Editing Push–pull output (section)

== Analog circuits ==
A conventional amplifier stage which is not push–pull is sometimes called [[Single-ended triode|single-ended]] to distinguish it from a push–pull circuit.

In analog push–pull power amplifiers the two output devices operate in [[antiphase]] (i.e. 180° apart). The two antiphase outputs are connected to the load in a way that causes the signal outputs to be added, but distortion components due to non-linearity in the output devices to be subtracted from each other; if the non-linearity of both output devices is similar, distortion is much reduced. Symmetrical push–pull circuits must cancel even order harmonics, like 2f, 4f, 6f and therefore promote odd order harmonics, like f, 3f, 5f when driven into the nonlinear range.

A push–pull amplifier produces less [[distortion]] than a single-ended one. This allows a [[Power amplifier classes#Class A|class-A]] or [[Power amplifier classes#Class AB|AB]] push–pull amplifier to have less distortion for the same power as the same devices used in single-ended configuration. Distortion can occur at the moment the outputs switch: the "hand-off" is not perfect. This is called crossover distortion. [[Power amplifier classes#Class AB|Class AB]] and [[Power amplifier classes#Class B|class B]] dissipate less power for the same output than class A; general distortion can be kept low by [[negative feedback]], and crossover distortion can be reduced by adding a 'bias current' to smoothen the hand-off.

A class-B push–pull amplifier is more efficient than a class-A power amplifier because each output device amplifies only half the output waveform and is cut off during the opposite half.  It can be shown that the theoretical full power efficiency (AC power in load compared to DC power consumed) of a push–pull stage is approximately 78.5%.  This compares with a class-A amplifier which has efficiency of 25% if directly driving the load and no more than 50% for a transformer coupled output.<ref name=Yunik73>Maurice Yunik ''Design of Modern Transistor Circuits'', Prentice-Hall 1973 {{ISBN|0-13-201285-5}} pp. 340-353</ref> A push–pull amplifier draws little power with zero signal, compared to a class-A amplifier that draws constant power. Power dissipation in the output devices is roughly one-fifth of the output power rating of the amplifier.<ref name=Yunik73/> A class-A amplifier, by contrast, must use a device capable of dissipating several times the output power.

The output of the amplifier may be direct-coupled to the load, coupled by a transformer, or connected through a dc blocking capacitor. Where both positive and negative power supplies are used, the load can be returned to the midpoint (ground) of the power supplies. A transformer allows a single polarity power supply to be used, but limits the low-frequency response of the amplifier. Similarly, with a single power supply, a capacitor can be used to block the DC level at the output of the amplifier.<ref>Donald G. Fink, ed. ''Electronics Engineer's Handbook'', McGraw Hill 1975 {{ISBN|978-0-07-020980-0}} pp. 13-23 through 13-24</ref>

Where bipolar junction transistors are used, the bias network must compensate for the negative temperature coefficient of the transistors' base to emitter voltage. This can be done by including a small value resistor between emitter and output. Also, the driving circuit can have silicon diodes mounted in thermal contact with the output transistors to provide compensation.
{{Further|OCL amplifier}}

=== Push–pull transistor output stages ===
[[File:Aura VA 100 Evolution 2 (4061759992) - closeup of output stage.jpg|thumb|upright=1.5|Typical transistor output stage of one channel of a 65 watt stereo amplifier from 1993. The 2 MOSFET push-pull output transistors (''FET2, FET4'') are bolted to the black [[heat sink]].  They are driven by transistors ''Q2, Q5, Q6,'' and ''Q7'']]
{{more citations needed section|date=November 2012}}
Categories include:

==== Transformer-output transistor power amplifiers ====
It is now very rare to use output transformers with transistor amplifiers, although such amplifiers offer the best opportunity for matching the output devices (with only PNP or only NPN devices required).

==== Totem pole push–pull output stages ====
Two matched transistors of the same polarity can be arranged to supply opposite halves of each cycle without the need for an output transformer, although in doing so the driver circuit often is asymmetric and one transistor will be used in a [[common-emitter]] configuration while the other is used as an [[emitter follower]]. This arrangement is less used today than during the 1970s; it can be implemented with few transistors (not so important today) but is relatively difficult to balance and to keep a low distortion.

==== Symmetrical push–pull ====
Each half of the output pair "mirror" the other, in that an NPN (or N-Channel [[FET]]) device in one half will be matched by a PNP (or P-Channel [[FET]]) in the other. This type of arrangement tends to give lower distortion than quasi-symmetric stages because even harmonics are cancelled more effectively with greater symmetry.

==== Quasi-symmetrical push–pull ====
In the past when good quality PNP complements for high power NPN silicon transistors were limited, a workaround was to use identical NPN output devices, but fed from complementary PNP and NPN driver circuits in such a way that the combination was close to being symmetrical (but never as good as having symmetry throughout). Distortion due to mismatched gain on each half of the cycle could be a significant problem.

