Analogue amplification stages incorporating just a single active device, such as a transistor or op-amp, – as in a DC amplifier -  usually allow amplification of either the negative-going or the positive-going part of an AC waveform alone; basically cutting out half of the cycle and therefore causing immense distortion. There are two ways of overcoming this: Those being amplifying both halves of the waveform separately and recombining them at the output; perhaps at a decoupling transformer, (Too much copper wire, inductance, and consequential weight for my liking.) or by using two interconnected active devices to amplify both halves of the waveform at once, before passing the full AC waveform on to the next stage.

The latter method can be achieved with a push-pull amplifier.

Pictured below is a circuit diagram of a very basic push-pull amplification stage. (Please excuse the freehand drawing.)

The input signal; an alternating waveform, is fed via capacitors 1 & 2 to the bases of Q1 and Q2. The functions of C1 & 2 individually and collectively are two-fold: The first is that they individually shield the base connections of their respective transistors from any stray DC voltages from a previous stage, also they separate any DC potentials present at the base junctions of Q1 & 2; therefore preventing unintentional and accidental biasing of one another.

Resistors 3 & 4 act as a potential divider biasing the base of Q1 to 0V7. Resistors 3 & 4 have the a similar effect effect on the base of Q2: However since Q2 is a PNP transistor, the values of resistor used in the case of the R3,4 pair are reversed; therefore giving the base of Q2 a negative bias with respect to that of Q1.

Preset potentiometers PR1 & 2 set the potential of the transistor-pair with respect to the supply rails; and consequently the swing-maximum of the output-waveform. Resistors R5 & 6 are a precaution to avoid the peak output level colliding with the supply voltage and therefore causing distortion.

Capacitor C3 provides AC decoupling for the transistor pair.

Note that the emitters of the transistors are connected together and the output taken from that connection. This allows the inclusion of the traditional collector load resistance in both cases.

A waveform appearing at the input flows through both C1 and C2 to the base of the individual transistors. If the waveform is on the positive-going half of the cycle it lowers the conduction of Q2 and raises the conduction of Q1. If the waveform is on the negative-going half of the cycle the reverse occurs: Hence the output polarity mirrors the input polarity to whatever degree of amplification is involved.

A Little Background Information:

Prior to the advent of digital electronics, this type of circuit configuration was widely used in analogue receiving and amplification devices throughout the 1950s, 60s, and to some extent even the 70s. Back in the 1950s before the transistor became widely used in electronic circuitry, they would use a pair of triode or pentode  thermionic valves in the place of the transistor pair. As technology developed the manufacturers developed smaller valves with two triode or pentode sections for the purpose, screened from one another by a metal electrode.

(Example: ECC82 (European Nomenculare), or equivalent 12AU7 (American nomenculare.) AF double-triode with a 6.3 Volt heater supply. The European equivalent with a higher heater voltage was the UCC82 which required a 32 Volt heater supply.)

Valves were also manufactured with a pre-amplification or oscillator stage included, usually a triode; along with a main amplification device, usually a pentode. (Example ECL85 (6.3V heater supply), UCL85(32 V heater supply), and PCL85 (17.5 Volt heater supply, commonly used in the audio output stages of televisions. – Right up until around 1973.))

During the late 1960s/early 70s, valves began to be excluded from the designs of electronic devices, in preference for the transistor; which was lighter, lasted a lot longer, required less voltage to function, and didn’t require an internal heater powered from an external source to make it work.

(There was a period at the very end of the 1960s which lasted a little way into the 70s where equipment manufacturers would produce valve/transistor hybrids, especially in the case of televisions. These exhibited a few benefits over valve-only technology; such as they took less time to warm up before they started working, and the amount of mains-hum distortion was reduced to a large extent.)

There: A free history-lesson along with the main subject. There’s value-for-money; even though there was no charge in the first place.

