A novel charge-mode control algorithm for PFC

1 Introduction

In a control system, if you want to control something, you need to sense it; this applies to power factor correction (PFC) applications as well. In offline power supplies with power levels >75W, PFC controls the input current to create a sinusoidal waveform (in other words, following the sinusoidal input AC voltage). In order to control the input current, it needs to be sensed.

The most common current-sensing method places a shunt resistor at the PFC ground return path (designated as R in Figure 1) to sense the input current. The sensed input current signal (I_SENSE) is then sent to an average current-mode controller [1] (shown in Figure 2). Because the current reference (I_REF) is modulated by the input voltage (V_IN), it is a sinusoidal waveform. The control loop forces the input current to follow I_REF, thus achieving a sinusoidal waveform.

Figure 1 A common current-sensing method for PFC.

Figure 2 Traditional average current-mode control for PFC.

Almost all continuous conduction mode (CCM) PFC controllers use traditional average current-mode control. Although traditional average current-mode control achieves a good power factor and has low total harmonic distortion, it also has some limitations, especially in totem-pole bridgeless PFC. This article presents a brand-new control algorithm: charge-mode control [2].

2 Charge-mode control

The charge-mode control algorithm is a new control concept: to control an object, you don’t really need to sense it – you can sense its consequence and then indirectly control the object. For PFC, instead of controlling the input current directly, this control algorithm controls how much electric charge is delivered to the PFC output in each switching cycle, and employs a special control law such that the input current becomes a sinusoidal waveform by controlling the electric charge.

There are a few ways to obtain the electric charge information. Figure 3 shows an example of using a current shunt and an operational amplifier (op amp) circuit, with the op amp configured as an integrator. When the PFC boost switch turns off, the inductor current starts to charge the PFC bulk capacitor. The shunt resistor senses this current, which is then integrated through the integrator. The peak value of the integrator output represents the total electric charge delivered to the PFC output in each switching cycle. This electric charge (V_CHARGE) is sampled by the controller as a control-loop feedback signal. The integrator discharges to zero through Q1 before the boost switch turns off.

Figure 3 Using the current shunt and op amp to obtain an electric charge.

Figure 4 shows another method, which employs a current transformer (CT) on the PFC output side. The CT output connects to capacitor C1. When the PFC boost switch turns off, the inductor current starts to charge the PFC bulk capacitor. The CT senses this current and its output charges C1. The voltage on C1 rises up; its peak voltage represents the total charge delivered to the PFC output. The controller samples the peak voltage V_CHARGE as a control-loop feedback signal. C1 discharges to 0V through Q1 before the boost switch turns off.

Figure 4 Using a CT to obtain an electric charge.

Figure 5 shows the typical signal waveform for charge-mode control.

Figure 5 Typical signal waveforms for charge-mode control.

3 Control law

Now that you know how to obtain the electric charge information for each switching cycle, let’s take a look at how to get the sinusoidal input current waveform using the new control law, see Figure 6.

Compared to the traditional control law shown in Figure 2, there are two differences:

The current-loop reference is modulated by V_IN², not by V_IN.
The feedback signal is the electric charge V_CHARGE, not I_SENSE.

Figure 6 Charge-mode control law for PFC.

From Figure 6, the current reference I_REF is given by:

Equation 1.

$I_{R E F} = \frac{A C}{B}$

where, I_REF is the current-loop reference, A is the voltage-loop output G_V, B is Vrms² used for V_IN feedforward control, and C is V_IN².

Looking at Figure 5, Equation 2 expresses the average inductor current in each switching cycle as:

Equation 2.

$I_{A V G} = \frac{{(I}_{1} + I_{2}) {(T}_{o n} + T_{o f f})}{2 T}$

where, I_AVG is the average inductor current, I₁ is the inductor current at the beginning of each switching cycle, I₂ is the inductor current peak value in each switching cycle, T_on is the boost switch Q turn on time, T_off is the boost diode D conduction time, and T is the switching period.

Equation 3 calculates the peak voltage of C1 (V_CHARGE) in each switching cycle as:

Equation 3.

$V_{C H A R G E} = \frac{(I_{1} + I_{2}) T_{o f f}}{2 C}$

where, C is the capacitance of C1.

In steady state, the control loop forces V_CHARGE to equal I_REF (see Equation 4):

Equation 4.

$V_{C H A R G E} = I_{R E F}$

For a boost-type converter in steady-state operation, the volt-seconds applied to the boost inductor must be balanced in each switching period (see Equation 5):

Equation 5.

$T_{o n} V_{I N} = T_{o f f} (V_{O U T} - V_{I N})$

Equation 6 combines Equation 1 through Equation 5:

Equation 6.

$I_{A V G} = \frac{G_{v} V_{O U T} C}{V_{r m s}^{2} T} V_{I N}$

In Equation 6, since both C and T are constant, and G_V, V_OUT and Vrms² do not change in steady state, I_AVG follows V_IN. When V_IN is a sinusoidal waveform, I_AVG is also a sinusoidal waveform, thus achieving PFC. Note that Equation 2 and Equation 3 are valid for both CCM and discontinuous conduction mode (DCM); therefore, Equation 6 is valid for both CCM and DCM operation.