The confusing definition of Q in LLC converters

[@fredflickinger643]

Sam, this crystallized my research into the operation LLC Resonant Converters especially when applying the FHA. Thank you!

[@sambenyaakov]

🙏🙂

[@Appregator]

Always good to hear you explain matters. Thanks so much.

[@sambenyaakov]

Thanks

[@vbidawat93]

@21:22 I believe it's a bit confusing here.. If you want to define the max boost you can call that as Qmin if we stick to Qs (series) nomenclature for quality factor. And this Qmin will define the max Re value where we get this gain. This can set the Fmin clamp If you set that as Qmax and corresponding Remin than you can never achieve the lower gain. The ZVS explanation was bang on and awesome. And one can rest assured that if your operating point is on the right side of the peak of a given curve you'll always have ZVS.

[@sambenyaakov]

Please see just posted video kzbin.info/www/bejne/mqrGgHhsnZiIb6s

[@anshultyagi7609]

Sir you are truly an inspiration ,what a classic explanation i always learn from your lectures thank you very much for all your knowledgeable sessions 😊

[@sambenyaakov]

Thanks, so nice of you.

[@godfreypoon5148]

This presentation really resonated with me, though I find currently I am always lagging.

[@sambenyaakov]

Watch closely the video: When resonating, lagging is good😀

[@godfreypoon5148]

@@sambenyaakov 😁

[@marcinszajner2924]

Parallel (C) is great for low power HV pplication, when you could use capacitance of transformer. LLC transformer are a bit complicated. Maybe some video about how design tranformmer and coil on one core? Of course great material.

[@sambenyaakov]

Thanks Marcin. Great suggestion. Will try.

[@vatsan2483]

Wonderful presentation sir.. Thanks a lot.. Now that I think about it.. The LLC Q factor curve.. it makes me think about Qs and Qp definitions.. but do you have any hypothesis on this confusion of Qs and Qp standard not well explained and struck with that??

[@sambenyaakov]

Beats me. Very good question. I am curious too. Thanks.

[@vbidawat93]

Well the confusion is not that big a problem, as LLC Q factor is not just a simple series or parallel case. The mode fsw< fres is where the Q factor is a combination of parallel and series both, however the optimal position of operation is at fres where Qs is perfect definition of Quality factor and I believe it's the reason why this has been adapted while talking about Q in LLC

[@robson6285]

Haa great, a new videolesson! I cannot wait until i finish my new job so i can watch and learn with high Quality factor about my real interest!

[@sambenyaakov]

👍

[@audiokees4045]

I have now find the right "knowings" I hope after simulering and reading stuff, Q is just the needed part for regulation, with some carefull calculations for max cap voltage and choose to get max power versus input voltage, to get zvs also the need of choosing wise, the boundary voltage. I have for the tube amp choose the LCC, here the transformer is a separate part in the tank, while LLC transformer is included in tank. I have seen after sims that for audio amp supply,s feedback is a difficult part, the dynamics of the amp is complicated and the fast transients can not be regulated, in current feedback it is faster but still. That is why the LLC is almost Always open loop for audio power. LCC is in mine eys better for audio because it can be regulated faster and also in open outputs, afcouse the amps do always draw current most idle currents idle is 250 ma or so, class A is a different step, these draw sometimes idle currents of 1 amp or more, LLC can do here fine. The transformer calculations I get now right, have them included in design schematics ltspice. I do now now that the LCC is a current source (transformer inductance LLC is replaced by a capacitor) and transformer is high impedance as current sources like that, that is why it is so high as possible in inductance (what transformer core fit), correct me if I am wrong. I did get it work with a FHA sim asc file and a simple q/resonance program, with the simple power transfer equation, more I do not need. I do not now if I am right because you are the teacher/professor in these technology, I did see that simulating do work quite oké. Now I do understand it much better but needed a long time to get it done, maybe because I am 67, brain is shrinking. have a nice day and keep posting, you have very good relaxed movies and always a pleasure to watch. have a nice day

[@sambenyaakov]

thanks for sharing

[@chennakesavulu4288]

Hii professor, thank you for the video. I want to understand what must be ratio self resonance frequency to switch frequency of the transformer in order neglecting the parasitic capacitance of transformer.? What happens when ratio too small.?

[@sambenyaakov]

I would say at least half a decade, If too close you might have wild oscillations

[@vbidawat93]

Hello Professor Sam Ben Yakov, Could you please make a video that explains do we get ZVS/ZCS for the Secondary side MOSFET based rectifier for a LLC converter. For primary it is clear turn on is ZVS and turn off is Hard commutation, however, turn off is not much loss as the current at the time of turn off is less and equal to magnetizing current.

[@sambenyaakov]

Thanks. Good subject. Will try.

[@Superfungus0]

For the case where Q is high, and there is ambiguity in the voltage information about which direction to move frequency to achieve regulation could you use a combination of voltage and phase information to maintain control? Other than the major disadvantage of requiring a complex, nonlinear controller would there be any advantages to that system?

[@sambenyaakov]

With proper design the converter will not reach thiscsituation.

[@chiefrunningfist]

Another great vid, thank you professor. 1. Is it possible to design 1 multi resonant converter to perform over a wide load range? Say 1ohm to 750ohm? Or does this require multiple resonant converter ccts? One for each section of power output, ie small, mid, high, and a controller to select from which cct to use given the load observed? 2. Is it possible to use multiple secondary windings with a multi resonant design?

