sg3525 feedback design

hi
i use this schematic to build a power supply with some changes(core is ETD39-frequency is 80KHz-target output voltage is 45*2-output capacitor is 2000uf-output inductor is 28uh)
but i have a problem with feedback and by this feedback transformer start to produce a noisy sound like"tik-tik-tik" and voltage change beetwen 40-45
this feedback is a crude feedback an dosn't control duty cycle continuously and just turn pwm on/off
so i need a better feedback and by searching net i realized that i should use TL431 for feedback.
now i want u help me design a good feedback using TL431 and 4N25.

Hi,

Your description makes me assume that the inductor is saturating.
Did you check this.

What about air gap on core?
Did you wind it on your own?

Klaus

KlausST said:

Hi,

Your description makes me assume that the inductor is saturating.
Did you check this.

What about air gap on core?
Did you wind it on your own?

Klaus

Click to expand...

for output inductor i use two EE25 core with about 0.5 mm gap and wind them on my own.inductor dosn't saturate
without feedback smps work without any noisy sound and duty cycle is maximum and voltage is 155/turn ratio
i connect a 300 watt load without feedback and circuit worked fine but voltage droped about 20 volt which is normal without feedback
i just want a good feedback with TL431

Hi,

If you haven't already seen these 431 in control loop application notes, I hope they can be of help.

View attachment Using TPS23753A with TL431 as EA slva433.pdf
View attachment The TL431 in the Control of Switching Power Supplies OnSemi TND381-D.PDF
View attachment The TL431 in loop control.pdf
View attachment Designing with the Advanced TL431, ATL431 slva685 TI.pdf
View attachment Closing the Loop with TL431 Kollman and Betten September 2003.pdf
View attachment 15 Designing with the TL431.pdf

Regards.

yoosefheidari said:

transformer start to produce a noisy sound like"tik-tik-tik" and voltage change beetwen 40-45

Click to expand...

The irregular operation could be unwanted 'hiccup' mode, or maybe a case of squegging. It works okay for a few cycles, then a momentary lockup. As this repeats back-and-forth, the transformer can make audible noises.

The noise is not necessarily arcing but it is natural to wonder if you're getting high voltage arcing in the transformer? Your schematic is not obvious whether it has a snubbing network. Its caption says 700W so the snubber should be able to handle 700W spikes.

I suspect that unsuitable feedback loop compensation is causing the observed instability. By connection a capacitor to comp pin you get pure I controller operation which is likely to give insufficient phase margin. Although TL431 circuit has additional features to tune the loop characteristic you'll probably need to modify the SG3525 compensation.

FvM said:

I suspect that unsuitable feedback loop compensation is causing the observed instability. By connection a capacitor to comp pin you get pure I controller operation which is likely to give insufficient phase margin. Although TL431 circuit has additional features to tune the loop characteristic you'll probably need to modify the SG3525 compensation.

Click to expand...

i think this problem caused by on/off state of pwm.in fact with this feedback transformer turns on and off continuously and cause that noise.
so i need a feedback that control output by decrease/increase duty cycle.
am i right?
i cant design a feedback on my own because i havn't compelete information about poles/zeros/bode plot/transien response/phase margin/gain margin and......
so i need much help.on the net i found this feedback
is that good without any problem?can i use this feedback without any change?or nees some modification?this is type I or II ? is compensation good?