Push-pull converter

From Wikipedia, the free encyclopedia

A push-pull converter is a type of DC to DC converter that uses a transformer to step the voltage of a DC power supply. Since it uses a transformer, its ratio is arbitrary but fixed. The primary advantages of push-pull converters are their simplicity and ability to scale up to high power throughput, earning them a place in industrial DC power applications.

The push-pull converter is similar to the flyback converter and especially the forward converter.

1 Circuit operation
2 Transistors
3 Timing
4 See also
5 External links

[edit] Circuit operation

The push-pull is an extremely simple circuit. The switches alternate the voltage across the supply side of the transformer, causing the transformer to function as it would for AC power and

[edit] Transistors

Usually the signal for switching is on some logic levels (lets say between 0 and 3 Volts) and can supply very little current.

The power-transistors are supplied with 24 Volt, for example.

N-type and p-type power transistors can be used. This is like CMOS is constructed. The gate (base) of the power transistors is tied via a resistor to one of the supply voltages. A p-type transistor is used to pull up the n-type power transistor gate (common source). A n-type transistor is used to pull down the p-type power transistor.

All power transistors can be n-type (often 3 times the gain of p-type). This is like TTL is constructed. Then the n-type transistor, which replaced the p-type has to be driven this way: The voltage is amplified by one p-transistor and one n-transistor in common base configuration to rail-to-tail amplitude. Then the power transistor is driven in common drain configuration to amplify the current.

In high frequency applications both transistors are driven with common source. In fact they are both pushing, pulling is done by a low pass filter ( coil ) in general and by a center tap of the transformer in the converter application. Because the transistors push alternating this device is also called a push-pull converter.

[edit] Timing

If both transistors are open, this is a short circuit. If both transistors are closed, high voltage peaks due to back EMF appear.

If the driver for transistor is powerful and fast enough, the back EMF has no time to charge the capacity of the windings and of the body-diode of the mosfets to high voltages.

If a microcontroller is used, it could measure the peak voltage and digitally adjust the timing for the transistors, so that the peak just appears (coming from no peak, starting from cold transistors in warm-up / boot-phase).

The cycle starts with no voltage and no current. The one transistors opens, a constant voltage is applied to the primary, current increases linearly, and a constant voltage is induced in the secondary. After some time T the transistor is closed, the parasitic capacities of the transistors and the transformer and the inductance of the transformer form an LC circuit which swings to the opposite polarity. Then the other transistor opens. For the same time T current flows back into the storage capacitor, then changes the direction automatically, and for another time T the current flows in the transformer. Then again the one transistor opens until the current is stopped. Then the cycle is finished, another cycle can start anytime later. The S-shaped current is needed to improve over the simpler converters and deal efficiently with remanence.