Power Electronics - Resonant Converters - Intro
Vložit
- čas přidán 30. 07. 2024
- This is the introduction to our video sequence on resonant DC-DC conveter. We focus our analysis on series LC and series LLC converters. Below are references where you can find more information on this topic.
"Survey of Resonant Converter Topologies," by Sheng-Yang Yu, Runruo Chen and Ananthakrishnan Viswanathan, Reproduced from 2018 Texas Instruments Power Supply Design Seminar SEM2300, SLUP376, Texas Instruments, 2018.
"Designing an LLC Resonant Half-Bridge Power Converter," by Hong Huang, Reproduced from 2010 Texas Instruments Power Supply Design Seminar SEM1900, SLUP263, Texas Instruments, 2010.
"Basic Principles of LLC Resonant Half Bridge Converter and DC/Dynamic Circuit Simulation Examples," by OnSemi, Application Note AND9408/D, 2016.
www.onsemi.com/pub/Collateral...
"Resonant LLC Converter: Operation and Design, 250W 33Vin 400Vout Design Example," by Sam Abdel-Rahman, Infineon Technologies, Application Note AN 2012-09 V1.0, 2012.
"An Introduction to LLC Resonant Half-bridge Converter," by ST Micro, Application Note AN2644, Rev 2, 2008.
www.st.com/resource/en/applic...
"Understanding and using LLC Converters to Great Advantage," by Sanjaya Maniktala, Microsemi, 2013.
www.microsemi.com/document-po...
this was actually very well explained probably the best version on YT.
Yaghiyah, thank you very much. Best wishes on your design. -Dr. K
a trick: you can watch series on Flixzone. Been using it for watching all kinds of movies during the lockdown.
@Clay Zakai Yea, I've been watching on flixzone for since november myself =)
I need to self-learn resonant converter, so this is a new concept to me. I always have to find this video when I forget how a resonant converter works. Thank you for a very nice explanation.
keep the great work going on. Thanks
Please wathch the music . Some people are listening on headphones . Never open with loud music !
Great introduction to resonant converters.
Thank you.
Very pure explanation. I hope you will make videos about it. These informations and you are very important.
Thank you Oguzhan. Best wishes on your design. -Dr. K
Great explanation........
Thanks Power Electronics 🙏🌹
Great experience
Explained well thank you
You are welcome. Thank you for watching. -Dr. K
Sir,
Great video. Great explanation. Salute for this.
Please make some more videos on power electronics topic to educate us.
Ranjeet, thank you for the kind words. I am glad this was helpful. -Dr. K
Great explanation love from India 👍🏽👍🏽👍🏽
Thank you.
So nice the world's best teacher
Thank you for the kind words. Best wishes on your design. -Dr. K.
Very good sir 👍
Thank you.
buen video 😊
Hello sir, could you please tell us what is the difference between filters, snubbers and resonant converters?
Wonderful explanation! Thank you for that! Just one small question: in the second phase that you've mentioned, just before the switching between M2 to M1, when the reverse current "uses" the body diode of M1 to hold its previous direction, I believe the polarization of the voltage over the body diode is on the other direction (opposite to the current), isn't it?
Yes, you are correct the polarization is opposite. It will be about 0.5V anode to cathode (source-drain). It is still near zero volts. This is a bit of an over simplification of the ZVZC characteristics for the LLC. Some of the source I have listed in the description section do a much better job of explaining the ZVZC necessary conditions. Thank you for asking and watching. -Dr. K
Excellent 👍
Thank you. -Dr. K
Thanks for the great video. Just to make sure I'm clear on this, with resonant circuits, we are switching at either zero-current or zero-voltage. As long as we switch below the resonant frequency, this condition will be achieved? That is, we do not need EXTERNAL sensors to measure current or voltage to waiting until zero-crossing in order to determine when to turn on/off the MOSFETs?
Hi Xichen, unfortunately its not that easy of an answer. Ideally ZVS and ZCS condition is only truly achieved at resonance. In a practical design, you will never be exactly at resonance because the load changes, The load impacts the Q factor and changes your resonant frequency. Typically you will want to be switching and operating in the inductive region (above resonant frequency( as this will keep current flowing in such a way as to create a ZVS condition when turning on the MOSFETs. Here's a very good description from TI. Please see the section on ZVS, www.ti.com/lit/pdf/slup263. I hope this helps. -Dr K
@@powerelectronicswithdr.k1017 Thank you.
great work giving the introduction to resonant converter.
sir i want to increases the voltage from 390v to 1000v is this converter is suitable for this design and the converter having a power of 1200watt rating. or any other converter is suitable let me know sir. i want to use isolation converters only .
