Phase shift smps topology

[darkfenriz]

Hi

I was wondering if I could build a phase-shift smps like this:

[areza]

why not try UCC3895 ??!!

[darkfenriz]

When I want to understand and learn, I try to build things from the most basic components. Using an specialized IC (hard to get in my country btw) and use modified circuit from the datasheet I won't learn much.
Have you analysed my schematic?
I tried to generate phase shift operation by T-flipflops triggered from rising/falling edges of normal PWM stage. Is it the right way?

Regards
Adam

[areza]

the schematics you shown is not a phase shift operation , cos one flip flop dont know the state of other flip flop,

check out UCC3895 data sheet timing and block diagram (page 7,8) , you may get the idea,

[Electrone]

It's simple to do with a 74C14 CMOS Schmitt trigger inverter chip. The 330pF capacitor isn't needed. The phase shift resistor can be replaced with a small signal isolation transformer for feedback. Or you could use a bidirectional optocoupler. Another option is to place a standard optocoupler inside a bridge rectifier composed of 1N4148 diodes so that it experiences DC even though the phase shift circuit is AC.

[darkfenriz]

Originally posted by areza
the schematics you shown is not a phase shift operation , cos one flip flop dont know the state of other flip flop,

The flipflops are 'set' at start-up at the same time and one flop toggles at the risng edge while the second toggles at the falling edge of PWM signal .
So there is a phase shift between two half-bridges equal in time with pulse width from the compataror.
How should it be different from what ucc3895 generates?


Electrone

I don't understand, the circuit from the attachmet seems to generate constant phase shift. Could you explain please?


Regards,
Adam

[Electrone]

Originally posted by darkfenriz



The flipflops are 'set' at start-up at the same time and one flop toggles at the risng edge while the second toggles at the falling edge of PWM signal .
So there is a phase shift between two half-bridges equal in time with pulse width from the compataror.
How should it be different from what ucc3895 generates?


Electrone

I don't understand, the circuit from the attachmet seems to generate constant phase shift. Could you explain please?


Regards,
Adam

Hello Adam

To use feedback from the output to vary the phase shift, replace the phase shift resistor with some type of AC feedback device like an optocoupler with FET output that works well with AC.

The phase shift resistor is the unmarked one between the two Schmitt trigger components.

Also, in my real circuits, I find best results if the phase shift capacitor has a value ranging from 1nF to 10nF. I usually use either 1nF or 10nF. 220pF shown in the diagram may prove too small. The value of that capacitor needs to be much greater than the parasitic capacitance that the output side of the feedback device has.

[areza]

a very nice phase shift indeed

below a primitive simulation

[darkfenriz]

Words work better than smiles
What is the problem?
difference between Q1 and Q2 after rectification is exactly CLK input signal.

[areza]

no problem untill u try it with a proto,
thunder, light, magic and smoke
serious now, let me know the result, i m really interested,

good luck