.
Thanks for comment. Indeed interesting and mind boggling
Many thanks professor. I really appreciate the efforts made to deliver such valuable knowledge.
Thanks
Hello, firstly thank you for the knowledge you give us. I have a question can you tell reasonable air gap height limit. For an example for transformer i design with that ferrite E55/28/21-3C94 i need air gap 30 mm, is that to much considering to EMI.
This is too much. You need to reconsider the core size.
@@sambenyaakov Perhaps It is a real challenge to design LLC converter. The limiting factor i see is delta B. Could you suggest core material with low losses. (I need to go to 0.4 T). Although i redesigned the resonant stage and now i need 6.5 mm air gap for the same core , but still seems to much for me.
I have one small question - is flat magnetics and planar magnetics the same thing?
Yes
Does this work for gaps as large as a centimeter with core length being 20 to 30 times longer than the gap?
I the gap is large, fringing become dominant and will make a major change in parameters such as coupling coefficient.
Sam, it is a very tricky thing assuming ur ~ 3000 - 4000. Ferrites or other materials with that high permeability are not generally usable due to low frequency range, low stability, high dependence on external magnetic field and low accuracy. If you go to realistic values of 1000 - 2000 you find out that a substantial portion of energy IS NOT STORED IN THE GAP but rather stored also in the magnetic core.
Dear Gregory, Thanks for comment. Permit me to disagree. Below is a table from EPCOS. Up to 100kHz the mur are around 3000. The ratio between energy stored in gap to energy stored in ferrite is =mur*(volume of gap)/(volume of ferrite) . Asssume 1mm long gap and 100mm ferrite length and mur of ferrite 1000, the energy stored in gap is 10 times that in the ferrite. I would say that this is substantial.
Regards
sam
__________________________
Power Transformer (low loss)
f1 MHz
PC200
µi = 800 frequency up to 4 MHz
Very interesting , once I read some PhD notes by Harold aspden written at Southampton university on his research into magnetic phenomena . It lightly showed that the gap in a laminated mains transformer had a larger energy density in the gap which he cut out. He was into many types of discourses into research at an astronomical level. Recommended refreshment here!
Thanks for comment.
It looks like in the discussion all participants periodically forget that any device that stores magnetic energy should store it in some volume that is proportional to the power to store for a while in order to transfer it to the load later. In order to store energy in some substance you have to be sure the substance is CAPABLE of: 1. Magnetizing and de-magnetizing, i.e. changing orientation of the magnetic dipoles it consists of (it may not consist of any) under the influence of the external magnetic field and 2. Providing a mode when ALL the dipoles change the orientation. If some of the dipoles cannot change their orientation in the changing magnetic field, they are saturated, and no magnetic field transfer is possible by these dipoles. The more intense magnetic field is applied to the substance displaying magnetic properties (H), the more sensitive the dipoles are to the external magnetic field (B) and the larger physical volume they take (Vol), the higher power they can transfer through the magnetization / de-magnetization process. This is why we should not forget about the air gap VOLUME or distributed gap magnetic core VOLUME while discussing about the gapped inductor design. We can define the air gap in the ferrite core but the cross-sectional area should always be present since we store energy in the VOLUME. Otherwise, the inductor will fail to transfer energy / power.
Gregory, I take issue with your assessment:
"In order to store energy in some substance you have to be sure the substance is CAPABLE of: 1. Magnetizing and de-magnetizing, i.e. changing orientation of the magnetic dipoles it consists of (it may not consist of any......." This scenario is possible but not a necessary condition. Think about an air core inductor. Furthermore we know from fundamental principles that the energy stored in any volume is the 'Volume Integral' of 1/2(B*H)dV, no magnetic dipoles are required.
Sam, you noticed that I was using a term "substance" to designate ANY material for the inductor core, air included. The air core inductor will definitely store energy but its efficiency would be very low due to the energy distribution "everywhere" since there would be no substance that concentrates the stored energy. And the fringing effect that kills inductors' efficiency would be ubiquitous. This is why we use ferromagnetic cores, which are capable of the magnetic field lines concentrating in a bunch and reducing stray parameters. Therefore, for real inductors the scenario I described is a necessary condition. By the way, in your lectures you describe inductors having ferromagnetic cores for the same reason, which is absolutely correct.Another issue is your statement that "...we know from fundamental principles that the energy stored in any volume is the 'Volume Integral' of 1/2(B*H)dV, no magnetic dipoles are required." is absolutely incorrect because the magnetic dipoles are hidden in the magnetic flux density B and they are required by the relative permeability factor that is a substantial part of the B value. Moreover, the magnetic dipoles make the vector B deviate from the vector H thus delivering some interesting effects to the magnetics but this is a different story.
@@gregorymirsky8707 I think the dipoles are hidden in magnetization M. H = B/μ - M.
@@j121212100 This is more related to dipole-dipole interaction.
Thanks for the video .
At 6:50 you come to the conclusion that the air gap is inversely proportional to the cross sectional area of the core by equating the volume of the air gap (Area x length) with an expression which contains B .
This is not a valid move because B actually depends on the cross sectional Area of the core (B=flux/Area = N.I/Area) with the consequence that the Area of the core has not been isolated but rather is present on both sides of the equation .
Applying the above in your equation we then have Area x Lg ~ K x Area squared (because B is squared) ( ~ stands for proportionality ) K= 2Muo.E/(N.I)squared
Dividing through by the Area gives Lg ~ Area . i.e. the air gap is directly proportional to the cross sectional Area .
The same result is obtained by substituting for B=flux/Area = N.I/Area in your original equation .
The entry point of an inductor design is an already specified B, based on saturation or losses (see relevant videos in my RUclips channel). The design needs to comply with the specified B and not the other way around.
@@sambenyaakov Thanks for the reply . However, you are avoiding my assertion that there is a wrong argebraic move at 6:50 which leads to an incorrect proportionality
relationship .
Hi, sorry but I do not agree.
1. As I have written earlier B is a fixed number in this derivation
2. (B=flux/Area = N.I/Area) is incorrect
B=mu*H = mu*N*I/le , le magnetic path length
3. The fact that the gap volume is proportional to the energy is well known and hence for a fixed energy lg is proportional to 1/Ae
Hi professor .
I tried a worked example , and yes, you are quite correct .
Sorry to doubt your superior knowledge - one lives and learns .
👍