Best Flyback converter explanation video on the internet till now . Thank you Professor Dr. Ben-Yaakov for creating all these Power electronics educational videos.
My one regret is that my college professors were not so interested in conveying such practical, easily understandable information. I think your style and depth of knowledge is wonderful, as is your willingness to share. Thank you, Dr. Ben-Yaakov.
@@sambenyaakov As the self-teaching individual that I am, I would like to extend my deepest thanks to you and the willingness to share your knowledge. Individuals such as myself greatly rely on the quality of information conveyance that individuals like you provide us with through these videos. Again, I can't express how deeply grateful I am. Thank you Mr. Sam, James M.
Thank you prof. for these lectures in English,I'm taking power converter course and my prof. suggested us to watch your lectures to get more clarity and I agree to him, these lectures definitely provide good clarity of the concepts.
I am an electronician and I admit that you explain wonderfully well. The teachers I had dreaded switching power supplies. Relatively complex concepts have become simple thanks to you.
Great introduction to the Flyback. I found also very useful the practical approach given to the presentation. Thank you for sharing your knowledge with us.
Hello! Good explanation! I`d like to add about a self-oscillating flyback converters with two transistors (like in a cheap mobile charger). These converters very dependents on a transformer , feedback winding turn ratio, there may be over-voltages or avalanche damage - because there is not any gate driver. The stability of this type of converter is not good enough - there is a hysteresis in feedback loop. I have an experience with this type of flyback converter - I only want to share my small experience ))) .Maybe I am not right - but integrated circuits with gate drivers and protection much increase reliability of a flyback converter .
Thanks for sharing. Discrete analog controllers for flyback and other converters are ingenious but passe. Indeed, IC controllers are very effetive and relaiable
Dear Mr Yaakov, I think that at 10mn, the fs frequency is missing in the peak current relationship when you are expressing the power. Thank you for this sharing
Hi Mlaid, You are correct! I hope others will see your comment. The initial expression for P is actually energy so to get power you need, as you indicated, to multiply it by the switching frequency. I will put your note and my answer in the TH-cam page of video. Thanks again.
When the switch is opened up in the primary circuit and the magnetic field around the primary inductor collapses, what becomes of the Back EMF spike? Obviously the large Back EMF spike will be coupled over onto the secondary inductor. However, I guess my question is, is the fact that this coupling the Back EMF spike into the secondary's coil - eliminate the need for the traditional 'snubber' circuit or 'snubber' diode across the primary inductor that is traditionally used to 'catch' and 'handle' the large Back EMF spike? Thank you kindly for an excellent explanation - this is a new circuit to me. EDIT: and then the question was answered later in your video - thank you!
There is no back EMF spike. The "back EMF" is the induced voltage by the magnetization, When the primary is opened the back EMF find it way to the secondary which is how the energy transferred.
Thank you Sam for the video; it was incredibly helpful. I'd like to know if the support material is available. Your work is greatly appreciated, please keep up with this valuable content. Thanks once more!
Excellent video as usual professor! One question tho @6:50, if energy = power x time, then shouldn't we have P = (1/2)*[(Ipk)^2]*Lm*fsw ? (fsw - switching frequency) And then after algebraic manipulation we will have P = [1/(2*Lm)] * (Vin*ton)^2*fsw ? Thank you!
Hi, Thank you for pointing out this typo. Too much work to change, I hope people will read your correction. Note that the corrected expression is given at th-cam.com/video/laHN4Xn59lI/w-d-xo.html
Dear professor, thanks for your excellent video. There is one thing bothering me for a while: the DCM flyback converter looks like a 1st order system with a single pole and a single zero as you illustrated. Why compensation is needed? As far as I know a 1st order system is always stable.
Good point regarding stability. But...A PI controller will increase the loop gaiat low frequency for lower DC error and better line regulation as well as PSRR.
@@sambenyaakov I see. So the plant is a first order system. With an error-feedback network the system becomes a higher order system and hence requires compensation. Another question is most DC-DC converter assumes a constant output voltage while the input voltage might vary within a certain range. Is there a boost converter scheme that's suitable for constant-input-voltage and variable-output-voltage? So far the best candidate I found is the constant-off-time hysteretic controller as it doesn't require a compensation network. But I couldn't find many controllers of this type.
