Push-pull output stage

from Wikipedia, the free encyclopedia
Two tubes for a transmitter output stage in push-pull circuit

A push-pull output stage is an electronic circuit and is used in the field of output stages. It is part of an amplifier circuit and has the task of amplifying an electrical signal to such an extent that devices connected to it can be operated. The name is derived from the fact that two components in the circuit work in opposite ways, whereby only one of the two is active depending on the design.

General

Applications of push-pull output stages are, for example, audio amplifiers , voltage converters or transmitters . Push-pull output stages are built with transistors such as bipolar transistors or field effect transistors . Historic designs use electron tubes .

A working resistance is not required with the push-pull circuit. As a result, a significantly higher degree of efficiency is achieved compared with other switching principles. In the case of power amplifiers, this is a major advantage over "single-ended circuits".

A special form of the push-pull stage is the bridge circuit . This has the advantage that, despite a unipolar voltage supply, both positive and negative voltage can be output.

Complementary output stage

Ironless power amplifier with complementary transistors
Complementary output stage in a CMOS inverter

The picture shows the basic circuit of a complementary output stage with single-ended control and asymmetrical operating voltage. The advantage here is that the direct current series connection of transistors Q4 and Q5 makes expensive transformers superfluous (ironless output stage). However, a base bias is necessary for the output stage transistors in order to minimize the takeover distortion, which is done with the two diodes D1 and D2. If these diodes are connected to the transistors in a thermally conductive manner, the forward voltage of the diodes changes to the same extent as that of the base-emitter paths of the transistors, which largely compensates for a change in the operating point (temperature compensation). Both Q4 and Q5 are just not conducting any quiescent current ; B operation is present. A higher base bias voltage (e.g. with three diodes) and the addition of emitter resistors can also be used to set operating modes with quiescent current (AB or A operation). The signal quality increases, but the power loss also increases.

Circuit description: R1 = 100 kΩ and R2 = 20 kΩ ensure that a voltage of 3.3 V is set at the connection point. Q1 and Q2 form a differential amplifier that compares the fraction R7 / (R7 + R8) of the output voltage with this 3.3 V and immediately takes any deviation as an opportunity to counteract this via Q3. The output voltage at the connection between Q4 and Q5 should be half the operating voltage so that the dynamic range is symmetrical upwards and downwards. This results in the values ​​R7 = 20 kΩ and R8 = 40 kΩ.

An electrolytic capacitor of around 1000 µF is placed in series so that direct current does not flow through the loudspeaker, which would heat it up and Q4 .

If the input voltage increases by 1 V, the output voltage must increase by 3 V so that the differential amplifier stops adjusting via Q3. The circuit thus amplifies the voltage by a factor of 3. If the input voltage increases by 1 V, the signal source must supply 1 V / 20 kΩ = 50 µA. However, Q4 and Q5 can safely conduct 10,000 times more current from the power supply to the loudspeaker, so the control power at the input is amplified 30,000 times.

This internal circuit can be found with minor modifications in many ICs that contain all components except the two capacitors. With an operating voltage of 20 V, the output voltage can deviate a maximum of 10 V upwards or downwards from the mean value, which enables an effective value of the output voltage of 7 V with a sinusoidal shape . With a 4-ohm loudspeaker, the maximum power is then 12 W. In practice, you still have to Subtract just under 1 W per power transistor and can then calculate with 10 W.

Quasi-complementary output stage

Circuit diagram of the quasi-complementary output stage

The quasi-complementary output stage , also known as the totem pole output in the field of digital technology , consists of two transistors of the same type. Until the 1970s, two PNP transistors were used for this, as there were no reliable germanium power transistors in NPN design. With the advent of silicon-based power transistors, this type of circuit was also used for NPN transistors until complementary types were available.

Quasi-complementary circuits are hardly used today for discreet audio amplifiers. In integrated circuits, however, they still play a role. In particular TTL gridset use the totem-pole circuit and an optimized short switching times push-pull output stage at the output.

The picture on the right shows an example circuit: T2 and T3 form the output stage, with T2 working as a collector circuit and T3 as an emitter circuit. T1 serves as a driver. The advantage of this circuit is that no NPN-PNP complementary type with identical electrical parameters is required. However, the CMOS technology has now largely replaced TTL and works exclusively with complementary transistors.

Push-pull output stage with transformer

Push-pull B output stage with two npn transistors

The picture on the right shows a basic circuit of a push-pull B output stage with two npn transistors. The push-pull control is carried out by the characteristic two transformers , the input transformer and the symmetrical output transformer, which reassembles both half-waves. This type of circuit is particularly widespread today in tube power amplifiers; in transistor amplifiers it was only common in the early days (until the 1960s) for devices with pnp transistors.

The operating point is set with the base voltage divider R1 / R2: The voltage at R1 must not exceed 0.55 V for silicon transistors, otherwise the quiescent current will rise sharply and the transistors may overheat. The working curve of each half is S-shaped. When graphically added, the characteristic curve of the push-pull amplifier results in a twofold S-shaped characteristic curve, from which a broad spectrum of harmonics with dominance of the odd harmonics results through Fourier transformation. Since the entire circuit is not linearized by any negative feedback, distortions are to be expected.

Web links