Converter Control

Why Converter Control is needed?

The output voltage of a converter changes due to disturbances in:

1. Circuit element values
2. Input Voltage
3. Load Resistance

So, we need to continuously control our switching elements so that our output voltage remains within an acceptable range.

How to control the switching element?

Switching elements operate in a closed feedback loop. The output of the converter is fed to a sum block. The output from the sum block is the difference between the output voltage and a reference voltage i.e. our desired output voltage. The output from the sum block (error between the reference and converter output voltage) is fed into the compensator.

The compensator is basically a lag-lead compensation circuit. In short, it shifts both the poles and zeroes of the input so that our error from the sum block will be stable, fast, and does not oscillates.

After the compensator, the compensated signal goes to the pulse width modulator.

The pulse width modulator consists of a comparator circuit which compares our compensated signal to a sawtooth wave generator. So, whenever our compensated signal is higher than the sawtooth wave generator, the comparator output is 1 and when the voltage is lower than the sawtooth wave generator, it gives 0. In this pulse are produced.

These pulses are then fed to the gate of MOSFET through a gate driver circuit. The function of the gate driver is to increase the voltage magnitude so as to improve the response time of MOSFET.

duty cycle of converter = Compensated voltage / Peak voltage of Sawtooth wave generator

Average AC Modelling:

Earlier, we used to model inductor voltage and capacitor current by their respective dc value. But to get more approximate responses, we will model is based on the quiescent model that we used in the analysis of transistors.

In this response, signals are approximated by their operating DC values and a small AC value around the operating value just like in transistors. So, inductor voltage and capacitor current will be approximated to low frequency averaged values.

Linearization- In AC modeling, we will always come across terms where we have to multiply two sinusoid waves. So, the average values will be non-linear. Hence, to apply the Laplace transform, we have to linearize it. Linearization can only be done by the Taylor series. In the Taylor series, we can neglect higher-order terms as its value will be very small.

#Converter Control