Greetings Prof. Invaluable knowledge U share with us indeed. A question if I may; @ 2:38 & with reference to the resonant topological diagram, mentioning constant duty cycle, does the D have to be 50% and how would one control output power if I need to keep the frequency constant?
Hello, thanks for your comment and question. Yes, the duty cycle is usually kept at 0.5. If you change the duty cycle you will increase the harmonic content of the output voltage generated by the half bridge. This will increase the resonant current distortion making it difficult to attain ZVS operation at different power levels. To keep the frequency constant it is better to use a full bridge instead of a half bridge and apply phase shift control to the full bridge. This way you can keep low distortion and achieve ZVS operation in a wide range.
Hello, Marcos! As always, thanks for your detailed video! A little question: why we should to use delta value of spikes at 32:34? Isn't the both spikes negative and positive represent losses? Or it represents stored and extracted from Cds energy only?
Yes to the latter question. This is energy handled by Cds. Only the difference between both spikes are losses. These losses are dissipated in the MOSFET's channel. Thanks for watching!
Thank you sir for the detailed video. However, could you please provide code/screenshot for the code written to generate voltage gain vs frequency plot for different value of normalised inductance and quality factor? I'm having difficulty in getting that plot. Thanks.
You only have to plot the equation of M_g shown at 6:50. You can use Winpython for this. In video Winpython #1 at th-cam.com/video/5aF-Iwf3dzg/w-d-xo.html, you can see an introduction to Winpython. It also shows how to plot functions. You can do the same for M_g.
@@MarcosAlonsoElectronicsthanks for your prompt response. Can you please elaborate how to find relevant values from Mg plot after plotting it on winpython since cursor (as available in ltspice) is not available in winpython. How do I find F0 and max and min gain in this case? Thanks.
Hi sir, in this video at time 22:33 you said that tdead is >85ns how you calculated . I have calculated using the formula 16*Ceq*fsw*Lm , the values i have taken as Ceq=0.2+0.2=0.4 , fsw=150kHz , Lm=210uH. i am getting 200ns . Kindly clarify the doubt sir.
Hi, thanks for your question. The formula is on the slide at 15:00. Using Ceq~Cds=0.2 nF (only one switch's because 2*Ceq is already considered in the formula), fsw=140 kHz (max. switching frequency), Lm=210 uH. We get t_dead=16*Ceq*fsw*Lm ~ 85 ns.
@@MarcosAlonsoElectronics thank you for the reply sir. One more question sir why switching frequency chosen as 140kHz instead of 150kHz? Is there any reason, sir?
Thank you proff. for the insight. If I want to do synchronous rectification of this LLC converter then how to provide the gate pulse for the other switches. Can you please give me the idea about that?
Usually SR is implemented using auxiliary windings on the transformer to drive the switches but there are also specific ICs for this purpose. See for example the UCC24624
Sir, If I measure voltage across transformer secondary without diode bridge rectifier then I should get a sine wave right? When I did hardware spike across transformer secondary was huge. Is this because switching is not proper?
Hello, if you remove the rectifier the resonant tank will be operating with no load. The waveforms can be strange. If you want to see sinusoidal waveforms without the rectifier you have to place an equivalent load resistor at the secondary.
@Marcos Alonso, for resonant converters (e.g. LLC), how someone can implement a small signal model in simulation? With voltage/current sources, like how is it done in the case of PWM dc dc converters? How could for example the control-output transfer function for an LLC converter be deduced in simulation with a small-signal model (so not using a switching or averaging model, but a small-signal model)? There is such a thing? Can you do a video showing how to do this?
Thanks for the suggestion. To do a dynamic model of the LLC resonant converter is more complicated than a conventional non-resonant DC-DC converter. It is not enough to consider averaged values because the state variable are AC so we need to consider the first harmonic too. This is a topic that I have in mind for a future video. Thanks!
@Marcos Alonso But you already know how to do the small signal model by calculation and simulation for resonant converters? It would be good then if you could accelerate the creation of these videos on how to calculate and simulate the small signal model for the transfer functions: control-output, audiosusceptibility, input impedance, output impedance, in open loop and closed loop at least in voltage mode control, for these resonant converters. Would you do this?
This was just a possible operating point that was selected to obtain a value of 12 V at the output for a given input voltage. You can choose other points within the switching frequency range. Remember that the switching frequency is our control parameter so it is variable. Thanks for watching!
Hi, thanks for your question. U2 is a voltage controlled oscillator (VCO) from our Simulink-compatible control library. Please, watch videos LTspice #11-14 and #20 for more info on this library. It is available from my website.