Hello Ben, it's nice presentation but @9.38 there is a missed T (fclk) in P formula like this: P=Ipk*Ipk*Lm /2*Tm without T we speak energy as previosly you mentioned
Thank you for this wonderful explanation. Query: Wouldn’t it be more efficient to utilize the back EMF of the secondary coil by dumping it into a capacitor which could then be used to drive the primary circuit? In theory you might be able to double the output of the power by using the resonant frequency, where the voltage peaks are maximized, and instead of dumping that power into a resistor, which turns it into wasted heat energy, you could use it to drive the primary coil or even dump it into a load? Just a thought…
@@sambenyaakov The ring down effect (back EMF) which starts as a voltage spike is typically maximized when the coil is tuned to its resonant frequency. In stead of dumping that into a resistor, you could send it to a separate capacitor bank or something and use it to power a load or to maybe drive the primary coil. That was my thought…
Hello sir, If possible, could you please throw some light or make a video on flyback operation when Vout< nVin. I am confused about its operation in that region. I found through simulations that, for Vout< n Vin, even if its designed for DCM, it shows CCM operation. Once it crosses that threshold it shows DCM operation. I am working on capacitor charging application. It would be also helpful, if you could include how parasitic capacitance play an important role in the maximum voltage the converter could reach?. If its not possible, recommendations of good reference books would be big help. Thanks a lot.
Dear Professor Yaakov, in calculating power P in the primary, it seems you have considered ton = 1/2 T and if so, P = 2 IPK^2 Lm * fs. In other words, 2 in the denominator should go to the numerator. And P as you know is the peak power. Thanks
@@raymonda.koosha5824 The energy transferred from primary to secondary is LIpk^2/2. This occurs fsw (1/T) times per second. So the power transferred is the energy times fsw (or divided by T). The duty cycle does affects the value of Ipk.
hi , how a buck or boost or flyback converter ic supply its own circuit ? I mean I always see an internal voltage regulator inside every diagram of these chips but if it get the current from the source directly it is not efficient
This is the best ever tutorial I have found on the internet, thanks, professor! But I have some questions regarding the 10:15 , Ipk = Vin*Ton / Lm, if the longer Ton looks like will result in larger Ipk. and if we have longer Ton, the Ipk value may vary but I think the Ip*Ton area may still be the same so why rms will be different? get confused here..
@@sambenyaakov thanks for explanation, I have draw the illustration and understand it now. You are the best teacher ever, even better than my professor 🤣
Hello sir! Its indeed an enlightening video. I have one doubt though, the clamp capacitor is initially charged to the clamp voltage during turn on of the switch. when the switch is turned off and when the switch voltage exceeds the clamp voltage , the diode is turned on and leakage energy has new path. My doubt is through which path the clamp capacitor is initially charged, as the diode is reverse biased. Could you please throw some light ? also in 20.35 you have mentioned ideally the switch voltage goes to zero when the current ceases in secondary. but shouldn't the switch voltage be equal to Vin ? Thank you.
Good questions. Thanks. Primary and secondary act also as a tranformer. If the cap voltage is low it will charge concurrently with the output. When the current stops the interwinding cap an primary resonant oscillate which reduces the voltage on fet.
Hi Sir , At 16:47 you were saying that time constant for the RC circuit should be larger than fs -> Hence basically in passive clamp you are not settling the RC circuit charge and discharge ; Is the passive clamp circuit just for passing those inductor (LC) undamped oscillations?
(Lm . Ip^2)/2 is the energy equation. Power equation must be multiplied with switching frequency. Am i right? Where as through out the video energy equation is mentioned as power.... If I'm right...
Can the peak current also be calculated the following way? Ipk= ton*Vin/Lm? Because the leakage current should be the same as the current through the coupled inductor, right? greetings Nick
Hi Sir, you are a great professor in making complex concepts easy. I need your favor. I am a junior research engineer. I am trying to develop a flyback converter just for understanding purposes. Our goal is something different. Flyback have following specs: Np=30T, Ns=2500T, fs=34kHz, D=47%, Vin=20V, RL=1Meg. Leakage Inductance Llk=200 uH. The Core is UU Core of MnZn Material having Effective Core area of Ae=191mm^2. Both primary and Secondary Windings are wound on separate legs. Primary Winding is of Single Copper Conductor of AWG=25. The issue which I am facing is that during ON Time (Ton), the inductor current doesn't increase linearly rather it oscillates and it seems that there are sinusoidal oscillations which are superimposed on the linearly Increasing Inductor Current. Due to this reason Output Voltage is much smaller in amplitude than the theoretical one. can you Plz explain the reason and Phenomenon which causes inductor current Oscillating. and also how can we solve that problem. Any helpful Information will be Appreciated. A Rough Plot of Primary Current During Ton is attached herewith: drive.google.com/file/d/1HHPh5b1vHrkywzdQzuXvkwrtDxF-Ipu2/view?usp=sharing
Hi Imran, You do have too much leakage inductance and apparently large inter winding capacitance. Not a good idea for flyback to wind on separate legs. Try as a first step to wind primary on secondary (with proper insulation in between).