Personally Speaking:

Just in case you’re wondering, I do just about remember those old days mentioned; especially the latter valve/transistor technology. I was just getting into electronics in those days. – And yes I was rather young to be messing about inside televisions et al. I practiced hobby electronics until fairly recently, when I got qualifications in the subject after a crash-refresher course at college. I became interested in computers in the late 1970s, around the time the Commodore Pet was released to market. Upon leaving school I followed the arts for a while, at the same time as running a small hobby-enterprise in analogue electronics, until I got back into both computers and digital electronics in the late 1990s.

(Oh yes. – Just in case you were wondering; I do remember the large B9D-base line-output pentode valve used in some televisions right into the 1980s; although I can’t remember the alphanumeric designation offhand. It started with a “P”, but that’s pretty obvious. – Most television valves did, other than maybe the HF triode-pentodes in the UHF/VHF tuners of some 1960s models. – Apart from some of the B8A valves of the early 1960s with the metallic base. – Now that is going back a bit too far.)

Ah I just remembered: PL504.  

. . . . . . . . .  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

I was just proof-reading after writing this lot and I remembered; I need to hyperlink. There is so much I should hyperlink. If this article takes longer than expected to produce then that’s part of the reason why.

Oh wow; this’ll be fun!

Advertisment : Word Press Link Cloak: Easily disguise and protect your affiliate links to increase your revenue. Click Here!

When designing a single-transistor-amplifier stage; be it an AF, RF, IF, whatever project, one must always take into account the gain of the transistor in question and design for it accordingly. It’s best not to attempt to utilise the full amount of available gain as the transistor will probably saturate and cause distortion. In fact a tri-stage amplifier with negative-feedback, (The subject of a later article maybe.) with only half the available gain of each transistor used will produce a much better and less distorted output than a single-transistor stage utilising all of the available gain.

For now, however, I’m going to concentrate on just a single-transistor stage, and on correctly biasing the transistor’s base proportional to its base/collector.

Let’s assume that our subject transistor has a voltage-gain of 20, and a Vb(max) of 3.3V. Let’s place it in circuit:

Ignoring the type and strength of input signal as far as this example is concerned, we intend to bias the base of the transistor from the supply rails using a straightforward potential divider (R1 & 2). We know that the base voltage must not exceed 3.3V, and we’re running the circuit from a 6 Volt rail. We use the equation V=Vin (+Vcc in this case, using a NPN transistor.) X R2 / (R1 X R2). In the example above, (Which is an example only rather than a functional circuit.) I’ve taken a guess and used a 1K negative resistor (R2) and a 2.7K resistor (R1) to the rail. This gives a working base voltage of 2.223 Volts. In most cases we’d tweak that voltage, by altering the resistance values, to as near 0.7 Volts as possible. (The transistor’s transconductance threshold, assuming that we are using a silicon rather than a germanium transistor.) For this example we’ll leave it at 2.223 Volts.

Note the use of resistances in the kilohm range: This is for the purpose of limiting the base current. Although we haven’t specified a current for the input signal we’re assuming that it’s in the several-tens-of-milliamperes range. When dealing with even smaller input signal currents use higher resistance to limit the biasing current further so as not to interfere with the input signal.

We can work out the base-bias current precisely using Ohm’s Law: I=V/R; 6 volts divided by (R1 + R2 = 3700Ohms) = 1.62 milliamperes: We’d therefore be looking for a gain of about 10, driving the transistor at half-available-gain, so it would be fair to say to use a total DC resistance of 100 Ohms in the emitter circuit, and a total DC resistance of 270 Ohms in the collector circuit, ignoring any reactance from any decoupling capacitor (Such as C3 in the second circuit.) in the emitter circuit.

Looking at that second circuit we note that there are DC decoupling capacitors (C1 & C2) on the input and output of the stage: these are included to prevent any DC component bleeding back into the stage before, as well as bleeding off from the collector circuit into the following stage. Note that I’ve sketched in a decoupling capacitor into the base circuit using a broken line: This may or may not be a good idea depending upon many factors; and that’s well beyond the scope of this post.

Advertisment : Word Press Link Cloak: Easily disguise and protect your affiliate links to increase your revenue. Click Here!