Sir is there any way we can measure converter input impedance using LTspice? I was trying with buck converter input impedance by placing series input ac source in the input side but unable to get it ..any suggestions?
Yes, it is possible. The procedure is similar to that shown in video LTspice #6: th-cam.com/video/JCHiMp47UGw/w-d-xo.html You have to inject the modulating signal in series with the input voltage. From the input voltage and input current perturbations you can calculate the input impedance and represent it from the log file data. This can be interesting for a future video.
@@MarcosAlonsoElectronics I was trying to design the input filter for buck converter so instead of calculating buck converter input impedance analytical ,I was trying with simulation. I tried to search in internet but it seems like no one measuring input impedance using LTspice that make me think how to do it.
U3 is a current sensor available from the Simulink-compatible control library. There is a playlist on this library on my channel. The library is available from my website. I use it to measure the resonant current as volts so that it can be represented together with the magnetizing current generated by the voltage source.
This is an OUTSTANDING presentation. Thank you for your work and sharing.
Thank you very much
Oh my god can't believe.... many many thanks 🙏
Thanks for watching!
Greetings Prof. Invaluable knowledge U share with us indeed. A question if I may; @ 2:38 & with reference to the resonant topological diagram, mentioning constant duty cycle, does the D have to be 50% and how would one control output power if I need to keep the frequency constant?
Hello, thanks for your comment and question.
Yes, the duty cycle is usually kept at 0.5. If you change the duty cycle you will increase the harmonic content of the output voltage generated by the half bridge. This will increase the resonant current distortion making it difficult to attain ZVS operation at different power levels. To keep the frequency constant it is better to use a full bridge instead of a half bridge and apply phase shift control to the full bridge. This way you can keep low distortion and achieve ZVS operation in a wide range.
What a wonderful video!
Thank you!
Been waiting for it. Thanks!
Thank you!
Thank you very much for sharing this.
You are welcome. Thanks for watching
Hello, Marcos!
As always, thanks for your detailed video!
A little question: why we should to use delta value of spikes at 32:34? Isn't the both spikes negative and positive represent losses? Or it represents stored and extracted from Cds energy only?
Yes to the latter question. This is energy handled by Cds. Only the difference between both spikes are losses. These losses are dissipated in the MOSFET's channel. Thanks for watching!
excellent video, very useful!
Thanks!
LTspice file is available here: github.com/marcosalonsoelectronics/PE-40
Thank you sir for the detailed video. However, could you please provide code/screenshot for the code written to generate voltage gain vs frequency plot for different value of normalised inductance and quality factor? I'm having difficulty in getting that plot. Thanks.
You only have to plot the equation of M_g shown at 6:50. You can use Winpython for this. In video Winpython #1 at th-cam.com/video/5aF-Iwf3dzg/w-d-xo.html, you can see an introduction to Winpython. It also shows how to plot functions. You can do the same for M_g.
@@MarcosAlonsoElectronicsthanks for your prompt response. Can you please elaborate how to find relevant values from Mg plot after plotting it on winpython since cursor (as available in ltspice) is not available in winpython. How do I find F0 and max and min gain in this case?
Thanks.
Hi sir, in this video at time 22:33 you said that tdead is >85ns how you calculated . I have calculated using the formula 16*Ceq*fsw*Lm , the values i have taken as Ceq=0.2+0.2=0.4 , fsw=150kHz , Lm=210uH. i am getting 200ns . Kindly clarify the doubt sir.
Hi, thanks for your question. The formula is on the slide at 15:00. Using Ceq~Cds=0.2 nF (only one switch's because 2*Ceq is already considered in the formula), fsw=140 kHz (max. switching frequency), Lm=210 uH. We get t_dead=16*Ceq*fsw*Lm ~ 85 ns.
@@MarcosAlonsoElectronics thank you for the reply sir. One more question sir why switching frequency chosen as 140kHz instead of 150kHz? Is there any reason, sir?
Thank you proff. for the insight. If I want to do synchronous rectification of this LLC converter then how to provide the gate pulse for the other switches. Can you please give me the idea about that?
Usually SR is implemented using auxiliary windings on the transformer to drive the switches but there are also specific ICs for this purpose. See for example the UCC24624
Sir, If I measure voltage across transformer secondary without diode bridge rectifier then I should get a sine wave right? When I did hardware spike across transformer secondary was huge. Is this because switching is not proper?
Hello, if you remove the rectifier the resonant tank will be operating with no load. The waveforms can be strange.
If you want to see sinusoidal waveforms without the rectifier you have to place an equivalent load resistor at the secondary.
What about current and voltage at mosfet look like from zvs ? Because vgs and vds show at present only.
