Half Bridge offline SMPS

[mag]

Hi All,
I am starting to develop a 500W offline SMPS.
I would like to use a quasi resonant half bridge topology with ZVS.

This is my first design of on offline SMPS and I have some doubt:

can you please tell me what is the difference between connecting the transformer primary to GND or between the two bus capacitors? (look at the attached drawing)

In any case the transformer will be AC coupled by the small 1uF capacitor an the voltage swing across the primary will always be
+/- 160V (assuming 230V line voltage).

Are these circuits equivalents or not?

thank you

ciao

-marco

[Bootstrapper]

The lower one is correct: With the capacitors you will form a potential that is between the supply rails. When you switch the lower FET on, you will apply this half supply voltage across the primary winding. So the primary current starts increasing. Then you switch the FET off, wait for some time and switch the upper FET on. Now you will apply the same voltage across the primary, but in reversed direction. That means the primary current starts flowing in opposite direction.

This way you generate the alternating voltage to the secondary winding.

The upper picture cannot be correct, since switching the lower FET on wouldn't apply any voltage across the primary.

Hope this helped

[wg_ski]

Actually, they are equivalent - in an ideal circuit simulator kind of way. The lower one is typically used and is better in practice because there is no significant DC voltage on the coupling capacitor. The upper one wll work, but I wouldn't do it at high power.

[darkfenriz]

Agreed,
For the higher circuit the cap will remain charged after switch off, which may (or may not) cause troubles to upper mosfet VGSmax

[mag]

Hi Darkfeniz,
for that I can add a bleed resistor in parallel to the cap to discharge it when it is off

[luka]

Hi Mag

looks good, 2nd one that is...Resistors can be 33k/2w or 47k or any you like of course

What about rest of circuit?

[mag]

Hi Luka,
the rest of the circuit is still to be defined.

Now I am thinking also about a double forward architecture (see attached). The two mosfets will be on at the same time and the two diodes will reset the transformer core (when the mosfets are off).

What do you think about this architecture?

thank you

[Electrone]

A common alternate practice is to use two film caps. One connected to the upper power rail, the other connected to the lower power rail, with the common point between the two connected to the transformer winding. It can help to decouple the power supplies, too.

[megajocke]

Two switch forward is pretty nice. It lets you use current mode control which is good. It doesn't have that many drawbacks compared to half bridge, especially at higher frequencies where you are core loss rather than saturation limited.

Using film capacitors to both rails in a half bridge might work or might create unwanted resonances between electrolytes, lead inductances and the films if you are unlucky.