Is there any reason why the LLC Resonant SMPS is not often written about as an amplifier suppkly SMPS in this forum?
I hazard a guess that Audio SMPS's don't often use PFC stages.....therfore, with a 90-265VAC input, the input voltage to the SMPS is very variable, and this does not suit an LLC Resoant SMPS.
Is this true?
I hazard a guess that Audio SMPS's don't often use PFC stages.....therfore, with a 90-265VAC input, the input voltage to the SMPS is very variable, and this does not suit an LLC Resoant SMPS.
Is this true?
I have the impression it's more that interest in resonant conversion only picked up recently. There's relatively little available in the way of design collateral or parts compared to buck/boost or flyback topologies. From the standpoint of keeping switching noise out of amp innards the continuous current in a resonant converter and the ability to have roughly equal and opposite transients on Vcc and Vee is rather interesting.
I took a look a while back and found I couldn't get a BOM together that was price competitive with Connextronic's supplies. Performance is an open question, but mostly a DIY build looks like something one would do for fun. In particular, I'd like to do a relatively low VA LLC series resonant converter that output a 5 to 15V dual supply. Haven't found a good supplier for the magnetics though. The few options I've come across so far are in excess of $1/VA; rather steep when you've got a small herd of unused toroids for linear supplies sitting around.
I took a look a while back and found I couldn't get a BOM together that was price competitive with Connextronic's supplies. Performance is an open question, but mostly a DIY build looks like something one would do for fun. In particular, I'd like to do a relatively low VA LLC series resonant converter that output a 5 to 15V dual supply. Haven't found a good supplier for the magnetics though. The few options I've come across so far are in excess of $1/VA; rather steep when you've got a small herd of unused toroids for linear supplies sitting around.
The "Frequency and Gain" section of this article by Jon Harper of Fairchildsemi explains why LLC resonant converters are terrible for universal mains.............
Using quasi-resonant and resonant converters
...they lack dynamic gain near the resonant frequency.
HiperLCS by powerint.com looks a very cost effective resoant converter solution, but as you know, really with resonant smps, you need a PFC stage up front.
Using quasi-resonant and resonant converters
...they lack dynamic gain near the resonant frequency.
HiperLCS by powerint.com looks a very cost effective resoant converter solution, but as you know, really with resonant smps, you need a PFC stage up front.
Beats me, we use LLC "converters" all the time for induction heating. The load can change wildly; when you put 10kW on a hunk of steel, it goes from magnetic (heavy load) to nonmagnetic (light load) rather quickly. Also, the resonant frequency changes considerably. The challenge is to keep it under control, without knowing a priori what the resonant frequency or load impedance is. But we're doing a pretty good job of it. I'd gladly welcome a simple, consistent power converter that has a fixed operating frequency (give or take) with low Q and variable power consumption!
Tim
Tim
Well, i've just found LLC components for an LLC resonant converter that can manage a vin variation of 193V.
They are :-
Lr = 40uH
Lm = 1mH
C = 100nF
Np/Ns = 0.78
Switching frequency at 330W load and 180VDC input is 97KHz
Switching frequency at 330W load and 373VDC input is 143KHz
Vout is 80V (its actually +/-40V, 4A)
The actual resonant frequency is 80KHz, but i could not operate near that, as i would be in danger of going into the capacitive region.
I am getting ZVS for FET at switch on but not at switch off.
The turn-off fet current is 6A83 so i am getting significant switching loss at turn off.
It loks to me that a nice normal LLC resonant converter cannot be done for our large vin range, and that we are having to go deeply into the inductive region to be able to manage it.
Diodes however are ZCS at switch on, but hard-switched off.
The Normal LLC resonant converter has ZVS for fet tunr on and off, and ZCS for diode turn on and off.
As you know a "normal" LLC resonant converter has a much higher Lr value than ours, that is, higher in relation to the magnetising inductor.
They are :-
Lr = 40uH
Lm = 1mH
C = 100nF
Np/Ns = 0.78
Switching frequency at 330W load and 180VDC input is 97KHz
Switching frequency at 330W load and 373VDC input is 143KHz
Vout is 80V (its actually +/-40V, 4A)
The actual resonant frequency is 80KHz, but i could not operate near that, as i would be in danger of going into the capacitive region.
I am getting ZVS for FET at switch on but not at switch off.
The turn-off fet current is 6A83 so i am getting significant switching loss at turn off.
It loks to me that a nice normal LLC resonant converter cannot be done for our large vin range, and that we are having to go deeply into the inductive region to be able to manage it.
Diodes however are ZCS at switch on, but hard-switched off.
The Normal LLC resonant converter has ZVS for fet tunr on and off, and ZCS for diode turn on and off.
As you know a "normal" LLC resonant converter has a much higher Lr value than ours, that is, higher in relation to the magnetising inductor.
I've just found an article about LLC resonant SMPS that doesnt make sense
http://www.eetimes.com/design/power.../Using-quasi-resonant-and-resonant-converters
...in the section "frequency and gain" it says why universal mains use is not possible for LLC reonant converters, and references Figure 3.
-Surely this makes no sense at all.......isnt it obvious that all you have to so is make low mains (USA) correspond to the 100KHz point, and then higher mains (Europe) will be served by increasing the frequency.?
http://www.eetimes.com/design/power.../Using-quasi-resonant-and-resonant-converters
...in the section "frequency and gain" it says why universal mains use is not possible for LLC reonant converters, and references Figure 3.
-Surely this makes no sense at all.......isnt it obvious that all you have to so is make low mains (USA) correspond to the 100KHz point, and then higher mains (Europe) will be served by increasing the frequency.?
Last edited:
I designed an LLC SMPS using a SMD IR device and it worked very well.
It was 45/0/-45 at 5 amps per rail.
It was happy with varying loads although I didn't run it off different mains voltages.
I did have a minimum load attached to the secondary.
I am currently working on a PIC based LLC SMPS.
The PIC simply outputs 80KHz (full power at resonance) or 160KHz (off resonance) depending on feedback logic level.
It was 45/0/-45 at 5 amps per rail.
It was happy with varying loads although I didn't run it off different mains voltages.
I did have a minimum load attached to the secondary.
I am currently working on a PIC based LLC SMPS.
The PIC simply outputs 80KHz (full power at resonance) or 160KHz (off resonance) depending on feedback logic level.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.