Do you know what are poles and zeros of the Error amplifier circuitry in this SMPS?

[Flyback]

Hello,
Please could you identify the poles and zeros in this Flyback SMPS Error amplifier set-up, which uses TL431 and opto.? (schematic of SMPS, with TL431/Opto Error amplifier shown)
Basso book on page 292-295 identifies them for an error amplifier circuit that's almost exactly identical as this one , but there is no R4 or C3 in the Basso Error amplifier circuit
(ie R4 is shorted, and C3 not fitted)

 

[MrAl]

Hello,

Do you have the original transfer function from the other circuit available?

 

[Flyback]

Hello,
Its on page 292, (equn 3-66) with the other two pages to show the values and poles.
(the poles and zeros are shown on 292 and 295)

to compare it with my schem...
czero1 = c1
rupper = r12
cpz = c2
rpz = r3
rpullup = 5K (1mA through 5V in side lt1243 controller))
rLED = R2

 

[MrAl]

Hi,

That might be enough but could you show pages 293 and 294 also?

 

[Jony130]

MrAI read this
https://cbasso.pagesperso-orange.fr/Downloads/Papers/The TL431 in loop control.pdf

 

[MrAl]

Hi,

Ok thanks Jony, but it's going to take a little time for me to read that so i'll have to do that later.

In the mean time, at first glance going by the G transfer function given on page 292 of the first link it appears that the new R and C produce a zero at:
f=1/(2*pi*R*C)

and a pole which replaces the 1/(2*pi*Rupper*Czero1) of:
f=(C+Cz)/(2*pi*Cz*C*R)

where
Cz=Czero1,
R is the new series resistor (in series with C), and
C is the new series capacitor (in series with R).

Note that Czero1 (or Cz for short) could be viewed as replacing C3 in the newer schematic rather than C1. This means all we did to the transfer function was replace C3 (which is now viewed as Cz) with a parallel network consisting of both C3 and the series combination of the new R and the new C. The new R is R4 and the new C is C1. This means the new zero is at:
f=1/(2*pi*R4*C1)

and the new pole that replaces the old pole:
f=(C1+Cz)/(2*pi*Cz*C1*R4)

That appears to be correct but i'll have to check again later when i have more time. A few pressing issues today.

 

[Flyback]

Basso pages gratefully supplied as kindly requested

 

[MrAl]

Hi,

Well i ended up getting back here much sooner than expected because what i had to do today is going to take much longer than expected and because it is outside i am going to have to wait for a warmer day, and hopefully that comes soon.

I went over the information i gave you in post #6 and it still appears to be correct so i guess that's all there is to it (except of course for the unmentioned pole at f=0). The thing i dont like about circuits like this though is that they seem overly complicated for example by including the opto internal LED as part of the feedback. But then the opto coupler is often used in regulators that need the output isolated, so if we didnt use that we' d have to find another way to get isolated feedback which could end up being more expensive. Too bad. In line frequency AC converters we used to throw in a simple transformer to get the sampled output AC voltage for the feedback. That of course meant we got isolation at the expense of a small transformer, and since we had a transformer department as part of the whole company that wasnt too costly either.

Can i ask what you are going to use this circuit for, or is it mainly an intellectual exercise?

 

[Flyback]

Thanks for that analysis, yes its for offline SMPS to power class D amplifier, where the resistor R4 is needed to improve transient response and cut out overshoot in vout at start-up.
Class D 's are "pulsey" loads and so the standard Basso type 3 didn't seem to offer the improvement of the added R4, to supply the highly transient load.

 

[MrAl]

Hi again,

Oh ok sounds interesting.

Just so you know, i made a mistake in copying the new pole in post #6 where i typed a multiplication sign "*" in place of where there was supposed to be an addtion sign "+", but i have fixed that now. If you copied any of that information you should therefore update your notes. The mistake had appeared in the second mention of the new pole but as i said it is fixed now.