SMPS feedback controller

[GeoAVR]

Hi all,

I have a question regarding the feedback network of an boost converter.
All power electronics books derive equations for the Vout of converters as a function of duty cycle.
But without a feedback controller these equations never hold true and they almost never describe how to build a feedback controller.

Is there a methodology for feedback design?

If i take the error = Vreference - Vout how to I create a Duty cycle?
Furthermore, in the simplest case of a P controller, how do I deside for the GAIN of the controller.

Thank you

 

[KlausST]

Hi,

please be more specific.

Select a device (or schematic / circuit) that best meets your specifications.

According this one can decide how to "manipulate" the feedback path.


Klaus

 

[--BawA--]

First of all , you have to know , how PWM signals are generated ,
When a DC voltage of a particular amplitude is COMPARED with the sawtooth or triangular wave ,The output of the comparator will be a PWM signal of a fixed duty cycle (you can easily visualize it).
If variation in dutycycle is intended , you have to vary the DC voltage, before comparing with sawtooth or triangular waveform,
However varying the amplitude of DC signal won't affect the frequency of output generated PWM signal.




Regards

 

[BradtheRad]

Feedback is more complicated in a boost converter than in a buck converter.

One issue is that your desired Vref does not exist for the first few cycles. When you finally have your Vref, it cannot be applied directly to your control device, since it is outside the supply rails. It must be divided down.

Vref (voltage regulation) is one type of feedback in a boost converter.

Another issue: Operation (oscillations) has to be continual. Oscillations can be driven by hysteresis feedback. You can get a stable output voltage, though unregulated.

Oscillation (hysteresis) feedback is a second type of feedback in a boost converter.

Another issue is that your boost converter cannot predict how many amperes need to go through the inductor. You want a particular output voltage to a given load. The boost converter does not know the formula to convert amperes to volts.

We know it's a matter of varying the duty cycle. We have a formula in mind. As for the circuit, it needs to apply more than one formula.

 

[GeoAVR]

Feedback is more complicated in a boost converter than in a buck converter.

One issue is that your desired Vref does not exist for the first few cycles. When you finally have your Vref, it cannot be applied directly to your control device, since it is outside the supply rails. It must be divided down.

Vref (voltage regulation) is one type of feedback in a boost converter.

Another issue: Operation (oscillations) has to be continual. Oscillations can be driven by hysteresis feedback. You can get a stable output voltage, though unregulated.

Oscillation (hysteresis) feedback is a second type of feedback in a boost converter.

Another issue is that your boost converter cannot predict how many amperes need to go through the inductor. You want a particular output voltage to a given load. The boost converter does not know the formula to convert amperes to volts.

We know it's a matter of varying the duty cycle. We have a formula in mind. As for the circuit, it needs to apply more than one formula.

Thanks for the answer!

It seems complicated. Do you have in mind a book that describes everything (including control) for Full-bridge boost converters?

 

[--BawA--]

Do you have in mind a book that describes everything (including control) for Full-bridge boost converters?

1) Switching Power Supply Design, 3rd Ed. by Abraham I. Pressman

2) power supply cookbook by Marty Brown

 

[BradtheRad]

Thanks for the answer!

It seems complicated. Do you have in mind a book that describes everything (including control) for Full-bridge boost converters?

Not sure about available books, but there is educational value in a simulation.

This one shows how to use a single op amp (via hysteresis) to keep a boost converter operating.



Output voltage is not self-regulating. If you change any component value, then it is liable to change the output voltage.

More components must be added, to achieve regulation.