@@sambenyaakov Yes you are right sir. Leakage inductance is large. Ok I try to wind on same leg. I will update you about the results. Thank you so much sir.
@@sambenyaakov Hi sir I have simulated the same circuit on LT Spice and added a stray capacitance of 35pF on secondary side before secondary Diode. The results are almost matching with those of hardware. This shows that oscillations in primary current are the results of large inductance and capacitance. They are making resonance. But my question is that how they are making resonance during On Time while leakage inductance is still getting energy from Source?
@@sambenyaakov Hi Sir! I have redesigned Flyback with Secondary and primary on same leg of UU core as per your instructions. Following are the specs: Vin=50V, Vo=10kV, Po=100W, fs=20kHz, D=48%, Lp= 114uH with air gap, Ro=1Meg, Np= 15T, Ns=1280T Ae=200mm^2 Now if I use same Volt/Turn on primary and secondary side then secondary turns come out to be Ns=10kV/(50/15) = 3000T But these turns are too large. So I wounded Ns=1280T. Then according to volt second equation Bac=(Vs*Toff)/(Ns*Ae) =(10kV*26U)/(1280*200U) =1.05T > 0.5T This shows that during energy transfer, core will saturate. My question is that Does above equation apply to the secondary side as well? If yes then how we can design flyback with above specs but with lower secondary turns without saturating? Also can this happen that secondary winding of flyback will saturate core even if primary winding is not saturating the core? Any help would be greatly appreciated.
Hello! Thank you very much for your lecture! Very good content! I Have question about formula for P = Ipk^2*Lm/2, because P is power, so may be right formula would be Ipk^2 * Lm* Fsw/2? So we need to product your formula to Fsw(switching frequency) ? I mean content on 6:50 min. Or I misunderstand something... Thank you!
Sir I had one question regarding the practical design of the flyback converter that I have seen in most of smps circuits. Sir there is a cermaic capacitor (1nf to 50nf) connected in parallel with the mosfet (drain-source). First I used to think it acts as a snubber during turn off of the mosfet, but soon realized that there should aslo be a snubber resistace in series with it to limit the discharging current, which is not present. I do not understand the reason behind it, as it will stress the current through the mosfet when it will turn on. Thanks.
Wow professor.. wicked videos. At 25:00 on this video you mention how you could look at the graph/poles and determine the output could be controlled using just a simple Type 2 / PI control. What concept/basis did you use to be able to determine PI vs PID (for example)? I've long forgotten my math, but more or less followed you to that point! :) Thanks for the education!
A great video as always. what happens if we run the flyback converter (or even other DC-DC converters) without load? I see that all analysis are based on the fact that there is an output (load) current and load resistor appears in some of the equations like here. what happens if there is no load? after the capacitor is charged, the diode will never conduct. Will the energy stored in the core be dissipated by heating up the core? If we add the feedback and PWM circuitry, will it even start if there is no load?
thank for the video, I want to know why you are using the average current in the calculation of the clamp current and not the RMS value. because we always use the RMS value in our calculation 18:52
Why rms? At steady state, there is about (save the ripple) a DC voltage on the capacitor so the current through the resistor is almost DC and equal to the average current o the leakage.
2 questions. I have looked at a few abstracts. On some they use the reflected voltage to calculate n and on others they use the vin over out. Which is correct? Also if Bmax is specified by the core and it is not exceeded in the calculations then why do we need the air gap. I understand that the gap changes the slope of the BH curve but is it required in all cases. Thank you
Thank you for your fast response. My understanding is not as deep as yours. In the case n=Vr/Vo vs n=Vo/Vin, they seem to yield different results for the turn ratio. Also on the second question, I have looked at your video. I am trying to design my first flyback converter to experiment with. At what point do I need to insert an air gap as part of the EE core I am using? Do I need the air gap if I stay below the Bmax?
If I have a flyback converter already operating in BCM, how do I send it to CCM or DCM without changing my input and output voltages? Just by changing duty cycle or also the inductor value must be changed?
For a fixed vin vout and power you can lower switching frequency it will go into DCM and higher frequency for CCM. But if your controller is BCM it will adjust itself again.
@8:10 Dr. Sam: If ton is higher, Duty cycle will be higher, that means peak current will be higher ( RMS will be higher) for variable switching frequency (boundary mode) as well as fixed freq DCM mode converter. Ipeak and L are inversely proportional. Or may be I am missing something here?