Hi Abhishek, there are numerous ways to adjust and regulate the voltage output for a resonant converter. I suggest starting with the turns ration for the transformer. If n < 1, you can step-up (boost) the output voltage. Here's a link to a very nice spreadsheet from Texas Instruments that helps with all the different design equations for an LLC converter www.ti.com/tool/UCC25600-DESIGN-CALC
@@powerelectronicswithdr.k1017 thank you sir .
i will check the design once and i will come back to you sir. instead of center tap transformer i can use without center tap transformer only for step up application is it possible?
Very well explained! Something that I don't understand is how you could switch the mosfet exactly when we are are near 0 current. Does it do it automatically? What would be connected to the gate of the mosfets?
No. The frequency of the H-bridge is selected to be at is at resonance frequency of the LC or LLC circuit. Hope this helps and thank yo for watching. -Dr. K
Great
What about the output is the Vin ~= Vout ?
Thank you for the superb explanation, sir !! I want to know do all the resonant converters provide soft-switching inherently?
Paramanand, you are welcome. Yes, but there are some necessary and sufficient conditions for soft-switching. Here's and excellent reference www.ti.com/seclit/ml/slup376/slup376.pdf titled "Survey of Resonant Converter Topologies," by Sheng-Yang Yu, Runruo Chen and Ananthakrishnan Viswanathan from Texas Instruments. Here you can find the 3 conditions that must be met in resonant converters for soft-switching. Best wishes on your designs. -Dr. K
@@powerelectronicswithdr.k1017 Thank you so much, Sir !!
Sir how to calculate Ls and Cs value i have a smps transformer ,design for 37KHz having Lm=0.348mH output inductance 18mH and 18mH (two secondary) and leakages is aprox 1.5uH. Primary having turn 8 and secondary having 57 and 57 turn on each limb. Input supply voltage 565vdc.
Hi Anil, I'm going to provide the following design tool from TI. This spread sheet is designed for their UCC25600 controller and automatically calculates the series capacitance, inductance, magnetizing inductance and turns ratio of the transformer. Typically, you will start your design with Input voltage range, maximum output power, and output voltage. You will also need to select a switching frequency. Your frequency is 37kHz. www.ti.com/tool/UCC25600-DESIGN-CALC . Best wishes on our design. -Dr. K
@@powerelectronicswithdr.k1017 thankyou sir one more question can i design LLC resonance convertor full bridge with IC sg3525 since it is an invertor is and i have design transformer of Lm=350 uH primary turn 8 two secondary 57 ,57 turn i have to find Ls Cs to charge capacitor in 1.5 msec @15kv... pls suggest
What I missing from this lecture, how the output voltage looks like on the secondary, if filter capaticitor is not available, only the load exist.
Why the Reference Manual of Onsemi says, that this converter is difficult to operate with light loads?
Hi John, for your first question, if the filter capacitor on the output is removed you would obtain a rectified sinusoidal waveform across the load assuming you are switching at or near the resonant frequency of operation. That frequency is dependent on the load, which we model as resistive. This leads to your second question. If the load is "light" then R_LOAD will be modelled with a high resistance (i.e. low output current). From equation 1 in the OnSemi paper you will find the Rac (load reflected back to the primary) will also have a larger value and this will reduce the resonant current through primary side of the transformer. There is a minimum current value which is required to overcome the magnetizing current of the transformer. If there current is too low, the result is discontinuous conduction mode (DCM). Also, the value of Q becomes low. The ST micro reference paper does a nice job of explaining some of the different DCM modes of the LLC converter and the implications of no-load operation. Operation and control of an LLC resonant converter is very complex and unfortunately this video just scratches the surface. Thanks for the questions. -Dr.K.
Hi, thanks for the video. Where can I find the next videos about this topic as you mentioned in your introduction?
czcams.com/video/V02dYsARF2g/video.html, czcams.com/video/YrDXj-yTNaM/video.html, czcams.com/video/6aaPwwkEVeQ/video.html
One of the best videos, could you please explain how to check and see the current wave in the primary side and check if it's sinusoidal? do I need a current probe? can you suggest one? and point exactly where to measure the current wave?
Is your question based on an actual circuit or on how to investigate the current via SPICE simulation? For a real circuit, the answer is "it depends." There's two main ways for measuring current. One method is to use a current-sense resistor and the other method is to use a Hall-effect sensor. There are pros & cons to each method. Unfortunately, I do not have an easy answer for your question. Best wishes. -Dr. K
@@powerelectronicswithdr.k1017 It's Ok you don't have an answer for my question, but which points in the circuit to probe for the current wave?