==== Super-symmetric output stages ====
Employing some duplication in the whole driver circuit, to allow symmetrical drive circuits can improve matching further, although driver asymmetry is a small fraction of the distortion generating process. Using a [[bridge-tied load]] arrangement allows a much greater degree of matching between positive and negative halves, compensating for the inevitable small differences between NPN and PNP devices.

==== Square-law push–pull ====
The output devices, usually [[MOSFET]]s or [[vacuum tube]]s, are configured so that their [[Power-law#square-law|square-law]] transfer characteristics (that generate second-harmonic [[distortion]] if used in a single-ended circuit) cancel distortion to a large extent. That is, as one transistor's gate-source voltage increases, the drive to the other device is reduced by the same amount and the drain (or plate) current change in the second device approximately corrects for the non-linearity in the increase of the first.<ref>{{cite journal | author=Ian Hegglun | title=Practical Square-law Class-A Amplifier Design | journal=Linear Audio  |volume=1}}</ref>

=== Push–pull tube (valve) output stages ===
{{See also|Valve audio amplifier – technical#The push–pull power amplifier}}
[[Vacuum tube]]s (valves) are not available in complementary types (as are PNP/NPN transistors), so the tube push–pull amplifier has a pair of identical output tubes or groups of tubes with the [[control grid]]s driven in antiphase. These tubes drive current through the two halves of the primary winding of a center-tapped output transformer. Signal currents add, while the distortion signals due to the non-linear [[Current–voltage characteristic|characteristic curve]]s of the tubes subtract. These amplifiers were first designed long before the development of solid-state electronic devices; they are still in use by both [[audiophile]]s and musicians who consider them to sound better.

Vacuum tube push–pull amplifiers usually use an output transformer, although [[Output transformerless|Output-transformerless (OTL)]] tube stages exist (such as the SEPP/SRPP and the White Cathode Follower below).{{citation needed|date=December 2012}} The phase-splitter stage is usually another vacuum tube but a transformer with a center-tapped secondary winding was occasionally used in some designs.  Because these are essentially square-law devices, the comments regarding [[Distortion#Cancellation of even-order harmonic distortion|distortion cancellation]] mentioned [[Push–pull output#Square-law push–pull|above]] apply to most push–pull tube designs when operated in [[Power amplifier classes#Class A|class A]] (i.e. neither device is driven to its non-conducting state).

A '''Single Ended Push–Pull''' ('''SEPP''', '''SRPP''' or '''mu-follower'''<ref>{{cite web|title=SRPP Decoded|url=http://www.tubecad.com/may2000/|website=The Tube CAD Journal|access-date=7 November 2016}}</ref>) output stage, originally called the '''Series-Balanced amplifier''' (US patent 2,310,342, Feb 1943). is similar to a totem-pole arrangement for transistors in that two devices are in series between the power supply rails, but the input drive goes ''only to one of the devices,'' the bottom one of the pair; hence the (seemingly contradictory) Single-Ended description. The output is taken from the cathode of the top (not directly driven) device, which acts part way between a constant current source and a cathode follower but receiving some drive from the plate (anode) circuit of the bottom device. The drive to each tube therefore might not be equal, but the circuit tends to keep the current through the bottom device somewhat constant throughout the signal, increasing the power gain and reducing distortion compared with a true single-tube single-ended output stage.

The transformer-less circuit with two tetrode tubes dates back to 1933: "THE USE OF A VACUUM TUBE AS A PLATE-FEED IMPEDANCE." by J.W.Horton in the Journal of the Franklin Institute 1933 volume 216 Issue 6

The '''White Cathode Follower''' (Patent 2,358,428, Sep. 1944 by E. L. C. White) is similar to the SEPP design above, but the signal input is to the ''top'' tube, acting as a cathode follower, but one where the bottom tube (in common cathode configuration) if fed (usually via a step-up transformer) from the current in the plate (anode) of the top device. It essentially reverses the roles of the two devices in SEPP. The bottom tube acts part way between a constant current sink and an equal partner in the push–pull workload. Again, the drive to each tube therefore might not be equal.

Transistor versions of the SEPP and White follower do exist, but are rare.

==== Ultra-linear push–pull ====
A so-called [[ultra-linear]] push–pull amplifier uses either [[pentode]]s or [[tetrode]]s with their [[screen grid]] fed from a percentage of the primary voltage on the output transformer. This gives efficiency and distortion that is a good compromise between triode (or [[Pentode#Triode-strapped pentode circuits|triode-strapped]]) power amplifier circuits and conventional pentode or tetrode output circuits where the screen is fed from a relatively constant voltage source.