SMPS feedback loop for flyback..deriving from buckboost equations?

[treez]

Hello,
I have the small signal feedback loop transfer function for a CCM, current mode, buckboost. (Basso book page 232,fig 2A-22)
I need to adjust it so that it can be used for a flyback...

-So does vout become vout*[(Np/Ns)^2].?
-does Cout become Cout*[(Ns/Np)^2].?
-does Rout become Rout/[(Ns/Np)^2].?
-does Cout's ESR become ESR/[(Ns/Np)^2]....?

 

[BradtheRad]

The usual recommendation with a flyback is to design for 50% duty cycle.

Your equations appear correct as guidelines...
although I believe capacitor ESR itself changes more with frequency rather than V or A going through it.
Whatever the ESR, its effect will be that if higher amperes go back and forth through it, then it will cause greater heat generation.

 

[treez]

thanks, though does output cap ESR need adjusting by the turns ratio for the purpose of assessing the power stage transfer function.
Incidentally, does anyone know why Basso's book doesn't give the control to output transfer function for a sepic?
In fact, nobody else does either, is it too hard to assess?

 

[BradtheRad]

thanks, though does output cap ESR need adjusting by the turns ratio for the purpose of assessing the power stage transfer function.

ESR eats a certain amount power, and increases ripple. If you see it having an effect (causing output voltage to drop), you have room to play with duty cycle and frequency, to bring it back up. Fifty percent duty cycle is not a rigid spec.

In this sense the flyback is like a boost converter. The lighter the load the higher the output voltage soars. The secondary winding needs to generate current as the flux field collapses, and it will produce whatever volt level is necessary.

Therefore turns ratio is not so much the crucial spec. Your control network can adjust frequency and duty cycle (within limits) so as to produce desired voltage at the load.

 

[treez]

Yes, but I am speaking here about the transfer function for bode plots of a flyback.

-as such I am wondering, supposing we transfer the cout and rload to the primary, then do we also have to transfer the esr to the primary, and also, does vout get multiplied by Np/Ns

.....when a trafo is involved in any smps, values have to be transferred to one side of the trafo, then put in the equation
I am not speaking about flyback operation.

The pages of the Basso book that give the power stage transfer function of a current mode ccm flyback (buckboost derivative) are attached. I attach them as I cannot answer this basic question from these pages. Please may you assist?

 

[BradtheRad]

Yes, but I am speaking here about the transfer function for bode plots of a flyback.

-as such I am wondering, supposing we transfer the cout and rload to the primary, then do we also have to transfer the esr to the primary, and also, does vout get multiplied by Np/Ns

.....when a trafo is involved in any smps, values have to be transferred to one side of the trafo, then put in the equation
I am not speaking about flyback operation.

The pages of the Basso book that give the power stage transfer function of a current mode ccm flyback (buckboost derivative) are attached. I attach them as I cannot answer this basic question from these pages. Please may you assist?

Action in the primary and secondary take place at mutually exclusive times. You could say the transformer is an inductor which (a) behaves like the primary during switch-On, then (b) it behaves like the secondary during switch-Off.

So if other parameters are unchanged, then Np/Ns applies. As you referred to Basso's statement:

"For a flyback topology, we can use the buck-boost equations."

Looking at his equations, I don't doubt they are correct, and I wish I could follow them long enough so that someday I could get an idea how they interrelate.

I appreciated his favorable mention of simulators: "Thanks to the SPICE implementation..." (page 225).

does output cap ESR need adjusting by the turns ratio for the purpose of assessing the power stage transfer function.

Even if ESR does not change per se, ESR has the effect of altering voltage in proportion to the current going through the capacitor.

The greater the current going into the capacitor, the more voltage is raised (temporarily) by ESR.

Then as the capacitor discharges, voltage drops a bit during that time due to ESR.

The effect of ESR is to hinder a capacitor's ability to reduce ripple. Therefore it could help if we use a bank of capacitors.

 

[treez]

yes but according to Basso on p235, for flyback , we can transfer things to primary and forget about Ls.

What I am asking is how the parameters of equation 2A-22 are adjusted due to the transformer being present...Page 235 does not tell how to adjust all parameters....it only tells how to adjust cout and Rload. I am asking about the rest of them

 

[BradtheRad]

Yes, the behavior is really too complex to be summed up in one or two pages of equations.

A transformer has many non-ideal characteristics to factor in. These are hard to nail down. (Whereas a single inductor is simpler to evaluate.)

Perhaps that is why Basso did not round out the entire list, but mentioned a few.

The obvious thing with a transformer is 10 or 15 percent losses.

Also the magnetic characteristics when the primary carries a certain amount of amperes, as compared to the secondary carrying a much greater current.

I wish I had more expertise in this area.

 

[treez]

yes though that's not to do with this feedback loop analysis, here its making out the equation, and even though its slightly ideal, at least its a basis