Thanks...I understand that you chose to use your referenced equation to solve...however...let's assume a buck converter simulation...would it not be easier to assume that one brings the inductor to full magnetic field and then at that point the charging circuit is disconnected...what ever the duty cycle time to do that...no more, no less. Then at that point in time...the problem now is simplified to that of the load circuit supplied by the flyback energy (Current source) imparted by the collapsing magnetic field...the problem then becomes much easier to conceive as well as approximate...and most likely also to simulate.
Dear Mr Yaakov, I love the way you explain electronic. I would like to point up the pole frequency. Based on your explanation this is 1/2*PI*C*(R+Rc) but the real frequency for the pole is 1/PI*C*(R+Rc). Could you explain this inconsistency? Thanks
Best Flyback converter explanation video on the internet till now . Thank you Professor Dr. Ben-Yaakov for creating all these Power electronics educational videos.
Wow, thanks!
My one regret is that my college professors were not so interested in conveying such practical, easily understandable information. I think your style and depth of knowledge is wonderful, as is your willingness to share. Thank you, Dr. Ben-Yaakov.
Hi Ken. Thanks . Comments like yours keep me going.
@@sambenyaakov As the self-teaching individual that I am, I would like to extend my deepest thanks to you and the willingness to share your knowledge. Individuals such as myself greatly rely on the quality of information conveyance that individuals like you provide us with through these videos. Again, I can't express how deeply grateful I am. Thank you Mr. Sam, James M.
@@spraynprey1044 Hi James, Thank you for the warm words. Comments like yours keep me going.
Thank you prof. for these lectures in English,I'm taking power converter course and my prof. suggested us to watch your lectures to get more clarity and I agree to him, these lectures definitely provide good clarity of the concepts.
Thanks for note. At what university are yo studying? And may I ask, what is the name of the Professor?
@@sambenyaakov Thank you for replying sir, I'm final year Btech EE student at Indian institute of technology Dharwad, my Prof. name is Dr.Satish naik
I am an electronician and I admit that you explain wonderfully well. The teachers I had dreaded switching power supplies. Relatively complex concepts have become simple thanks to you.
Glad to read that. Thanks.
شكرآ جزيلا على هذا الفيديو المفيد جدآ. تحياتي لك من تركيا.
Thank you very much.
Thank you. I learned a lot.
This was much better understanding then just throwing a coil and controller on a board 👍
👍😊🙏
Thank You so much sir for this wonderful video
Thanks
Truly grateful that you keep producing these videos!
Thanks.
Great introduction to the Flyback. I found also very useful the practical approach given to the presentation. Thank you for sharing your knowledge with us.
Thanks for comment.
Thank you for your sharing Dr. Ben-Yaakov. 谢谢您!
Welcome!
Thank you, Dr. Ben-Yaakov! This knowledge has been a critical stepping stone in the overall re-design of our EE senior project.
😊
Enourmously information-packed videos!
Thanks
Very Nice Video Sir. Grateful to you for imparting such knowledge to us.
Best Regards,
Sanjay
Thanks for taking the time to write the note. Comments like yours keep me going.
I totally agree with Ken Coffman!! Thank you Dr. Yaakov
Thanks
I thought I knew this - but I was missing a few things - this really is great stuff !
👍😊
Thank you so very much!
I'm making a BCM flyback myself and this is a great help!
Great.
Hello! Good explanation! I`d like to add about a self-oscillating flyback converters with two transistors (like in a cheap mobile charger). These converters very dependents on a transformer , feedback winding turn ratio, there may be over-voltages or avalanche damage - because there is not any gate driver. The stability of this type of converter is not good enough - there is a hysteresis in feedback loop. I have an experience with this type of flyback converter - I only want to share my small experience ))) .Maybe I am not right - but integrated circuits with gate drivers and protection much increase reliability of a flyback converter .
Thanks for sharing. Discrete analog controllers for flyback and other converters are ingenious but passe. Indeed, IC controllers are very effetive and relaiable
Awesome professor!I learned a lot from your video, thank you!
Thanks
Thank you very much Dr. Yaakov!
😊
Dear Mr Yaakov, I think that at 10mn, the fs frequency is missing in the peak current relationship when you are expressing the power.
Thank you for this sharing
Hi Mlaid, You are correct! I hope others will see your comment. The initial expression for P is actually energy so to get power you need, as you indicated, to multiply it by the switching frequency. I will put your note and my answer in the TH-cam page of video. Thanks again.
