Power Electronics - Switching Losses in a MOSFET

[@ashutoshyadav1270]

Thank you Sir, great videos. Would be quite helpful for your viewers if you could organize your videos in playlists.

[@fatihe.1338]

Hi, in 5:15,What is the O in parentheses in VDS(O) and QGD(O)? Shouldn't CGD and VDS be inversely proportional? Can you explain a bit? Also, in the AN123 file you shared, the calculated and measured values ​​are very different from each other. Can you explain why?

[@powerelectronicswithdr.k1017]

Hello Fatihe.1338, My handwriting is not good. That should be VDS(D) and D stands for the datasheet. The switching times can be very difficult to obtain accurate values as these times change based on the load requirements and the DC bus supply. Here is a good video from TI regarding the info on a MOSFET datasheet as it relates to turn-on/off times www.ti.com/video/series/mosfet-101.html#

[@markoula7211]

Hi, thank you for the great tutorial! I have a question regarding the Id current at the end of Miller plateau. If the Vgs is not at the maximum value at the end of plateau, how does the Id current reaches maximum value? My understanding is that the current should increase after plateau because Vgs is increasing and opening the channel of the mosfet further? Could you elaborate this? Thank you, Marko.

[@powerelectronicswithdr.k1017]

That's a great question. It is actually easier to illustrate if you have an Ids vs Vds graph from a MOSFET in front of you. In that graph, you will find different curves for the different valuse of Vgs. As Vgs starts to increase past the thershold voltage value, the drain current will start to increase from 0A along a straight vertical line in the Ids vs Vds graph. It will reach a value of Vgs that is not the maximum value, but the current will be at maximum value. The point will then start to move along this Vgs line (or very close to it) towards the left and begin to enter the Ohmic region of the MOSFET. Also note that the increasing Vgs curves are nearly on top of each other in the Ohmic region. Once in that region Vgs will reach its maximum value, yet the current does not increase drastically. The Miller plateau is the region where the point is moving from the right to the left in the Ids vs Vds curve at a near constant Vgs value. I do hope this explanation helps, but it is much easier to visualize with a point moving in the Ids vs Vds graph. Best wishes on your design. -Dr. K

[@thuyavansathiamurthi3592]

This is by far the simplest and most comprehensive explanation of the timing of mosfets. It answered all the questions I've had about the power losses and I can't thank you enough!

[@powerelectronicswithdr.k1017]

You're welcome! -Dr. K

[@random2059]

do extra heat on mosfet can be a cause of permanent /change in #threshold voltage/ ......that's can ruin the mosfet in long run.... please reply

[@muraliselvam3394]

in the formula some words I can't understand. Would you show clear text?

[@sth4someone912]

Thanks for preparing the video. Could you please prepare a video on how to calculate the switching loss by the aid of simulation tools such as cadence or spice? Thanks

[@powerelectronicswithdr.k1017]

Hello, if you go to kzbin.info/www/bejne/o6SZhZx7j659js0 and at about the 11:00 minute mark I will illustrate how to check the instantaneous power in the simulated DC chopper circuit's MOSFET. You can also window the switching waveform and LT Spice will compute the average loss. Hope this helps. -Dr. K.

[@SMV1972]

Отличный урок! Спасибо вам большое!!

[@musonglouiskatche6219]

Thank you for this interesting video. Please can you share with me the proofs for the turn-on and turn-off times displayed in this video?

[@powerelectronicswithdr.k1017]

Hi Musong, those are estimates taken from www.vishay.com/docs/73217/an608a.pdf reference equations 17-20 in this document. Thanks for watching. -Dr K.

[@cammasters2053]

Hi Dr. K, I'm building a model of a bidirectional DC-DC converter in simulink and logging efficiency against switching frequency. I am expecting a peak followed by a downturn within the graph where switching losses dominate. But no matter how high the switching frequency is the efficiency keeps increasing which I think is wrong. Any ideas on how to resolve this?

[@powerelectronicswithdr.k1017]

Hi Camrone, unfortunately I have never used Simulink for analysis of DC/DC converters. My suggestion is to look at all the losses in all the devices to see which device is not being modelled correctly as you sweep the operating frequency point. Keep me posted. -Dr. K

[@cammasters2053]

@@powerelectronicswithdr.k1017 Hi Dr. K, think I have fixed the problem, I don't think switching losses are considered in Simulink so i have modelled them manually, I do have another question though, I've been looking at mosfet datasheets and have noticed there is a turn on delay time and a turn on rise time, (same for turn off) when calculating switching losses do i need to consider both of these times or just rise time? i.e. is t-on = t-d(on) + t-rise or t-on = t-rise ? Thanks in advance

[@powerelectronicswithdr.k1017]

@@cammasters2053 Hi Camrone, here's a link to info from TI on the MOSFET datasheets. training.ti.com/mosfet101training.ti.com/mosfet101. This is a really good series and Video 1.4 discusses the Ton, Trise, Td, etc. The vendors follow JEDEC standards for testing and reporting information. Hope this helps. -Dr. K