Hi Sam, Thanks a lot for the content, There is a Thesis from Caltech in 1980, entitled "Problems in analysis and design of switching regulators", written by SP Hsu, In the second chapter, it does reluctance (in fact, permeance) modeling of the multi-winding transformer, and transforms the coupled inductor model into the conventional Pi equivalent circuit model (in which each winding has a leakage inductance and a magnetizing inductance in parallel with the first winding), showing that it is sufficient and accurate enough in the case of a three-winding transformer (with a few approximations)
Hi Sam, One comment I would make about your different equivalent models - I agree with your comment about them all being equivalent if you apply the appropriate transformations. I have designed several integrated magnetic devices using custom cores that allow a flux shunt to be inserted between the primary and secondary windings. Based on these designs they do end up being somewhat asymmetric and I can implement the leakage inductance so that it is associated with one particular winding - resulting in the bulk of the leakage associated with one winding and a much lesser amount with the other. This does make a difference for the solution as to whether you run the leakage on the primary or secondary from a basis that the volt-seconds applied across the transformer can be different depending on which side you place the leakage. From a simulation model you are correct that you can work out equivalent transformations between the primary leakage and secondary leakage design. That being said from a practical designe perspective it can be beneficial to use one of the designs over the other on the basis of lowering the applied volt-seconds across the transformer. This is a very unique case however as this difference only comes about from the unique core geometries that I have developed. Virtually all LLC designs that I have come across based on industry standard core designs will end up creating symetrical designs - I.E., the leakage ends up being distributed evenly between the primary and secondary windings.
Hi, I think that you have missed the point. In as much that the origin of the leakage can be associated with one port or another, you can always extract the coupled inductor model and then transform it to an equivalent circuit in which the leakage can be put in one side or the other. See ua-cam.com/video/TUfaN40T1Yg/v-deo.html
Thank you! I have a suggestion for new contribution on this channel: Induktive cooker and the typical schematics of these devices. I disassembled some weeks ago one of this. I found inside just 2 power electronics parts: a diode bridge and a 600 V MOSFET. The coil has a parallel capacitor comprising a resonance circuit. It's a miracle for me, how a single MOSFET is able to control the power transfer into the cooking pot. At list I would expect a MOSFETs half bridge.
I could have said that this is a test🙂but this typo is on the original scematics in an earlier video where I have put a correction note in the description. Now it is used to grade the viewers of the video. You got an A+.
Hi Professor, this is also a topic I am working on recently. THanks for your explanation. However, I have tried these four modeling methods in LTSPICE and seen that the two L-models give different results as other two. I have double checked and made sure the parameters as the same as yours. Additionally, all models give like twice higher than your results at 7:42. How come is this? BTW, all your schematics at 3:38 missed Cr.
I believe the L-model has a wrong equivalent turns ratio. The inductance on the side sec. side should be: Lm*1k/(N*N)/(KP1*Kp1) - verified by simulation.
@@sambenyaakov I do now that this is for coupling the windings, I use 1 most of the time. When use 0.85 output voltage is lower, and that is right sim. I thought U did use it for resonant simming. I usetransformer Lp for sim.
Hi Sam,
Thanks a lot for the content, There is a Thesis from Caltech in 1980, entitled "Problems in analysis and design of switching regulators", written by SP Hsu, In the second chapter, it does reluctance (in fact, permeance) modeling of the multi-winding transformer, and transforms the coupled inductor model into the conventional Pi equivalent circuit model (in which each winding has a leakage inductance and a magnetizing inductance in parallel with the first winding), showing that it is sufficient and accurate enough in the case of a three-winding transformer (with a few approximations)
Thanks for the reference,
Hi Sam,
One comment I would make about your different equivalent models - I agree with your comment about them all being equivalent if you apply the appropriate transformations.
I have designed several integrated magnetic devices using custom cores that allow a flux shunt to be inserted between the primary and secondary windings. Based on these designs they do end up being somewhat asymmetric and I can implement the leakage inductance so that it is associated with one particular winding - resulting in the bulk of the leakage associated with one winding and a much lesser amount with the other. This does make a difference for the solution as to whether you run the leakage on the primary or secondary from a basis that the volt-seconds applied across the transformer can be different depending on which side you place the leakage.
From a simulation model you are correct that you can work out equivalent transformations between the primary leakage and secondary leakage design. That being said from a practical designe perspective it can be beneficial to use one of the designs over the other on the basis of lowering the applied volt-seconds across the transformer.
This is a very unique case however as this difference only comes about from the unique core geometries that I have developed.
Virtually all LLC designs that I have come across based on industry standard core designs will end up creating symetrical designs - I.E., the leakage ends up being distributed evenly between the primary and secondary windings.
Hi, I think that you have missed the point. In as much that the origin of the leakage can be associated with one port or another, you can always extract the coupled inductor model and then transform it to an equivalent circuit in which the leakage can be put in one side or the other. See ua-cam.com/video/TUfaN40T1Yg/v-deo.html
Thank you! I have a suggestion for new contribution on this channel:
Induktive cooker and the typical schematics of these devices. I disassembled some weeks ago one of this. I found inside just 2 power electronics parts: a diode bridge and a 600 V MOSFET. The coil has a parallel capacitor comprising a resonance circuit. It's a miracle for me, how a single MOSFET is able to control the power transfer into the cooking pot. At list I would expect a MOSFETs half bridge.
Good topic. Perhaps. The one MOSFET is probably in E class topology.
Amazing content 💐
Thanks
Great video as always, thanks Sam
By the way, isn't the diode bridge rotated in the schematic at 1:46?
I could have said that this is a test🙂but this typo is on the original scematics in an earlier video where I have put a correction note in the description. Now it is used to grade the viewers of the video. You got an A+.
@@sambenyaakov hahaha thanks :)
Hi Professor, this is also a topic I am working on recently. THanks for your explanation. However, I have tried these four modeling methods in LTSPICE and seen that the two L-models give different results as other two. I have double checked and made sure the parameters as the same as yours. Additionally, all models give like twice higher than your results at 7:42. How come is this? BTW, all your schematics at 3:38 missed Cr.
I believe the L-model has a wrong equivalent turns ratio. The inductance on the side sec. side should be: Lm*1k/(N*N)/(KP1*Kp1) - verified by simulation.
👍🙏❤
🙏👍🙂
Nice Video professor. Interesting, I use the same method!
@Robert Bolanos sir could you make detailed video on LLC Resonant converter like flyback converter??
🙏👍
I see you do in ltspice sim the leakage and use K1 L* L* 0.85? . Now I see how you did it, thanks,
Please indicate the minute of video you are referring to.
@@sambenyaakov It was on 4.28 minutes start time. Have a very nice christmas.
@@audiokees4045 K LA LB {XX} is the prompt to define the coupling coefficient , 0.85 is the number used for XX in the simulation
@@sambenyaakov I do now that this is for coupling the windings, I use 1 most of the time. When use 0.85 output voltage is lower, and that is right sim. I thought U did use it for resonant simming. I usetransformer Lp for sim.
@@sambenyaakov Ohh oke, I did use 1 that is to high then to get trustable results from transformer. Thanks.