[@sambenyaakov]

1. At low power (high resistance) burst mode is probably the way to go. 2. Definitely

[@andrewandrosow4797]

Hello! As I expected - my experiment was unsuccessful. I had: Lr=7uH, Cr=100nF, Lm=25uH, Vin=16V, Vout=7V Iout=2A. The capacitor was warm and the transformer (with air gap) was quite hot. So, this topology, as I thought probably is not for low voltages and high currents .So I am going to half-bridge PWM topology with quasi-resonant serial LC circuit (a leakage inductance with a serial capacitor) - it makes current more sinusoidal with heavy loads. Maybe I have not enough experience - my degree is not in power electronics - I am an engineer in radio, television and radio communication - but I learned power electronics by myself.

[@sambenyaakov]

Thanks for sharing. Don't give up. Persistence pays.

[@salaheddinearribi4735]

شكرا يعقوب

[@sambenyaakov]

🙏

[@forestxu8369]

I think the 4 diodes rectifier is connected wrong.

[@sambenyaakov]

Indeed.

[@RaedMohsen]

While calculating the ration of Preac/Preal, why was the contribution of Cr ignored? Since it has an opposite sign it will decrease the total Preac of the tank.

[@sambenyaakov]

Hi Raed, to begin with, energy is only positive. But the more fundamental answer is that in resonant networks, the reactive energy is moving from cap to inductor, At the zero cross of current there is no energy in L and all is in C and at the peak current all energy is in the inductor. So we are dealing with one reactive power.

[@MunnaKumar-sc7qo]

Thanks for the information sir. Can you share the LTspice file

[@sambenyaakov]

Send me your email address

[@tuhinmitra935]

Hi Sir, it is a very good video, But one thing I would like to mention that @8:31 , the formula is wrong. It should be Qp/Fn = (Re/Wsw.L) not (Qp*Fn) I think. What do you think sir? Kindly let us know, please sir.

[@sambenyaakov]

Hi Tuhin, thanks for note. You are correct. It is not only the math that is wrong but also the basic intuitive assumptions are incorrect. Thank you for the observation. I am going to delete this video in a couple of days and may post a corrected one later. You can write to me at [e-mail]

[@KratosGhostofSparta-pp1cr]

Hi. Can you send me the link of Dr. Steigerwald article? thank you

[@sambenyaakov]

R. L. Steigerwald, "A comparison of half-bridge resonant converter topologies," in IEEE Transactions on Power Electronics, vol. 3, no. 2, pp. 174-182, April 1988, doi: 10.1109/63.4347. keywords: {Resonance;Topology;Power supplies;Voltage},

[@andrewandrosow4797]

This type of converter in my opinion can`t working with low voltages - as a result of simulation I had for 200kHz Vout =V, Pot=20W, Vin_min=19, Vin_max=36, Re=3Ohm , Lr=1uH, Cr=470nF, Lm=6uH. Moreover to reduce output voltage a controller must goes in burst mode.Such a big capacitor and a small resistor , I think, will not working good.

[@sambenyaakov]

If you mean low input voltage than this is correct.

[@andrewandrosow4797]

@@sambenyaakov thanks for replay! I made a mistake (I didn't wrote exectly) Vout=from3,7-to7V. I am making an expiriment (because I already have a prototype) in future - I increase Z of the resonant tank (to decrease reactive power and capacitance) and, that is essential, the converter be worked only above serial resonance. In my case it may have sense.If I will not have a success - I re-build this converer to PWM, but with a serial capacitor - to acheive a soft switching of rectifier and reducing harmonics with heavy loads (with low loads it isn't effective)

[@jorgerive7335]

Great video, thank you. I’m wondering why you’re referring to the resonance between the different cases as parallel or series, when in fact it is the resistance that is in parallel or series with the reactive element, and not that the reactive elements are in parallel or series. To me, all the cases with L in series with C are series resonant circuits, and it is only the loaded Q in your various examples that are either series or parallel. Could you please clarify this? Thank you! Love your channel, btw.

[@sambenyaakov]

Hi Jorge, thank you for your comment which could help to highlight another misconception. Parallel resonant converters (the more accurate name is, in fact, parallel LOADED resonant converters, but the “loaded” is usually dropped) are NOT PARALLEL RESONANT NETWORKS. Looking up the vast literature, one will find, in fact, that in BOTH series and parallel resonant converters, the voltage source, inductor and capacitor are in SERIES. The difference is in the loading. In series converters the load is in series with L and C while in parallel resonant converters the loading is in parallel to (usually) the capacitor. This is a convention that is used since the pioneering work of Professor Erickson (and, at the time, his students Vitulski and Johnson (e.g. S. Johnson and R. W. Erickson, "Steady-State Analysis and Design of the Parallel Resonant Converter," IEEE Power Electronics Specialists Conference, 1986 Record, pp. 154-165, June 1986). Generic parallel resonant converters (as in parallel networks) are much less used since they require a current source. I have done some work in this area, some of which can be found by looking for “Ben-Yaakov push pull resonant converter”. Hence, my definitions of Qs and Qp are consistent with what is known in the literature as series resonant converters and parallel resonant converters (see also TI, Survey of Resonant Converter Topologies).

[@biswajit681]

Sir please cover DC link capacitor lifetime calculation ... please please I request you to cover it

[@sambenyaakov]

Do you have any particular caps in mind ? There are new products like the Hybrid of KEMET.

[@biswajit681]

@@sambenyaakov sir conductive polymer hybrid aluminium electrolytic capacitor part no EEHZK1V331V from Panasonic or any capacitor of similar type

[@bensalemn271]

thank you professor

[@sambenyaakov]

Thanks

[@tamaseduard5145]

👍🙏👍

[@sambenyaakov]

👍🙏

[@justinn771]

as always. no further comment needed in his presentation.

[@sambenyaakov]

👍🙏