When the switch is opened up in the primary circuit and the magnetic field around the primary inductor collapses, what becomes of the Back EMF spike? Obviously the large Back EMF spike will be coupled over onto the secondary inductor. However, I guess my question is, is the fact that this coupling the Back EMF spike into the secondary's coil - eliminate the need for the traditional 'snubber' circuit or 'snubber' diode across the primary inductor that is traditionally used to 'catch' and 'handle' the large Back EMF spike?
Thank you kindly for an excellent explanation - this is a new circuit to me.
EDIT: and then the question was answered later in your video - thank you!
There is no back EMF spike. The "back EMF" is the induced voltage by the magnetization, When the primary is opened the back EMF find it way to the secondary which is how the energy transferred.
Thanks Professor Ben, I have to design an 80W isolated power supply for IGBT gate drivers. This video is very useful.
Thanks for comment?
Thank you sir. this video help me in designing fly-back.
👍
Hi Professor, This is really great content. Please consider creating a video on a comparison of all major derived converters. Thanks
Look up my TH-cam channel there are videos on many other typologies
@@sambenyaakov
In solving the DC-DC converter problems how can we determine the mode of operation (continuous, discontinuous, or boundary) ?
Nicely explaination.
😊
@@sambenyaakov May GOD Bless you alway.
Oh thanks a lot professor. I'm going to watch it
Great video! Perfect teaching skills.
Nicely explained. One suggestion is to add a working example or link to someone going through the process of applying the principles presented.
Good point. Thanks. Will try next time.
Thank you Sam for the video; it was incredibly helpful. I'd like to know if the support material is available. Your work is greatly appreciated, please keep up with this valuable content. Thanks once more!
Thanks for comment. Please look up my youtu e channel for related videos
Very nice video. Typically DCM transformer core is smaller than CCM core!
See th-cam.com/video/-SxEo_Usj0g/w-d-xo.html , The qestion is how do you compare: same losses? same switching frequency? Both?
Excellent video as usual professor!
One question tho @6:50, if energy = power x time, then shouldn't we have P = (1/2)*[(Ipk)^2]*Lm*fsw ? (fsw - switching frequency) And then after algebraic manipulation we will have P = [1/(2*Lm)] * (Vin*ton)^2*fsw ? Thank you!
Hi, Thank you for pointing out this typo. Too much work to change, I hope people will read your correction. Note that the corrected expression is given at
th-cam.com/video/laHN4Xn59lI/w-d-xo.html
Thank you
Thanks
Dear professor, thanks for your excellent video. There is one thing bothering me for a while: the DCM flyback converter looks like a 1st order system with a single pole and a single zero as you illustrated. Why compensation is needed? As far as I know a 1st order system is always stable.
Good point regarding stability. But...A PI controller will increase the loop gaiat low frequency for lower DC error and better line regulation as well as PSRR.
@@sambenyaakov I see. So the plant is a first order system. With an error-feedback network the system becomes a higher order system and hence requires compensation. Another question is most DC-DC converter assumes a constant output voltage while the input voltage might vary within a certain range. Is there a boost converter scheme that's suitable for constant-input-voltage and variable-output-voltage? So far the best candidate I found is the constant-off-time hysteretic controller as it doesn't require a compensation network. But I couldn't find many controllers of this type.
Hello Ben, it's nice presentation but @9.38 there is a missed T (fclk) in P formula like this: P=Ipk*Ipk*Lm /2*Tm without T we speak energy as previosly you mentioned
Yes, you are right but this typo has already been noticed and it is shown in the description part of video.
Beautiful!
😊🙏
Thank you for this wonderful explanation. Query: Wouldn’t it be more efficient to utilize the back EMF of the secondary coil by dumping it into a capacitor which could then be used to drive the primary circuit? In theory you might be able to double the output of the power by using the resonant frequency, where the voltage peaks are maximized, and instead of dumping that power into a resistor, which turns it into wasted heat energy, you could use it to drive the primary coil or even dump it into a load? Just a thought…
Thanks for the comment. Not sure I follow you.
@@sambenyaakov The ring down effect (back EMF) which starts as a voltage spike is typically maximized when the coil is tuned to its resonant frequency. In stead of dumping that into a resistor, you could send it to a separate capacitor bank or something and use it to power a load or to maybe drive the primary coil. That was my thought…
@@BrentLeVasseur Got it. See: th-cam.com/video/ySC-SvoQa3U/w-d-xo.html
Thanks for uploading, please make a video on closed-loop design of buck-boost converters.Thanks.
Perhaps in the future. Thanks.
