Half bridge LLC vs Full bridge LLC converter?

Hello,
Supposing that one has a wide output voltage range, then is it better to use a full bridge LLC converter than a half bridge LLC converter? I say this because the Q is lower for a given load when a full bridge LLC is used. -And that surely means that the full bridge LLC can manage a wider output voltage range?

Our spec is...
vin = 390VDC
Vout = 250 - 420VDC
Pout = 3kW

Half-bridge topologies often end up delivering 1/2 the supply V to the next section. This may or may not be true of your LLC converter. It's worth verifying because a large step-up might be required.

i have designed power supply based on half bridge LLC topology (220ac to 24V dc, 100 Wt).
but when i tryed to use LLC topology in another PS with regulated output, i faced problem with regulation. once LLC is not PWM, but freq controled, analysys shows that for quite narrow voltage regulation, very wide freq change need. so before choose topology try to plot voltage gain char-c

Using LLC in this way requires a wide freq control range, e.g. 100kHz - 400kHz, or a high Q resonant tank (as well) in order to get the o/p volts down at no load and high line, and to get full power at low line.

A double stage resonant is perhaps a better approach

Using LLC in this way requires a wide freq control range, e.g. 100kHz - 400kHz, or a high Q resonant tank (as well) in order to get the o/p volts down at no load and high line

Click to expand...

Thanks, too true, we had a consultancy do one for us, and they managed to avoid going to overly high switching frequency by regulating the PFC output voltage down to 340V (this is the input voltage to the LLC converter) .......then we asked them what happens when mains = 265VAC, and its just not possible for the PFC to regulate is output any lower than 370V? They put the phone down on us, and now our lawyers are trying to chace them up.
So much for getting stuff designed by consultancies.

Always ask consultants for reference customers, we are happy to supply these kind of referrals to new customers, as we do defence work and for large OEM's (Eaton etc) it seems to keep us in work...

Thanks for the advice, Eaton I know well, being a former employee in the fuel pumps division.

To answer your original question full bridge LLC is preferred at 3kW, 390VDC, as the fets only have to carry 8A or so, so there is a wide choice, and cheaper than higher current parts....

Usually the Tx design is a bit easier with the thinner pri wires too...

Now that one can buy 1200V 80milli-ohm SiC mosfets the conduction losses can be got quite low, and with the low sw losses in the LLC the efficiency can be very good, still need low loss chokes and Tx and AC caps though....

treez said:

Thanks, too true, we had a consultancy do one for us, and they managed to avoid going to overly high switching frequency by regulating the PFC output voltage down to 340V (this is the input voltage to the LLC converter) .......then we asked them what happens when mains = 265VAC, and its just not possible for the PFC to regulate is output any lower than 370V? They put the phone down on us, and now our lawyers are trying to chase them up.
So much for getting stuff designed by consultancies.

Click to expand...

If your Design Spec didn't state the input limit tolerances, then its the fault of the person wrote the specs.

We've used this trick in reverse, when we needed to go to 155V on our 110VDC rectifiers (1.5kW) we lifted the boost bus from 385VDC to 430VDC as the o/p volts required went from 138V max to 155V max... worked well.

Biggest advantage of full bridge over half bridge I can see is that maybe your resonant capacitor won't need nearly such a high DC voltage rating.

The resonant cap in an LLC has high freq, high current and as a result if often rated at 630V/1kV or higher to be able to handle the current without overheating.

The difference between full bridge (FB) and half bridge (HB) can be seen if the whole converter is modelled as a chain of source (FB or HB), filter network (resonant tank and magnetising inductance) and rectifier. The selection of FB or HB changes the source voltage swing from 0.5xVi (HB) to 1xVi (FB). I wouldn't expect any change in quality factor Q since the two source variants (FB/HB) require scaling of the filter network - resonant tank and transformer turns ratio - such that the shape of the voltage transfer curve stays the same. Further mathematical proof would be interesting though...

As mentioned before in another answer FB has a benefit of lower current since the full input voltage swing is applied. In terms of power dissipation, the benefit for the switches might be minimal since FB requires two switches in series. There is however a thermal advantage since the power is now dissipated in more switches. Mind that the number of drivers required also doubles. The real benefit should come from the resonant tank. The primary side resonant current halves. This in combination with the scaled resonant tank and transfer ratio might become an advantage.

Another advantage of the FB could be the option to make a combination between frequency control and phase shifting control between the two legs. Maybe (I have no real arguments) this helps in relation to the control of the large output voltage range.