I had a similar issue few weeks ago with a high voltage generator on the primary side of the transformer. Those high current peaks as on the graph at 15:23 didn't seem to appear due the frequency itself, they were rather dependent of the point where the switching met the ringing of the circuit (based on the inductance of the primary coil and the MOSFET capacitance probably). The worst case was when I tried to force the current by turning it on when it had to counteract the current slope generated by the ringing of the preceding excitation. Even if it was not the strongest, first period of the ringing but a later one, the sharp current peaks were present. By tuning the timing of the excitation and the ringing I was able to reduce those peaks near to zero, just as on the upper part of the picture. Could my issue had some completely different cause than you addressed in this presentation? In my circuit there was a simple N-MOSFET pulling down the primary coil.
for Buck or Boost converter, beside switching loss, is BCM mode always outperform CCM in term of overall efficiency? Or it limit in low-moderate load current case as BCM need very large ripple results in high Irms*dcr loss?
Dear Professor, I do not understand why the main loss is the product of drain voltage and current during hard switching turn-on. What with the capacitance parallel to the transistor which is discharging through the transistor during the transistor turn-on?
@@sambenyaakov Dear Professor, thank you for the reply. I'm looking forward to this video. I want to precise my question. In ua-cam.com/video/I7PvLSC-kzA/v-deo.html you describe hard switching turn-on. In simulation results, we can see the current of the transistor (M1) which is the current of the inductor and reverses recovering current of the diode. This current and the transistor voltage cause the switching power losses. But what with the energy stored in the capacitance of the transistor M1 ? Why we do not observe the current which discharge the transistor capacitance like in flyback ua-cam.com/video/I7PvLSC-kzA/v-deo.html ?
You are correct, there is an additional loss due to the discharge of the Coss. I am not sure if LTspice takes tis current into account when plotting the power loss.
No, you dont, The voltage is clamped to the bus and the current depends on the load. Working at a slightly higher frequency than resonance is the best.
I see then it´s best to put the frequency slightly above resonance for normal operation and for short circuit protection pushing the frequency way higher because as you say we still have ZVS benefit at higher frequencies. Thanks for the nice video !
Very helpful Professor. Thank you for your continuous effort in spreading knowledge!
Thanks
Thanks again Professor.
Thanks
I had a similar issue few weeks ago with a high voltage generator on the primary side of the transformer. Those high current peaks as on the graph at 15:23 didn't seem to appear due the frequency itself, they were rather dependent of the point where the switching met the ringing of the circuit (based on the inductance of the primary coil and the MOSFET capacitance probably). The worst case was when I tried to force the current by turning it on when it had to counteract the current slope generated by the ringing of the preceding excitation. Even if it was not the strongest, first period of the ringing but a later one, the sharp current peaks were present. By tuning the timing of the excitation and the ringing I was able to reduce those peaks near to zero, just as on the upper part of the picture. Could my issue had some completely different cause than you addressed in this presentation?
In my circuit there was a simple N-MOSFET pulling down the primary coil.
Thanks for sharing. It appears that the tuning did the job of achieving soft switching.
for Buck or Boost converter, beside switching loss, is BCM mode always outperform CCM in term of overall efficiency? Or it limit in low-moderate load current case as BCM need very large ripple results in high Irms*dcr loss?
At higher currents the conduction loss increase due to the high rms values.
שלום שמואל, תודה רבה על הסרטונים, מעשיר מאוד.
אשמח אם אוכל להתייעץ אתך בנוגע למעגל HB ואחריו LC resonance tank, הספק גבוה.
בהחלט.
sam.benyaakov@gmail.com
👍🙏❤️
👍🙏😊
Dear Professor,
I do not understand why the main loss is the product of drain voltage and current during hard switching turn-on. What with the capacitance parallel to the transistor which is discharging through the transistor during the transistor turn-on?
You have keen eyes. Thanks. There is a problem with the explanation. I was carried away. This part was deleted and I plan to post a correction video.
@@sambenyaakov Dear Professor, thank you for the reply. I'm looking forward to this video.
I want to precise my question. In ua-cam.com/video/I7PvLSC-kzA/v-deo.html you describe hard switching turn-on. In simulation results, we can see the current of the transistor (M1) which is the current of the inductor and reverses recovering current of the diode. This current and the transistor voltage cause the switching power losses. But what with the energy stored in the capacitance of the transistor M1 ? Why we do not observe the current which discharge the transistor capacitance like in flyback ua-cam.com/video/I7PvLSC-kzA/v-deo.html ?
You are correct, there is an additional loss due to the discharge of the Coss. I am not sure if LTspice takes tis current into account when plotting the power loss.
@@sambenyaakov thank you.
If you hit resonance exactly at the LLC converter you destroy your switches from overvoltage right ?
No, you dont, The voltage is clamped to the bus and the current depends on the load. Working at a slightly higher frequency than resonance is the best.
I see then it´s best to put the frequency slightly above resonance for normal operation and for short circuit protection pushing the frequency way higher because as you say we still have ZVS benefit at higher frequencies. Thanks for the nice video !
Indeed
Thanks professor. Pls doublecheck the slide 9, it looks the overlap loss waveform of Rg=1ohm needs to be corrected to that of Rg=50ohm.
Thanks for comment. I am sorry, but this slide has inaccuracies. I have deleted this section. I will post a correcting video.
Sir, in the very first slide you have shown diode across the MOSFET. Is it body diode or external diode because the current direction is reversed.
These are body diodes. he graph shows the magnitude of the current not polarity as is clear from the narration.
Is zvs turn off really helpfull? Why would one do that, if i were to transfer current to upper device.?
In PWM converters the duty cycle controls the power transfer so do have to turn on and off the transistors, unless I did not understand the question.
But if i need to control the impact of the noise generated by the source, i need fixed frequency of commutation.
You mean synchronizing to another source? Well, you can use a variable inductor as in ua-cam.com/video/jmIdYymO7rE/v-deo.html
@@sambenyaakov I knew about all these, separate aspects, but never thought about putting it together. Very smart!
Hi Sir, nay u help to design transformer for flyback converter in step byxstep process.
Have you seen ua-cam.com/video/Y0WWj2dO_h8/v-deo.html ?