Please check the voltage drop polarity across diode at 8.33 it should be reversed
Jacob, great catch! Yes, the polarity of the voltage should be reversed. Thank you for watching. -Dr. K
Sir, if we will use a MOSFET instead of diodes on the secondary side of the transformer (i.e. on the rectifier side), then what will happen ? I mean how will we achieve ZVS or ZCS in that case?
Hi Paramanand, great question. The use of MOSFETs instead of diodes is called synchronous rectification and is very common for achieving higher efficiency. Luckily we will have near ZCS on the secondary side. If you look at the current flow through the secondary coil(s) and you are switching the half-bridge near the resonant frequency, the current will be sinusoidal. Therefore, the secondary side diodes commutate at near zero current. Therefore the loss in the secondary is primarily conduction based. To reduce this we can use MOSFETs and utilize the body diode to reduce the the voltage across the drain to source when we turn the MOSFET on and with the lower Rds, reduce the conduction loss in the secondary rectifier circuit. Hope this helps and thank you for the question. Best wishes on your design. -Dr. K.
@@powerelectronicswithdr.k1017 Thank you so much sir!
Hi there Dr.K nice place here. I have a question about a LCC converter, the LLC I have some succes with but because I use high voltages and some lower voltages, I had more succes with a LCC who do a lot better regulation. I can calculate the capacitors but I have not ebough info how to do the Transformer who seems not a part of the resonance but is a load, now I do not now how to calculate that, maybe a normal half bridge transformer calculation will do? Can you explane in some words? Thanks in advance.,
Hi Audio Kees, yes the transformer will appear to the resonant circuit as a load that is in parallel with either the shunt inductor for the LLC topoloty or the shunt capacitor in the LCC topoloty. The transformer provides isolation and also enables you to either step up or down the voltage on the secondary side (i.e. boosting or bucking). Are you looking to size for the voltage boost/buck or how to reflect the DC load back to the DC/AC resonant side of the circuit? -Dr. K
Sir!! one more doubt, if a converter is not a resonant one, then to achieve the soft switching, I mean to achieve ZVS or ZCS, do we need to do some special arrangement? Like, for a simple buck converter, how can we achieve the soft switching?
For both ZVS and ZCS you typically have to be using either a resonant or quasi resonant topoloty. It is possible to get ZCS switching conditions with a simple buck converter when going into Discontinuous Current Mode (DCM). The downside is that you have often have a much larger current ripple through your inductor and your output capacitor has to supply that additional charge to the load. Hope this helps. -Dr. K
@@powerelectronicswithdr.k1017 Thank you, Sir !!
What kind of transformer contains Lm shunt inductor inside the trafo?? Who sells them? Can't you just use a separate inductor for Lm?
8:26 Where does the current come from that flow thru M2 body diode? Ground?
How do the mosfets know when to switch?
Can ZVS/ZCS be used in an LLC?
Yes, you can use a separate inductor for Lm. However, for cost reduction you can have whats called an integrated design where Lm is part of the transformer. You must control the switching of the MOSFETs. The LLC control is similar to a VCO where the output voltage controls the switching frequency of the half-bridge. ZVS and ZCS are paramount to the efficient operation of the LLC converter. You must make sure to add dead-time between switching the MOSSFETs to force the current through the body diodes. This is why you want to operate in the inductive region. Best wishes on your design. -Dr. K
Why does the current start to decrease halfway through M1 being closed?
Hi Nick, great question. First, that graph is illustrating a steady-state behavior for a constant load and we are switching at resonance. At resonance, the switching voltage and the current are in-phase. Second, we could do a Fourier Analysis of that switching voltage and see that the fundamental frequency of the voltage wave is sinusoidal at the resonant frequency (there would be a 3rd harmonic, 5th harmonic, etc). Therefore, the current is following the fundamental frequency component and the higher order odd-harmonics are being filtered out from the current (the tank circuit acts as a 2nd order band-pass filter). If we weren't switching at resonance, the current would be rising or already falling (lagging or leading) at the halfway point of M1 being closed. Hope this explanation helps. The resonant converter can be a tricky circuit to understand. Thanks for watching and best wishes on your design. -Dr.K
I have now done some tests with all kinds of resonance, I ask myself can I use a series resonant because I do need only a rock solid voltage and low emi, seems the the series resonance do have low stresses on parts and as such lower emi, because the supply,s are for audio use, and there are audio tubes involved I need a supply who do not radiate in it making things worse. An LCC is also a nice idea, LLC do not well because of the high voltages involved. The supply has to do only keep voltage on one value.
One question arrives what concern the transformer for a series resonant, what to do with the leakage inductance? need that to include in the total Lr inductance? Otherwise I get a light version of LLC.