Hello sir, If possible, could you please throw some light or make a video on flyback operation when Vout< nVin. I am confused about its operation in that region. I found through simulations that, for Vout< n Vin, even if its designed for DCM, it shows CCM operation. Once it crosses that threshold it shows DCM operation. I am working on capacitor charging application. It would be also helpful, if you could include how parasitic capacitance play an important role in the maximum voltage the converter could reach?. If its not possible, recommendations of good reference books would be big help. Thanks a lot.
Send me an email to sby@bgu.ac.il and I will reply.
@@sambenyaakov thank you.
Dear Professor Yaakov, in calculating power P in the primary, it seems you have considered ton = 1/2 T and if so, P = 2 IPK^2 Lm * fs. In other words, 2 in the denominator should go to the numerator. And P as you know is the peak power. Thanks
Hi, Raymond. Please clarify to which minute in the video are you referring to?
@@sambenyaakov This is at 4:38/25:40 on Slide # 5 Thank you so much for sharing such educational videos. Very Best Regards
@@raymonda.koosha5824 The energy transferred from primary to secondary is LIpk^2/2. This occurs fsw (1/T) times per second. So the power transferred is the energy times fsw (or divided by T). The duty cycle does affects the value of Ipk.
Great video, thank you professor
Thanks for comment.
Very good video , Thanks a lot .
😊
Hi, Dr. Ben-Yaakov. What drawing tool are you using for this presentation? it looks really clear and simple like Texas Instrument's datasheets.
Visio for basic drawings and a tablet with pen
@@sambenyaakov Thanks for your reply. Do you mean that you have your own stencil? It looks really good.
@@kkh8623 Thanks. Yes, we prepared templates for various components and then copy and paste in Visio,
hi , how a buck or boost or flyback converter ic supply its own circuit ? I mean I always see an internal voltage regulator inside every diagram of these chips but if it get the current from the source directly it is not efficient
Auxilliary power supply
This is the best ever tutorial I have found on the internet, thanks, professor! But I have some questions regarding the 10:15 , Ipk = Vin*Ton / Lm, if the longer Ton looks like will result in larger Ipk. and if we have longer Ton, the Ipk value may vary but I think the Ip*Ton area may still be the same so why rms will be different? get confused here..
RMS is not related to the average. In a triangle waveform
Irms=Ipk/(Sqrt3)
@@sambenyaakov thanks for explanation, I have draw the illustration and understand it now. You are the best teacher ever, even better than my professor 🤣
@@pkhugovan 👍🙏😊
Very nice!
Thanks
Hello sir! Its indeed an enlightening video. I have one doubt though, the clamp capacitor is initially charged to the clamp voltage during turn on of the switch. when the switch is turned off and when the switch voltage exceeds the clamp voltage , the diode is turned on and leakage energy has new path. My doubt is through which path the clamp capacitor is initially charged, as the diode is reverse biased. Could you please throw some light ? also in 20.35 you have mentioned ideally the switch voltage goes to zero when the current ceases in secondary. but shouldn't the switch voltage be equal to Vin ?
Thank you.
Good questions. Thanks. Primary and secondary act also as a tranformer. If the cap voltage is low it will charge concurrently with the output.
When the current stops the interwinding cap an primary resonant oscillate which reduces the voltage on fet.
@@sambenyaakov Thank you sir.
Hi Sir , At 16:47 you were saying that time constant for the RC circuit should be larger than fs -> Hence basically in passive clamp you are not settling the RC circuit charge and discharge ; Is the passive clamp circuit just for passing those inductor (LC) undamped oscillations?
Indeed
Hi, is there a way to reduce the resonance which happens after the switch go off?
By a snubber
very good
Thanks.
(Lm . Ip^2)/2 is the energy equation.
Power equation must be multiplied with switching frequency. Am i right?
Where as through out the video energy equation is mentioned as power.... If I'm right...
Can the peak current also be calculated the following way? Ipk= ton*Vin/Lm? Because the leakage current should be the same as the current through the coupled inductor, right? greetings Nick
Certainly, but..you don't know ton. The requirement is P from which you calculate everything including ton.
Hi , is Lm the mutual inductance or primary coil inductance ?
Magnetization.
could you please make video about the feedback and PWM circuitry that controls the switching in any flyback converter?
