Feature: Power supplies


Figure 3: Simplifi ed typical TPPFC application diagram using the ON Semiconductor NCP1680


80+ Titanium standard T e traditional way to actively correct power factor is to use a boost converter from rectifi ed mains to a DC level higher than the mains voltage peak, see Figure 1 leſt . Pulse-width modulation is used to regulate the DC level and simultaneously force the line current to follow the line voltage waveform. T e technique works well and is easy to control in continuous,


discontinuous and critical conduction modes, relating to whether the boost inductor energy is fully exhausted each cycle or not. However, there is also pressure to increase AC-DC converter effi ciency, with the strictest 80+ Titanium standard level for servers mandating up to 96% effi ciency at 230VAC input at 50% load. Typically, 2% loss is allowed for the DC-DC stage, leaving just 2% for the line rectifi cation and PFC stage, but more than 1% is easily lost in the bridge rectifi er alone and up to around 1.7% at low line. A more effi cient technique has therefore been developed, the


bridgeless Totem Pole PFC, or TPPFC, (Figure 1, right), in which the boost diode is replaced by a synchronous rectifi er, enabling the boost transistor and boost diode, Q1 and Q2, to swap functions, depending on mains polarity. Now only two line-rectifi er diodes are necessary, and they can also be synchronous rectifi ers Q3 and Q4 as shown, for even better effi ciency. With perfect switches, an ideal inductor and no diode voltage drops,


the effi ciency of the TPPFC circuit can approach 100%. However, real switches have conduction and switching losses and, although ultra-low on-resistance MOSFETs can be used (even paralleled) to achieve low conduction losses, this invariably increases dynamic losses, which means a balance must be struck. Dynamic losses stem from reverse recovery of the MOSFET,


confi gured as the boost synchronous rectifi er, when its body diode conducts in the switching ‘dead’ time and also from charge and discharge of the switch output capacitance. T e eff ect on effi ciency can be so severe that silicon MOSFETs, even the ‘superjunction’ types, can’t be used in the circuit when operating in continuous-conduction mode. Consequently, wide-bandgap switches in silicon carbide and gallium nitride have to be considered.


Continuous conduction mode (CCM) is favoured at higher power


because peak switch and inductor currents can be set low, reducing rms values and keeping conduction and inductor core losses low. T is is a ‘hard’ switching mode however, with reverse recovery and output capacitance eff ects causing high dynamic losses. At low power, discontinuous conduction mode (DCM) has low


turn-on losses, as at this point boost-diode current has fallen to zero and therefore there is no charge to recover. However, peak and rms currents can be unmanageable, causing high ohmic and core losses, so this mode is unsuitable for high power.


Compromise A good compromise, usable up to a few hundred watts or more with interleaving is to operate in critical conduction mode, or CrM. In this mode, switching frequency is varied to force the circuit to operate on the border between CCM and DCM as load current or line voltage changes. T e benefi ts of low turn-on loss are retained, whilst limiting the peak current to 2x average for reasonable conduction and core loss; see Figure 2. However, turn-off in CrM produces hard switching commutation,


with any forward recovery of the boost diode, causing some loss and output voltage overshoot. T e CrM’s variable switching frequency also has the disadvantage that at, light load, frequency can be very high, producing more switching losses and degraded effi ciency. T is relationship is given by:


T e equation would imply an inverse relationship of switching


frequency to input power; a 20% to 100% load power or 5x change should produce a 5x change in frequency for constant effi ciency. However, higher frequency reduces effi ciency anyway, so the factors interact. T e relationship between frequency and rms line voltage is more complex, producing typically more than 2:1 variation in frequency over the line range, peaking at mid-voltage.


www.electronicsworld.co.uk October 2021 17


Page 1  |  Page 2  |  Page 3  |  Page 4  |  Page 5  |  Page 6  |  Page 7  |  Page 8  |  Page 9  |  Page 10  |  Page 11  |  Page 12  |  Page 13  |  Page 14  |  Page 15  |  Page 16  |  Page 17  |  Page 18  |  Page 19  |  Page 20  |  Page 21  |  Page 22  |  Page 23  |  Page 24  |  Page 25  |  Page 26  |  Page 27  |  Page 28  |  Page 29  |  Page 30  |  Page 31  |  Page 32  |  Page 33  |  Page 34  |  Page 35  |  Page 36  |  Page 37  |  Page 38  |  Page 39  |  Page 40  |  Page 41  |  Page 42  |  Page 43  |  Page 44  |  Page 45  |  Page 46