By series resonance, I assume you mean just an series LC converter. While the output might be "rock-solid", what about the input voltage to the converter? Also, what about the output load? Is that constant too? The variation in the load will impact the range of Q that you will need to operate over. One last thing to consider, if you are using a transformer to either step up or down the voltage, the magnetizing inductance on the primary side is utilized in the design of the LLC converter. Just some things to consider. If the input, output and load do not vary drastically, then an LC series converter will work fine. Best wishes on your design. -Dr. K
@@powerelectronicswithdr.k1017 Th There Dr k, I give some more info, you are right I was not that clear. I am busy with a supply for a hybrid amplifier, the smps resonance is for the tubes part of it, but also I do have to do one for the power amp itself, for the last one I can just use a resonance supply who is open loop setting on the resonance hot spot, the feedback is not able to regulate fast enough for the audio power peaks and impulses, a open loop do, with big capacitors and a more wide soft start , I did sim also a lcc in open loop, that supply was not so happy with that ramming up voltages in low load, a audio amp has only when idle max 200 to 500 mA idle current. I have here a LCC 2 x 60 volt supply who go in burst mode when audio low level, and that give ripple problems, it has feedback so need to remove that and adjust frequency.
Oke the real question was that I need a supply for the tube part, that is a 300 watt max version with multiply output voltages, 19 volts 10 amp, 350 volts 0.12 amps, 2 x 130 volts 0.250 amps. I have more then I need for the next reason, to be able to experiment with other audio tubes, like the ones with 5 amp filament currents. So it need to regulate from minimum of 500 ma when put high voltage with it I presume that it has to work from 1 amp as min, but when a tube get defective and the load is complete gone it need also to protect itself there can a burst mode be the solution or a resonance supply who can work in open load. The input voltage is from grid without a pfc I need to be careful with EMI who can radiate into such a tube preamp.
I did sim different supply,s and came on LCC or a old fashion serial resonance system, has low stress on parts and low emi, emi is really a important factor.
In the mean time I have learn some more stuff, only the transformer making and calculate windings, inductance N was still not clear, but also here I get better, I am a very bad calculator, always as even on school, some kind of number blindness I guess.
Last question, the series resonance need a transformer with very low leakage? otherwise I get a LLC. for the LCC the transformer needs to have high reactance because it do load a current source resonance tank, I did see in sim that when I play with inductance the Q and the lower resonance did change so I need to know what inductance is best, I did see papers where se use 500uH primarily to 5000 uH values. I still have not much clearance about.
How do you think about the new gan mosfets, because these have no body diode affecting soft switching? These mosfets are very good for class D amps.
And yes I am 67 so brain is more slow, but a lot of fun.
And sorry for the long story.
thanks and regards.
In real world, it works? Because of capacitor will block the DC voltage right?
Sir, I couldn't understand that how current is being zero in switch before switching off.
The current is sinusoidal and is in phase with the fundamental harmonic of the switching voltage square wave. This is because the load looks purely resistive at this frequency. Also, the other higher order harmonics are attenuated and only the resonant frequency is well supported. There's a plot of the current waveform in the video. Hope this helps. -Dr. K
3:45 Isn't it the opposite... if n>1, we boost the system as the output voltage in the secondary is greater than the primary one. Otherwise thank you for your content!
Hi Zluckyk, you are welcome. The ratio Vp/Vs = n/1 or Vs = Vp/n. As and example, if Vp = 100V and n = 4, then Vs = 25V. So when n > 1, we are bucking (stepping down) and when n < 1 we are boosting (stepping up). Hope this helps. Best wishes on your design. -Dr. K
Thank you Dr K. It was because i learned it in this way n=Vs/Vp
what is wrong with adding too much music
If you don't care to hear the actual lesson, nothing wrong with drowning it in music.
Good Video. But tune down the music. Or leave it out altogether..
Hi Just Random, thank you for the feedback. I'll lower the volume on my bumper music for future videos but will not eliminate it. My love of music is what got me into electrical engineering and electronics well over 40 years ago. Hope the video helped. Best wishes on your designs. -Dr K
Very basic knowledge but explained in a very confusing way. Where is the reference for MOSFET switching in High Side Switching Mode? Do you have a MOSFET driver supplying pulses at the gates of M1? The back -BEM will do what? M1 will never switch unless there is a driver IC or bootstrap for M1. Yes it is adviseable to run the pulses at 90% duty cycle. But your way of explanation is vague, incoplete and not description. Instead of building an understanding, you are confusing more. I disliked your video.
Please do not confuse my audience. You only switch with a 50% duty cycle for a resonant converter. You are correct in that gate drivers are required. This video is not about MOSFET gate drivers. This an explanation of how resonant converter functions and has soft switching. Best wishes on your designs.