Hi Sir, you are a great professor in making complex concepts easy. I need your favor. I am a junior research engineer. I am trying to develop a flyback converter just for understanding purposes. Our goal is something different. Flyback have following specs: Np=30T, Ns=2500T, fs=34kHz, D=47%, Vin=20V, RL=1Meg. Leakage Inductance Llk=200 uH. The Core is UU Core of MnZn Material having Effective Core area of Ae=191mm^2. Both primary and Secondary Windings are wound on separate legs. Primary Winding is of Single Copper Conductor of AWG=25. The issue which I am facing is that during ON Time (Ton), the inductor current doesn't increase linearly rather it oscillates and it seems that there are sinusoidal oscillations which are superimposed on the linearly Increasing Inductor Current. Due to this reason Output Voltage is much smaller in amplitude than the theoretical one. can you Plz explain the reason and Phenomenon which causes inductor current Oscillating. and also how can we solve that problem. Any helpful Information will be Appreciated. A Rough Plot of Primary Current During Ton is attached herewith:
drive.google.com/file/d/1HHPh5b1vHrkywzdQzuXvkwrtDxF-Ipu2/view?usp=sharing
Hi Imran, You do have too much leakage inductance and apparently large inter winding capacitance. Not a good idea for flyback to wind on separate legs. Try as a first step to wind primary on secondary (with proper insulation in between).
@@sambenyaakov
Yes you are right sir. Leakage inductance is large. Ok I try to wind on same leg. I will update you about the results. Thank you so much sir.
@@sambenyaakov
Hi sir
I have simulated the same circuit on LT Spice and added a stray capacitance of 35pF on secondary side before secondary Diode. The results are almost matching with those of hardware. This shows that oscillations in primary current are the results of large inductance and capacitance. They are making resonance. But my question is that how they are making resonance during On Time while leakage inductance is still getting energy from Source?
Inter winding capacitance of magnetic element
@@sambenyaakov
Hi Sir!
I have redesigned Flyback with Secondary and primary on same leg of UU core as per your instructions. Following are the specs:
Vin=50V, Vo=10kV, Po=100W, fs=20kHz, D=48%,
Lp= 114uH with air gap, Ro=1Meg, Np= 15T, Ns=1280T
Ae=200mm^2
Now if I use same Volt/Turn on primary and secondary side then secondary turns come out to be Ns=10kV/(50/15) = 3000T
But these turns are too large. So I wounded Ns=1280T. Then according to volt second equation
Bac=(Vs*Toff)/(Ns*Ae)
=(10kV*26U)/(1280*200U)
=1.05T > 0.5T
This shows that during energy transfer, core will saturate.
My question is that Does above equation apply to the secondary side as well? If yes then how we can design flyback with above specs but with lower secondary turns without saturating? Also can this happen that secondary winding of flyback will saturate core even if primary winding is not saturating the core? Any help would be greatly appreciated.
Hello! Thank you very much for your lecture! Very good content! I Have question about formula for P = Ipk^2*Lm/2, because P is power, so may be right formula would be Ipk^2 * Lm* Fsw/2? So we need to product your formula to Fsw(switching frequency) ? I mean content on 6:50 min. Or I misunderstand something... Thank you!
Yes, correct. This is pointed out in the TH-cam page of this video.
The secondary inductor should be formed as Lm/n^2, am I right?
To which minute/slide_number are you referring to, please?
@@sambenyaakov Sorry, I thought the turns ratio Np:Ns is represented by n:1, it's 1:n in your presentation.
Sir I had one question regarding the practical design of the flyback converter that I have seen in most of smps circuits. Sir there is a cermaic capacitor (1nf to 50nf) connected in parallel with the mosfet (drain-source). First I used to think it acts as a snubber during turn off of the mosfet, but soon realized that there should aslo be a snubber resistace in series with it to limit the discharging current, which is not present. I do not understand the reason behind it, as it will stress the current through the mosfet when it will turn on. Thanks.
Tis would be for DCM. The reason could be either helping quasi resonant operation of lower the frequency of the parasitic oscillations.
@@sambenyaakov Thankyou Sir!
Hi Mr Yaakov, I want to have two output for my application ( eg: +-5V). So, is there any difference in procedure of calculation?
thanks a lot!
You need to take into account the two secondaries as explained in video.
Wow professor.. wicked videos. At 25:00 on this video you mention how you could look at the graph/poles and determine the output could be controlled using just a simple Type 2 / PI control. What concept/basis did you use to be able to determine PI vs PID (for example)? I've long forgotten my math, but more or less followed you to that point! :) Thanks for the education!
@25, this is because a current source into an RC is first order.
Aha! Thanks a bunch professor!
What do you mean by offline and online application ?
Which minute in video do you refer to?
Thanks Prof. !!
Thanks
Just looked at your channel. I do not speak the language but it looks impressive.
