The PSU FAQ

[Behemot]

up?

[PeteS in CA]

It's probably way late to respond, but there are 4 rectifiers in a bridge rectifier. Two conduct during one half of the AC cycle, and the other 2 during the other half cycle. Thus the filter cap(s) get charged during the full cycle, or the full wave. Hence the name "full wave bridge bridge".

[Behemot]

I meant the secondary rectifiers, there is no full bridge being used, just two diodes with common output. In every cycle, just one diode is used, yes? But I have learnt there is also something called duty cycle, so, in reality every diode is able to rectify all the power of the whle rectifier because after short time the second diode will do the rectifying. The first could "rest" in this time, and otherwise, right?

[PeteS in CA]

Ah, sorry for misunderstanding. The answer depends on the topology, whether it's a forward converter or a half- or full-bridge.

In a bridge topology current flows through the transformer primary winding one way then the other in alternating half-cycles (with "dead time" in between each half cycle of conduction). The secondaries are center-tapped, with that center tap being output return. The ends of the winding each have a rectifier (or one of the rectifiers in a dual rectifier), which conduct in alternating cycles.

In a forward converter current flows through the transformer primary winding in just one direction. So you have an "on" time and an "off". This simplified schematic on Wikipedia shows how the output rectifiers are connected in a forward converter. There is one rectifier (or half of a dual), D1, that conducts during the "on" time, charging the output inductor through the load and back to the other end of the winding. During the "off" time, the transformer and D1 are not active, but the inductor is charged and able to provide current to the load. The second rectifier, D2, is connected to complete that current path. During "off" time, current flows out of the inductor, through the load to the output return, from return into the second rectifier, and through the rectifier into the other end of the inductor. The current through the inductor is "continuous". It is either being charged while current is also being supplied to the load or partly discharging to supply current to the load. Thus, the output inductor averages the rectified voltage pulses to supply a continuous current to the load, resulting in a DC voltage.

[Behemot]

I see, thanks for explaining this :-) In both modes icreasing the switching frequency means increasing forwarded power, right? In first, there are shorter dead times, in the second, power is pumped more often to the inductor.

Anyway, the fact that the single diodes are able to let the same amount of current through as is the maximum of whole device (two diodes in parallel) is conected with the duty cycle, right?

[JonathanAnon]

Would it be handy to try to find an image of some basic switch mode power supply and try to mark out on it all of the areas that are referred to in the original post... It would be easier to visualize the different sections.. I'd draw it up in Photoshop...

[joshnz]

Originally Posted by Behemot

I see, thanks for explaining this :-) In both modes icreasing the switching frequency means increasing forwarded power, right? In first, there are shorter dead times, in the second, power is pumped more often to the inductor.

Anyway, the fact that the single diodes are able to let the same amount of current through as is the maximum of whole device (two diodes in parallel) is conected with the duty cycle, right?

Duty cycle refers to the on time


almost all switching power supply's use this method to control the output most have a maximum duty cycle of 80% due to core saturation that means the transformer core or inductor core has reached the maximum of magnetic flux it can hold.

[Behemot]

Yep, but as there are two diodes in the bridge topology (with double the secondary wiring on transformer and return pin), when one has duty cycle of e. g. 30 %, the other gets the 70 %, right? So in all they have 100 % and can do twice the current per diode (means, total output power equals to power of device).

[joshnz]

Originally Posted by Behemot

when one has duty cycle of e. g. 30 %, the other gets the 70 %, right?

no they both will get 30% given to them 30% on one half of the cycle and 30% on the other. this setup doubles the the number of output pulses giving more stable voltage and higher output current.

If I have my schematic soft on this machine I could explain it better.

[Behemot]

So when one half of the wiring gets high pulse, the other one gets the same low, but with reversed polarity, so basically both get the same (two minus give plus)?

[joshnz]

Originally Posted by Behemot

So when one half of the wiring gets high pulse, the other one gets the same low, but with reversed polarity, so basically both get the same (two minus give plus)?

?
in my attachment you can see how the two halfs are prossesed
excuse my crude drawing done it in paint
also i made a error in the pic d1 and d2 are the same part.

[joshnz]

Originally Posted by joshnz

?
in my attachment you can see how the two halfs are prossesed
excuse my crude drawing done it in paint
also i made a error in the pic d1 and d2 are the same part.

this is another way it is done you double the diodes current capacity but only one output pulse per cycle and the winding on the transformer carry's twice the current if was run at the same frequency.

[Behemot]

So there is also common wire on the primary, I see. It clears a lot of things now

EDIT: and theres also famous push-pull.

[itwentboom]

Not sure if to mention this here, but it is interesting to note most of the computer SMPSUs come from Taiwan/China. I can see the economics of manufacture of the far east but do they do the design and development too?

I'm in the UK and trying to design/manufacture in this country is usually not a good idea, just the usual costs of.

[Behemot]

Yes, majority of worldwide successful companies producing good PSU's are Taiwan based. Just a couple (actually ) is from USA I think. Maybe few more I am not so familiar with too.

[german318]

Thanks for the info