Rectifier Bridge With No 2Vf drop
The venerable full-wave rectifier bridge (Fig. 1) is a common, familiar circuit for converting an AC input voltage to a DC output voltage. It is also useful for translating a DC input of arbitrary polarity into a DC output of known polarity, as is commonly required in electronic telephones or other telephony devices, and has application in protecting against battery reversal in battery-powered circuits.
A drawback of the classic four-diode rectifier bridge is the unavoidable forward voltage drop (Vf) of two diodes when current is flowing. With conventional silicon diodes, this could typically amount to 1.5 volts or more. The result of this is wasted power and reduced efficiency in power supply applications, or loss of working voltage in telephony or battery-powered applications.
In telephony applications in particular, it is possible for a device to have as little as 4 volts available to it under worst case conditions of loop current and line length. Since most integrated circuits, telephony or otherwise, are decidedly unfriendly about power supply reversals, it is common practice for the line-powered electronics to be surrounded by a full-wave rectifier bridge in order to guarantee power supply polarity. But with only 4 volts of line voltage, a 1.5-volt drop in the rectifier would leave only 2.5 volts for the electronics!
Similarly, in battery powered circuits, it is often the case that the loss of efficiency caused by series diodes to protect against inadvertent battery reversal is unacceptable.
The circuit shown in Fig. 2 eliminates this drawback by replacing the diodes with MOSFETS. The four MOSFETs are connected in such a way as to conduct in opposing pairs. Which pair conducts is a function of the polarity of the applied voltage. The conducting pair is such as to steer the applied voltage to the appropriate output terminals so as to always maintain the same polarity at the output. In other words, the circuit rectifies.
Interestingly, if one looks at the intrinsic drain-to-source body diodes of the MOSFETs, ignoring the MOSFETs themselves, they form the conventional rectifier bridge configuration. Indeed, when voltage is first applied, the circuit acts the same as a conventional rectifier bridge in that the forward voltage drop of two diodes (2Vf) appears between the input and the output. But as soon as the applied voltage exceeds the turn-on threshold of two MOSFETs (or more precisely, the sum of an N-channel threshold and a P-channel threshold), the appropriate pair of MOSFETs turns on, effectively bypassing the pair of diodes that is conducting. The voltage-drop performance of the bridge is now a function of drain-to-source resistance (RDS(on)), which, with modern MOSFETs, is pretty darn good! In telephone line applications, a voltage drop in the millivolt range can easily be achieved. Also, with low-threshold MOSFETs achieving thresholds in the 1-volt range these days, it is possible to construct a bridge where the MOSFET turn-on occurs not long after the diode turn-on as the applied voltage ramps up.
A limitation of the circuit, as shown, is that the applied voltage cannot exceed the gate-to-source voltage (VGS) rating of the MOSFETs. Typically, this is 20 volts. For higher voltage applications, it is possible to put a resistor in series with each gate and use a zener clamp between the gate and source of each MOSFET to limit the VGS experienced by any individual MOSFET, as shown in Fig. 3. With such a provision, the primary limitation on applied voltage then becomes the drain-to-source breakdown (BVDS) rating of the MOSFETs.
One caveat of the FET bridge circuit: do not use it as the rectifier in front of a capacitor-input power supply! In a conventional rectifier bridge, the diodes prevent the backflow of current from the power supply input capacitor as the applied voltage drops below the voltage on the capacitor. With this design, the MOSFETs act like switches rather than one-way valves for current flow. They don’t care which way current flows, hence the input capacitor of the power supply will be discharged to near zero volts with each half-cycle of the applied AC power! This limits the power supply applications for this circuit to inductive- or resistive-input designs.
However, it would be possible to use this circuit with a polarized capacitor in power-factor correction applications. Correction of an inductive power factor would normally require a non-polarized capacitor directly across the AC line. By putting the FET bridge circuit in front of the capacitor, a polarized capacitor could be used instead which may be advantageous in terms of size and cost. I haven't tried this particular application, so I can't vouch for it, but if you have success with the idea, please let me know.