Current-sensing scheme improves PFC on/off sequences

PFC (power-factor-correction) preconverters typically use the step-up, or boost, configuration, because this type of converter is relatively easy to implement (Figure 1). However, this topology requires the output voltage to be higher than the input voltage. When this condition is not the case—for example, with on/off sequences or under load conditions—some inrush current flows through the boost inductor and diode to abruptly charge the output capacitor. For instance, before start-up, the output capacitor discharges. When you plug in the PFC stage, the output capacitor attempts to charge resonantly to twice V_IN. During this sequence, the current can largely exceed the levels obtained during normal operation. Too often, these uncontrolled inrush currents make PFC-stage designers nervous during on/off reliability tests. Except for On Semiconductor's (www.onsemi.com) MC33260, which monitors the entire loop current, available controllers cannot detect this overcurrent state. These controllers may turn on the power switch while a huge and potentially destructive current flows through the inductor.

In the boost structure of Figure 1, the controller turns the power switch, Q₁, on and off to control the L₁ inductor current. Figure 2 illustrates two phases:

• Switch Q₁ is on. The boost structure's input voltage (V_IN, the rectified ac-line voltage) appears across the inductor, L₁, which charges linearly.
• Switch Q₁ is off. The diode, D₁, turns on and drives the inductor current toward the output capacitor, C₁. The inductor current ramps down with a slope equal to (V_OUT–V_IN)/L₁. V_OUT must be higher than V_IN to properly discharge the inductor.

If V_IN exceeds V_OUT, an inrush current flows through L₁ and D₁ and charges output capacitor C₁. Designers generally place a diode, D₂, between V_IN and V_OUT. This diode conducts a major part of the inrush current, thus improving the safety of the first power-switch turn-on. However, when the output voltage is in the neighborhood of V_IN, the current that ramps up during this first switch-on time generally cannot significantly decay during the off-time. As a consequence, the following turn-on operation may occur while the inductor is still charged. Moreover, if V_OUT needs several switching cycles to significantly charge (for example, under heavy load conditions), the power MOSFET faces a succession of stressful turn-ons that may jeopardize the circuit's reliability.

When you cannot use the MC33260, you can use the circuit in Figure 3 to improve the reliability of the PFC stage. You can test this configuration using the MC33262. Typically, the current-sense resistor, R₁ connects between the power MOSFET's source and ground, and the negative terminal of the output capacitor, C₁, connects to ground. As a result, R₁ senses only the power-switch current. In the modified circuit of Figure 3, the output capacitor's charging current also passes through R₁. As a result, the sense resistor senses the entire inductor current. The MC33262 keeps the power switch off as long as the sensed current is higher than the setpoint that an internal multiplier establishes. When the sensed current is below this setpoint, the MOSFET turns on upon core-reset detection. If detection is impossible, a 600-µsec watchdog timer reactivates the power switch. In the case of on/off sequences, the core-reset information is generally unavailable, and the MOSFET turns on in the following cases:

• 600 µsec after the preconverter switches on, regardless of the inductor current, in the traditional application, or
• once the coil current measures lower than setpoint in the modified application.

Figure 4 clearly shows that no switching takes place when the input current is high, as long as the current is higher than the setpoint. In Figure 3's example, the situation is even better. In effect, the 600-µsec timer delays the power switch's turn-on even after the current-sensing block allows the turn-on, so the MOSFET finally switches on when the inductor current is zero. You can see that the modified application schematic increases the power dissipated in the current-sense resistor. In effect, the losses are P_MODIF=1/6R_SENSE(I_PKMAX)², where R_SENSE is the current-sense resistor (R₁ in Figure 3), and I_PKMAX is the maximum inductor peak current (obtained at the top of the sinusoid). In this application, you can compute the losses using the following equation:

 EQUATION 1

where V_AC is the rms input voltage and V_OUT is the output voltage. The relative increase in dissipation then conforms to the following equation:

EQUATION 2

From Equation 2, you can determine that if V_AC=90V and V_OUT=400V, the dissipation increases by 36%. If V_AC=180V and V_OUT=400V, the dissipation increases by 117%. The simple modification of the sensing resistor's location significantly increases the robustness of the PFC preconverter during on/off tests at the price of a reasonable increase of the power dissipation in the sensing resistor.

Is this the best Design Idea in this issue? Select at www.ednmag.com.