# How does a transistor act as an amplifier considering energy conservation?

We all know energy cannot be formed from nothing.

So how does a bipolar junction transistor (BJT) — for example — amplify the base current by beta and outputs it to collector current?

Where is the “catch?” Is there somewhere else where were losing energy?

• You are confusing the idea of energy with current and probably also failing to recall that there are voltage sources with stored energy in them as part of such circuits.
– jonk
Aug 29 '16 at 17:18
• Think of BJT's current variable resistors between supply and load with a sensitive input controlled by Beta. It is not efficient as switched voltage with very low resistance with duty cycle modulated to vary the output voltage to the load such as in class D amplifiers and SMPS. Aug 29 '16 at 17:39
• Don't forget that amplifier is a device that allow us control the flow of "high power" with the help of a "low power". In order to Amplifier effect occur, two things are necessary: source of energy (power supply) and a device for controlling the flow of this energy - > the amplifier.
– G36
Aug 29 '16 at 17:56
• The analogy I always liked for what a transistor does is the button on the top of an aerosol spray can - your tiny force (finger) is releasing a large force (compressed gas). Aug 29 '16 at 23:22
• @slebetman, please consider my resriction (DIRECT control without conversion of physical quantities in between). Hence, a small current cannot DIRECTLY control a large current. For BJT`s it is the electrical field within the depeletion region which controls the current. By the way, this applies to ALL conductive bodies - only the E-field causes the movement of charged carriers (which is current). And any change of the current is caused by a corresponding change of the E-field - and not by any other current. How should this work? Nobody has explained this up to now.
– LvW
Aug 31 '16 at 7:17

The base current in a transistor controls the collector current. The energy comes from the power supply. It is not generated within the transistor.

• To clarify: a small current applied to the base can control a larger current from collector to emitter; thus it "amplifies" your output (e.g. from an Arduino pin) to another device. Aug 30 '16 at 4:18
• To clarify better (in particular, the last paragraphs on page 1387): ieeexplore.ieee.org/stamp/stamp.jsp?arnumber=775421
– LvW
Aug 30 '16 at 7:39
• @LvW Can you point out whats wrong and which sentence is correct? That would be very helpful. Aug 30 '16 at 22:38
• The answer can be found in the link I have provided. Of course, the collector current is NOT controlled by the base current. It is the electrical field (determined by Vbe) that matters only.
– LvW
Aug 31 '16 at 7:09
• Because some people prefer the water analogy: Can somebody imagine that a change in the water flow of a small river would be able to change the water flow of a larger river WITH GAIN ? (Example: 1m³ per second (small river) causes a change of 100m³ per second (large river).
– LvW
Aug 31 '16 at 10:40

The 'catch' is that a transistor only controls the flow of current; it does not itself generate power. The power would come from some other part of the circuit, perhaps from the electric company via a power supply or from a battery.

Now, one possible point of confusion is that transistors can be represented as equivalent circuits that contain a so-called "current source". This sounds like it could generate power, yes? Well, it turns out that a "source" can actually either generate or absorb power, depending on the relationship between the current through it and the voltage across it. The main thing a "source" does is fix the current (in the case of a current source) or the voltage (in the case of a voltage source) to a specific value. For example, a lithium-ion battery acts like a voltage source. If you connect a resistor across the battery terminals, the battery will supply current to hold the voltage more-or-less constant. However, if you connect an external power supply, the battery will start to charge, absorbing energy while trying to hold the voltage constant.

Now, there are several different 'models' or 'equivalent circuits' of different types of transistors, all of which use dependent sources in some way. The trick is that these models are only valid under particular operating conditions, and it turns out that there isn't any set of conditions under which a transistor will ever generate power. This isn't a trick of the mathematics, the reason for this is that there is nothing inside of a transistor that's capable of generating power; the only thing a transistor can do is generate a voltage drop to oppose the flow of current. It usually turns out that transistors end up dissipating a lot of power and end up needing to be mounted on large heat sinks.

