feedback dc dc power supply

[franticEB]

hi to all,
i'm trying to realize the push-pull in attached images...



in order to overcome the isolation barrier i want to take the feedback for pwm regulation from the output stage of the buck as in figure.
A problem arises from this idea... the output voltage after the inductance is not continuos but it goes to zero with switching of push pull mofset as in figure 2.



How could realize this kind of feedback? Could you help me?

On texas website on ic web page there are datasheets and application notes from which i started my design.

 

[FvM]

Ideally, the average voltage on both sides of the storage inductor will be proportional to the secondary output voltage, obviously with less ripple on the right side of the inductor. In so far you are just asking about suitable filtering.

In a real circuit, there will be a load dependent voltage drop at the transformer (mainly by leakage inductance) and the primary and secondary switching transistors. You have to check if the regulation performance achieved by primary sensing is sufficient.

 

[franticEB]

how i could modify the classical type 3 compensation network in order to achieve this scope?

 

[crutschow]

Your could use a diode peak detector with a capacitor filter to remove the negative spikes. Of course the effect of that capacitance must be included in the feedback analysis.

Since there is now no direct feedback from the output voltage to the control circuit, the loop should require minimal (possibly just 1st order) compensation.