Adjustable voltage-mode smps

[zkaiser]

I am in the middle of designing to make smps front pre-regulator stage for bench top supply with linear CC/CV reg. And I would like opinions is this pcb layout prone to make some heavy ringing issues ? Operating PWM frequency would be 33kHz because I need to be able to produce short impulses in order to achieve regulation from near zero to 32V at the output...

Topology is Half-bridge for 230V mains, output needs to be nearly const. ~80W until voltage drops below 13.3V when current is limited to 6A, pwm controller is sg3525 and isolated low voltage side is monitored with lm358 which one half is actually for thermal protection...

[areza]

may be bipolar switcher with proportional driver is better with veriable duty cycle when using a driver transformer

[zkaiser]

may be bipolar switcher with proportional driver is better with veriable duty cycle when using a driver transformer

Somehow I think that for fast and wide range changing duty cycle, MOSFET are better solution especially for such a low power, however here is the current schematic and I still have problem with output voltages below 5V depending on load.

[zkaiser]

Is there any conventional solution for squealing noise when duty is near zero and controller is entering chaotic pulse skipping mode, because even though a have some bleeding load 270ohm resistor + const. current ~200mA, output voltages below 8V producing undesired effect. If I just increase idle load this is direct throwing of efficiency, certainly not good thing to do ?

[Sch3mat1c]

Put some acoustic padding around, pulse skipping is more efficient than asking the FETs to swing 150V+ for hardly any output power. In other words, by trying to make it continuous, you incur much higher switching losses.

Tim

[zkaiser]

Here are waveforms at some points...

Proportions respectively are (5V/5us, 20V/5us,20V/0.1us , 20V/5us).

First picture is gate-source voltage at one of irf740 with max duty, which is around 0.95, second is voltage just before output inductor, third is magnified spike from fourth waveform loaded with 3.6A (Vout=8.6V passed through halogen 60W car light bulb).
Second waveform is only bleeding load at Vout=8.6V, when pwm is still continuous and regular, below this and I need to load more or squealing sound is here.

[armandjones82]

Originally Posted by zkaiser

Here are waveforms at some points...

Proportions respectively are (5V/5us, 20V/5us,20V/0.1us , 20V/5us).

First picture is gate-source voltage at one of irf740 with max duty, which is around 0.95, second is voltage just before output inductor, third is magnified spike from fourth waveform loaded with 3.6A (Vout=8.6V passed through halogen 60W car light bulb).
Second waveform is only bleeding load at Vout=8.6V, when pwm is still continuous and regular, below this and I need to load more or squealing sound is here.

That was a great idea on what you have raised there. I think this would really help us to really get things straight and good for your project.