A BJT is generally used as a power amplifier and the the power gained by the output signal comes from the DC power source that it uses.

Amplifying current on its own can be done without power amplification using a transformer but, if you want a power amp (i.e. the product of volts and amperes increased) then you need a power source.

Energy is average power times time: $E = P \times t$. Energy is conserved. Since power is just instant moments of energy, it is also a fact that, in general, power is also conserved. So you will often see statements like "power in must be equal to or greater than power out." On average and on human scales of time and locality, that's pretty much also true. But it is the conservation of energy that really rules the universe as we understand it.

Power, itself, though is like a kind of coinage. Each coin has two faces: voltage and current. Power is the product of volts times current: $P = V \times I$. Note that neither of these are time. So there is no implication here that current ($I$) must be conserved, because voltage ($V$) can be adjusted. There is also no implication that voltage is conserved, because current can be adjusted. On average what is conserved is energy and power. Not voltage and current.

The energy per unit time (power) comes from the power supply. That energy and power provides heat and electrical and transducer activities and it does have to follow basic energy conservation ideas. But the BJT transistor consumes a small amount of "re-combination" current provided to the base-emitter region in order to activate a larger in-rush of collector current. What's missing here is any discussion of the voltages involved and the rest of the circuit, too, most especially including the energy sources. The BJT transistor draws from those sources and those sources lose energy well in excess of the very tiny local effect by the BJT as one tiny part of a much larger circuit and energy source system.

The conservation law applies to the closed system as a black box. But tiny local increases in energy are possible, so long as those increases come from somewhere else you are ignoring. Just as life itself on Earth may appear to be organization coming from disorganization and violate entropy laws, the fact is that the sun's entropy increases far, far more than any small local decreases in entropy represented by an isolated life form on Earth. You have to take an appropriately complete system when applying conservation laws.

It is possible to view such an amplifier as a circuit which transfers DC power (from the power supply) into signal power - available at the amplifiers output.

There are so-called "power amplifiers", which - however - do NOT amplify power. But it is their main task not to amplify a signal voltage but to PROVIDE as much signal power as possible at its output. And the output-to-input power ratio is the so-called "efficiency".

Lets consider a typical NPN transistor, and consider two electrical paths through it: collector to emitter and base to emitter. These are the only paths through which current will flow when the device is being used in a conventional manner as an amplifier.

When no forward bias voltage (relative to the emitter) is present on the base terminal, no significant current flows through any path in the device.

When voltage on the base terminal (relative to the emitter) is in the forward bias range (depends on the specifics of the device, but perhaps 0.6V), a small amount of current flows through the base to the emitter. This has the effect of altering the current-carrying ability of the device in the other path - from the collector to the emitter. The small amount of current flowing from base to emitter permits a much larger current to flow from collector to emitter. Base-emitter current doesn't drive collector-emitter current. Think of it more like operating a valve. In the forward bias range, tiny changes in base-emitter current causes big changes in the current-carrying ability of the collector-emitter path. Thus we have an amplifier. To flow through the collector, current still has to be provided by an external source (battery, power supply, etc.).

Pretty much anything described as an amplifier works in an analogous way, whether electrical or even mechanical. There is a power supply which provides the output power, an input signal, and the amplification device which simply controls the flow of power through the device in response to the input signal. No conservation laws need be violated.

A transistor is a ... transformable resistor. It work as the 'tap' at your basin. When you with a little force turn the knob of the tap, the water flows out but it's not 'generate' by the tap. In the same way, when you drive with a small current into the base of a transistor (B-E junction), a greater current flows through the collector (C-E junction) but the current it's not 'generate' inside the transistor. The power (voltage and current) is generated by the power supply, in the same way that the pressure and the flow is not generate by the tap of your basin but by the water plant of your city.

• Welcome to EE.SE, umlibero. I notice you are adding answers to several old questions. This one had an answer accepted four years ago and your answer is similar to several others. Watch the new questions instead or ones that have not got an accepted answer! Nov 15 '20 at 22:28