Dear Professor, Very helpful presentation. The variation of power loss with temperature is quite interesting. While working with ferrites, I observed from the manufacturer datasheet(TDK software) that the permeability of ungapped ferrites increase with temperature. In that window the power loss decreases with temperature as well. If the ferrites are gapped, the permeability does not vary much. the thermal stability is quite good for the gapped cores I guess. Thank you.

When you mentioned (4:20) that a typical saturation limit for cores is 300 mT or so, and then mentioned (17:50) that a typical excitation limit for powder cores is 50 mT or so, the ratio of the two is constant (about 8). This made me realize something I had not before : We know the formula for the saturation limit based on current : Is is Bmax = Imax L / (N Ae). Here Imax is the maximum current through your inductor. Now your work (around 21:50) also gives a formula for delta-B : For boost converters, at D=0.5, delta-B = 1 / (N Ae) * Vout/(4*fpwm). Now this is cool, because with Bmax/delta-B being constant 8, that gives us : Vout/Imax = (L fpwm) / 2. This provides a super-simple way to determine an optimal L for a given power converter Vout, Imax and PWM frequency fpwm : L = 2 Vout / (Imax fpwm) So if you have a converter with inductor current 20 A, Vout = 200 V, fpwm=80kHz, then L = 250 uH is a good choice for an inductor based on magnetic saturation / exctation limits. I find that really cool !

Woah! Full Circle! That's amazing that for a given L, Bsat = (1/(nAe))*VdT AND Isat = ( Bsat . Ae ) / √( L . AL ); dT = L / V * Isat !!! That explains why there is no free lunch. I am still stuck on something, for a given core, say Ve = 8320 mm^3, Bmax = 250mT, switch frequency of 100 kHz, and material, say N41, a chart indicates 1,500kW/m^3. Do you multiply the volume by that loss value? Like would that mean my core would really be losing TWELVE WATTS?!? That seems truly insane.

Hello Mr. Ben-Yaakov, this Steinmetz Formula is only for sinusoidal waveforms correct. Otherweis you should use the Äquivalent Frequency instead to the normal switching frequency.

I also studied your video on soft-switching pwm converters. Now for a high power SMPS (2000W>) I think it's most useful to use the fullbridge topology. To achieve maximum efficiency you'd like to have minimum switching losses and core losses. In that presentation you showed an input inductor(Li) between Q1 and Q3 and transformer, now I assume that brings some core losses. What would be the best design for that inductor when Vin

Dear Professor, I can't understand why having DC bias tend to increase loss?? In fact if a core (DC+ripples) works at a positive side of its BH curve there is not any need to demagnetization that reduces a considerable internal frictions. Also I believe the hysteresis band will be much narrower this way. I appreciate your helpful answer for my question in advance. Thanks

Thank you very much for this video! I have a question. Do you have a video about synchronous rectification? Since losses can be really big. Or is it a topic that's not worth it? Thanks again!

Hi Prof Sam, Thanks for the video to introduce the core losses in details. But I have question regarding the core losses versus DCR losses. Would you think the majority losses for the inductor core is came from (DCR losses)? Because I think normally DCR losses is higher than core losses. So in this case, if we want to improve the total losses of the inductor core. Should we decrease the inductance values (for saturation capability), fixed the particular core size (based on the PCB requirement) and increase an appropriate ranges of switching frequency (reduce the current ripple). Because I was concerned about the core loss compared to DCR loss from the video in 9:12 Thank you so much Pro sam

Dear Sir Can you show me how to know the delta B ? Using datasheet of core material, right ?

Dr. Sam, Can you please try to make a video on planar transformer? Here is an app note from TI: www.ti.com/download/trng/docs/seminar/Topic4LD.pdf , but it is little confusing for me because I have never designed power supply using a planar Transformer