Offline Dual SMPS with split output

Well, it's a pretty crazy design and not very efficient or cost effective.
1. It uses half wave rectification at the input - why?
2. It duplicates all the components.

Suggestions: use a bridge rectifier at the input,
remove all the bottom half of the schematic,
use a different transformer and rectifiers to produce the positive and negative outputs,
add some voltage feedback to regulate the output voltage.

Also note it is difficult to provide the VCC to the ICs, particularly on the lower IC. You can't share VCC lines!

Brian.

Thanks, its two half waves...so kind of "full wave"...also, you only go through one diode at a time ...not two diodes in series like in a full wave bridge, which mean wasted power.
Regulation for each of the output rails will be very tight, unlike a two coil secondary on a single flyback where there will be coupling issues.....and the rails may wander apart.
It does duplicate...but that means for a 75W offline flyback, you now need no heatsinks on the diodes or fets...unlike a single 75W flyback where it'd be adviseable to heatdsink the semi's.

If the output power was say 180W....you could use the 2 flybacks as shown (wit a bit of variation of components etc).......and this would be less custom magnetics than eg a two transistor forward.......2 tran forward, ayk, needs custom transformer, custom output inductor, and custom high side drive transformer....as well as two fets on primary and two fets on secondary.
Ayk, 180w would usually need a PFC, but not always....in any case, the same method as shown could be used with dual PFCs.

You didn't mention the load before but I would still raise the same concerns. Uneven loads on the outputs would also cause asymmetric loading of the AC input and that might cause other problems.
At the very least I would use a bridge rectifier and two identical circuits with one having its output rectifier/cap reversed.

Brian.

Hi,

I would it even more simplify:
Mains --> full bridge rectifier --> capacitor
(Filters and so on as you like)

Then I would power both supplies from this one capacitor.

There is no need for the negative supply to be powered by an extra input path.

And - as Brian mentiones - it always gives best distribution on both mains half waves.

++++
All this is my idea for a new design ... but I guess it´s too much effort for a modification on the existing design.

++++
Btw: There are ready to buy supplies ...

Klaus

Uneven loads on the outputs would also cause asymmetric loading of the AC input and that might cause other problems.

Click to expand...

Thanks, this was also our concern, but we are unable to find what EN standard governs this? if any?

Thanks, also which one, right or left, gives best placing of the filter inductors and NTC's? Common mode noise will be different with these, which one is best?

The current, whether intended or not is the same in a series path so I would expect near identical performance.
Do you need two NTCs at all, I would think one in series with the AC feed should be sufficient and put the filter inductors one in each of the AC feed wires.

Brian.