The waveforms are shown at 13:57
@Marcos Alonso, for resonant converters (e.g. LLC), how someone can implement a small signal model in simulation? With voltage/current sources, like how is it done in the case of PWM dc dc converters? How could for example the control-output transfer function for an LLC converter be deduced in simulation with a small-signal model (so not using a switching or averaging model, but a small-signal model)? There is such a thing? Can you do a video showing how to do this?
Thanks for the suggestion. To do a dynamic model of the LLC resonant converter is more complicated than a conventional non-resonant DC-DC converter. It is not enough to consider averaged values because the state variable are AC so we need to consider the first harmonic too.
This is a topic that I have in mind for a future video. Thanks!
@Marcos Alonso But you already know how to do the small signal model by calculation and simulation for resonant converters? It would be good then if you could accelerate the creation of these videos on how to calculate and simulate the small signal model for the transfer functions: control-output, audiosusceptibility, input impedance, output impedance, in open loop and closed loop at least in voltage mode control, for these resonant converters. Would you do this?
Dear Prof. Alonso, could you please do a similar video on PSFB? Maybe even in Qspice
Hello, thanks for your suggestion. I will consider
Thanks, are there any papers describing High Power charging applications of LLC?. 50kW and above?
Thanks for your question. I have no information on this type of high-power application.
@@MarcosAlonsoElectronics No problem, thank you for all your videos.
Thank you for the reply sir. One more question, sir, why switching frequency chosen as 140kHz instead of 150kHz? Is there any reason, sir?
This was just a possible operating point that was selected to obtain a value of 12 V at the output for a given input voltage. You can choose other points within the switching frequency range. Remember that the switching frequency is our control parameter so it is variable.
Thanks for watching!
@@MarcosAlonsoElectronics thank you very much sir.
As switching frequency is different from resonant frequency then how the resonance occurs in the circuit
Resonant converters usually operate around resonant frequency and not only exactly at resonant frequency.
Thanks for your work. What can you say about ZCS forward converter? This is very interesting theme.
Thanks for your comment. I will think about it.
👍👍👍👍👍👍👍👍👍👍
Thanks!
Hai Marcos,
enjoyed the video, what is the component U2?
Hi, thanks for your question. U2 is a voltage controlled oscillator (VCO) from our Simulink-compatible control library. Please, watch videos LTspice #11-14 and #20 for more info on this library. It is available from my website.
@@MarcosAlonsoElectronics Thanks for you reply, please do a video of LLC closed control
Sir is there any way we can measure converter input impedance using LTspice?
I was trying with buck converter input impedance by placing series input ac source in the input side but unable to get it ..any suggestions?
Yes, it is possible. The procedure is similar to that shown in video LTspice #6: th-cam.com/video/JCHiMp47UGw/w-d-xo.html
You have to inject the modulating signal in series with the input voltage. From the input voltage and input current perturbations you can calculate the input impedance and represent it from the log file data.
This can be interesting for a future video.
@@MarcosAlonsoElectronics I was trying to design the input filter for buck converter so instead of calculating buck converter input impedance analytical ,I was trying with simulation. I tried to search in internet but it seems like no one measuring input impedance using LTspice that make me think how to do it.
@@biswajit681 I will post a video about how to do this soon. This is on my to-do list now.
@@MarcosAlonsoElectronics Many Thanks sir ...
Thank you sir!
Thanks for watching!
Hello! good video! I made experiments - the LLC converter can operate with current density 4A/mm sq.I mean litz wire (wire wound like steel rope)
Hello, thanks for your comment and for sharing your expertise
Bạn có thể cho tôi xin tài liệu mô phỏng LPspice như trên đươc không. Cảm ơn
Tệp mô phỏng hiện có sẵn ở đây:
github.com/marcosalonsoelectronics/PE-40
Cảm ơn bạn đã ghé thăm kênh!
Hello sir, I have done this ,can you pls make video on close loop control of LLC
Thanks for the suggestion. I will consider.
Hi sir ..is everything alright?? I hope you are doing well
Hi, I am good, thanks for asking!
just very busy these last weeks, not much time to post videos
but please see my last video posted today, thanks!
Sir what is U3?
U3 is a current sensor available from the Simulink-compatible control library. There is a playlist on this library on my channel. The library is available from my website.
I use it to measure the resonant current as volts so that it can be represented together with the magnetizing current generated by the voltage source.
Or can you please make video on vco
This video shows how to implement a VCO: th-cam.com/video/BSiRYkqVfqU/w-d-xo.html
Thank you for sharing this.
Thanks for watching!
thank you very much 🙏
Thanks!