Designing LLC Converter with variable output voltage and current

[tictac]

Hi all,

I would like to get your advice for designing the power architecture(s) for the following power supply:
Input: 304-456 Vac ( 3 phase)
Output: 5v to 50V - 10A to 300A

I has studied this App note from TI for calculating parameters.
**broken link removed**

I want to know Can I use LLC for variable output voltage and current? At the above app note describe about the ability of LLC for line and load regulation

"The LLC resonant converter has many additional benefits over conventional resonant convert
ers. For example, it can regulate the output over wide line and load variations with a relatively
small variation of switching frequency, while maintaining excellent efficiency. It can also achieve zerovoltage
switching (ZVS) over the entire operating range. Using the LLC resonant configuration in
an isolated half-bridge topology will be described next, followed by the procedure for designing this topology."

I use calculations that describe at the app note but It calculates for fixed power with fixed output voltage and current . I want to change this calculations for my requirements(variable output voltage and current)

At first it calculates n(transformer ratio) and then Mg-min and Mg-max for fixed output voltage.(Mg=Gain)
Mg-min= n*(Vo-min+Vf)/(Vin-max/2)
Mg-max=n*(Vo-max+Vf+Vloss)/(Vin-min/2)
There is an example at this App note that assumes output voltage like this. ==> "Vo-min=Vo*(1-1%) and Vo-max=Vo*(1+1%)"
I want to know how can I change this calculations for my defined smps. Can I change Vo-min=Vo*(1-90%) ??

Please help me for starting this.
Regards.

 

[mtwieg]

The key word in that paragraph is "relatively" compared to other resonant converters. Yes the LLC is better than SRC and PRC converters, but it is still nowhere near as flexible as non-resonant topologies. I'm no expert, but I'm pretty sure your wide operation range is not feasible with the LLC. You can plug in your specs to the design equations and see that it will require an extremely large frequency range, even with a very high Q.

 

[tictac]

Thanks you for response.
yes, you are right.
If you see the figure 6d at the attached document about LLC design , I can think if I select Ln=5 and Qe=0.5 , so I can get the maximum power with the switching frequency near to the resonance frequency and with increasing it I can decrease the output voltage to the minimum. what is your idea about it?

So , What can I do for this design? If I use pure full-bridge topology for it ,then the losses because of hard-switching is very high. I should use ZVS for it. How can I run this technique in this design ? Do you know better solution for this range of power ?

Regards.

 

[mtwieg]

Thanks you for response.
yes, you are right.
If you see the figure 6d at the attached document about LLC design , I can think if I select Ln=5 and Qe=0.5 , so I can get the maximum power with the switching frequency near to the resonance frequency and with increasing it I can decrease the output voltage to the minimum. what is your idea about it?

From your specs, your Ma needs to span a 15:1 ratio, and Qe must span a 30:1 ratio. Even if you were to manage a maximum Qe of 10, then you would have to operate at Qe=0.33 at light load, and from looking at the plots the intercept of Ma=1/15 with Q=0.3 doesn't even show up. It looks like your frequency span would be on the order of 20 or so, even with Ln=20. I can't say it's impossible, but it's well beyond the typical LLC application.

So , What can I do for this design? If I use pure full-bridge topology for it ,then the losses because of hard-switching is very high. I should use ZVS for it. How can I run this technique in this design ? Do you know better solution for this range of power ?

A phase shifted bridge can be quite efficient if designed correctly. Advanced designs adaptively change the dead time to achieve optimal ZVS without diode clamping.

Also I presume you're going to use an active PFC front end, so you're stuck with hard switching anyways there...

 

[tictac]

From your specs, your Ma needs to span a 15:1 ratio

what the Ma point to? can you tell me about 15:1 ? how you get it?

and Qe must span a 30:1 ratio

how you get it?

A phase shifted bridge can be quite efficient if designed correctly

As I know s phase shifted bridge only can operate for fixed output power, Isnt it?

Thanks you so much for your answers.

 

[mtwieg]

what the Ma point to? can you tell me about 15:1 ? how you get it?

Whoops meant to type Mg. You Mg is the DC gain of the converter, so if will span from Mg_min=Vout_min/Vin_max to Mg_max=Vout_max/Vin_min. So if we normalize Mg to Mg_max=1, that means you need to operate down to Mg_min=1/15.

how you get it?

Qe is inversely proportional to load resistance, and your load current must vary by a factor of 30, as per your specification. Actually it might be as high as 300, when also accounting for the factor of 10 variation of output voltage.

As I know s phase shifted bridge only can operate for fixed output power, Isnt it?

No, it can be controlled over a wide range, though efficiency may not be very high over a very wide range, unless exceptional control techniques like adaptive dead time are used.

 

[tictac]

Hi
I want to design my lcc circuit with above specification with L6599. I want to know can I use this controller for lcc resonant topology?
I use the below link for testing different values of resonance parameters.
**broken link removed**

Please let me know about it.

 

[Orson Cart]

Hello there, your input range and output range are not practically achievable with a standard LLC converter.
Better to look at driving a series resonant converter with phase shift PWM. We have designed and built these up to 14kW.
Regards, Orson, pwrtrnx. c o m

 

[tictac]

Hi
Can you describe about the reasons of unachievable of building this converter with LLC?

I find many article for designing HV smps with LCC not LLC. you can search "LCC high voltage " in google.
Can you help me for building it,please?
Regards

 

[tictac]

Sorry, I forgot the main topic of this thread.
Now I want to design another smps with below specification

Vin=220 Vac
Vout=1KV to 30KV , Iout=0 to 20mA

I want to use LLC for it. Is it possible to use LLC for this converter?
Regards.

 

[Orson Cart]

You need to start a new thread...