A great video as always. what happens if we run the flyback converter (or even other DC-DC converters) without load? I see that all analysis are based on the fact that there is an output (load) current and load resistor appears in some of the equations like here. what happens if there is no load? after the capacitor is charged, the diode will never conduct. Will the energy stored in the core be dissipated by heating up the core? If we add the feedback and PWM circuitry, will it even start if there is no load?
thank for the video,
I want to know why you are using the average current in the calculation of the clamp current and not the RMS value. because we always use the RMS value in our calculation 18:52
Why rms? At steady state, there is about (save the ripple) a DC voltage on the capacitor so the current through the resistor is almost DC and equal to the average current o the leakage.
2 questions. I have looked at a few abstracts. On some they use the reflected voltage to calculate n and on others they use the vin over out. Which is correct? Also if Bmax is specified by the core and it is not exceeded in the calculations then why do we need the air gap. I understand that the gap changes the slope of the BH curve but is it required in all cases. Thank you
The two approaches seem to be the same. As for need of gap see
th-cam.com/video/CCS_sPgb6U4/w-d-xo.html
th-cam.com/video/z17qFQ12CfE/w-d-xo.html
Thank you for your fast response. My understanding is not as deep as yours. In the case n=Vr/Vo vs n=Vo/Vin, they seem to yield different results for the turn ratio. Also on the second question, I have looked at your video. I am trying to design my first flyback converter to experiment with. At what point do I need to insert an air gap as part of the EE core I am using? Do I need the air gap if I stay below the Bmax?
If I have a flyback converter already operating in BCM, how do I send it to CCM or DCM without changing my input and output voltages? Just by changing duty cycle or also the inductor value must be changed?
For a fixed vin vout and power you can lower switching frequency it will go into DCM and higher frequency for CCM. But if your controller is BCM it will adjust itself again.
Very nice presentation Schmuel. Are you planning a separate lecture for active clamp operation ? Also can you post lecture on phase shift converter ?
Yes indeed. Thanks for comment.
@8:10 Dr. Sam: If ton is higher, Duty cycle will be higher, that means peak current will be higher ( RMS will be higher) for variable switching frequency (boundary mode) as well as fixed freq DCM mode converter. Ipeak and L are inversely proportional. Or may be I am missing something here?
Hi Swaraji, if by design ton is longer , Lm will have to be larger for same power. The end result will be a lower peak current
Sam Ben-Yaakov yes you are correct Dr.Sam. I checked my equations . Great video by the way
Thanks
Why the stored energy inside choke should not be 1/2*Lm*Iavg instead of peak current? what if the current shape not be triangular in general not here
Energy stored in an inductor is 1/2 L I^2 so the energy of the leakage inductor at he turn off instance is related to Ipk of main inductor.
thank you for reply
Thanks...I understand that you chose to use your referenced equation to solve...however...let's assume a buck converter simulation...would it not be easier to assume that one brings the inductor to full magnetic field and then at that point the charging circuit is disconnected...what ever the duty cycle time to do that...no more, no less. Then at that point in time...the problem now is simplified to that of the load circuit supplied by the flyback energy (Current source) imparted by the collapsing magnetic field...the problem then becomes much easier to conceive as well as approximate...and most likely also to simulate.
This is too simplistic Switch mode converters do not work this way.
@@sambenyaakov Deflection is simplistic as well
is this [L1 = ( A*N^2*m/l )] the equition to find the indactans in the flyback converter??
Yes. This is the equation for the inductance of any inductor.
thank you very very much professor
DCM Flybacks use a smaller core size then comparable CCM Flybacks or Forward Converters...
can we use same method for CCM ??
Yes. With adjustments.
@@sambenyaakov thank you so much sir
I want to design a dc- dc converter for 180w rating how bad is that using flyback topology for that.
Dear Mr Yaakov, I love the way you explain electronic. I would like to point up the pole frequency. Based on your explanation this is 1/2*PI*C*(R+Rc) but the real frequency for the pole is 1/PI*C*(R+Rc). Could you explain this inconsistency? Thanks
Please indicate minute in video you are referring to.
prof.could you tel me how i can choose the value of Lm ?
Lm in CCM is a function of target ripple. in DCM it is related to power transfer.
how can i choose the value of the Capacitor?
Which cap?
@@sambenyaakov the capacitor at the output
Based on current ripple of cap and desired voltage ripple. But...watch for losses due to ESR.
what do you mean by Lm??
This is the inductance id coupled inductor. Yo can define it from primary or secondary.
excuse this sample just is is theory in action problem is not simple.please in action fasten circuit an show to us .thank you
Can this flyback model exchange energy 2 way sir? I plan to use it to balance the voltage between two 15V sources
For bidirectional you need also a switch at the secondary.
@@sambenyaakovThank you for the answer sir, can you tell me if the calculation formulas are different with that model?