You are asking very important fundamental questions, whose answers seem easy for many experts, but is actually more difficult than they seem, because
1. "active/passive device" definitions are often used improperly (or in a misleading way).
2. systems exhibiting linear behavior are often considered to be a "more likely and more natural" behavior than a non-linear behavior.
3. transistors are useful not only for amplification but also in digital/switching applications.
I am aware that now I should start to clarify these issues and then give you a satisfying answer. However, I need some time to write all these, because I see it will not be easy to make them "convincing". It's funny to ask for this, but please give me some time to complete my own work (whose deadline has come) and I will return soon to complete my answer. :)
one part of the answer is: For circuits made exclusively from linear elements, the superposition principle applies: If you connect the circuit to several sources (of voltage or current) the resulting voltages and currents inside the circuit are just the sum of the effects of each source connected to the circuit separately (the other sources replaced by shorts for voltage sources and breaks for current sources). As a consequence, the only frequencies occurring inside the circuit are those introduced by the sources.
The basic idea behind electronic amplifiers is to have a circuit with three ports: a signal input, a signal output, and another input for the DC supply. The circuit takes energy from the DC supply, and shifts (part of) it to the frequency or spectrum of the input signal, providing an amplified output signal.
With a linear circuit providing the three ports mentioned above, the output voltage (or current) is just the sum of the output voltage caused by the input signal and the output voltage caused by the supply voltage. In the absence of the supply voltage, the output signal cannot be larger than the input signal (where would the additional energy come from?). The presence of the supply voltage does not change the output signal caused by the input. So, the output of any linear circuit can provide only the (in many cases even attenuated) input signal plus a DC voltage.
With non-linear elements, the superposition principle doesn't apply, thus enabling implementations of the above mentioned idea of amplification.
If, for practical purposes, we restrict the term "passive" to resistors, capacitors, inductors, and diodes (rectifiers), we find that circuits comprising exclusively passive components can generate output frequencies not included in the input, provided that at least one element behaves non-linearly. In particular, harmonics can be generated, that is, frequencies which are an integer multiple of the base frequency. E.g., the power of a 1 kHz input signal can be distributed to 1 kHz, 2 kHz, 3 kHz etc. But there are no harmonics to the base frequency 0 Hz, so no power is shifted from DC to any other frequency.
To turn DC power into AC power is the outstanding ability of active elements.
If only using passive component to build circuits will get loading effect between input nodes and output nodes. Then the signals in them will not to be under control, the system is easy to go into unstable.
The best advantage of active components is that avoide loading effect.
There are 2 kinds of amplification to be discerned:
voltage amplification
power amplification.
The first is easily implemented with passive elements: for any AC signal with a DC component of 0, a transformer can do the trick. And a transformer is linear (within given limits) and passive.
For the latter let me argue from the energy view: the (power) amplified signal contains more energy than the input signal. On the other hand, passive components are not able to inject energy. So a power amplifier must contain some active element(s) that provide the power difference.
Whether you need nonlinear elements: this might be a philosophical discussion: at least on the macro level an OpAmp is a linear device capable to amplify a signal. From this point of view, a nonlinear element is not necessary to implement power amplification.
As for the 3rd question ("... shift the dc potential to some frequency that the input signal contains?"): sorry - I do not understand this question.
Yours is a very interesting question. And, in fact, it can be difficult to even tell which electrical components are actually passive or active. Whatever answers you are provided by the RG members, they must be able to account for the behavior of the following power amplifiers: 1) the carbon button microphone as an amplifier, and 2) the magnetic amplifier.
One usually does not think of a two--terminal electrical transducer such as the carbon button microphone - found in the handset of older telephones - as an active device capable of power amplification, but it is nevertheless. For citations discussing this active device, see pp. 26-29 [PDF pages 34-37] of the following thesis available on RG as Research Coherers, A Review
.
As has already been discussed by U. Dreher , the transformer is capable of increasing an AC voltage, but it normally acts as a passive device and so does not provide power amplification. However, special versions of the iron-core transformer were used to provide power amplification, and were dubbed 'magnetic amplifiers', see the following citations:
[1] Anonymous (Prepared by the Bureau of Naval Personnel); Basic Electricity (Second revised and enlarged edition); Dover Publications, Inc.; 1970; pp. 393-414, Chapter 20 - Magnetic Amplifiers.
[2] G. F. Pittman, Jr., R. O. Decker; Magnetic Amplifiers; in Eugene M. Grabbe, Simon Ramo, Dean E. Wooldridge; Handbook of Automation, Computation, and Control, Volume 3, Systems and Components; John Wiley & Sons, Inc.; 1961; pp. 25-01 - 25-39.
1) Passive elements can not provide amplification. Amplification can ONLY be done with the help of active elements.
2) If we look into the I-V curve of an element, the V/I ratio must be negative in at least one of the quadrants for an active element (i.e. V/I ratio must be negative either for V>0 or V
It is a really good question and it is worth discussing. I will try to be intuitive.
1) Actually, passive elements provide voltage boosting, and it is called passive voltage-boosting which you can especially observe in the matching circuits. (it is frequently used in harvester circuits to exceed the threshold voltage of the diodes or MOSs) Transformers are the other good example that U. Dreher has already mentioned. However, you cannot provide power amplification using a passive device. I need to add that I do not like to say power amplification because it sounds like we are violating the thermodynamic laws. It might be better to say that an active device uses the DC battery to amplify the power at the input, but overall the power supplied by the battery plus the power of the input signal is equal to power at the load plus power converted to heat energy (PDC+PIN=PLOAD+PHEAT). The most popular example might be a sound card in a speaker which receives a small power at the input but delivers high power to the load (driving the speaker coil) by using a DC battery or power grid so that we can hear the sound. Briefly, an active device uses the DC battery and amplifies the power of the input signal without violating the thermodynamic laws (of course). You cannot do something like that with a passive device.
Ali Zeki made a good point that an active device may also provide switching. Normally, a switch is not an active device, but can be realized by an active device (transistor) and also by a passive device (diode).
An interesting point is that even if we call a diode a passive device, Gunn diodes are called active devices because they can provide negative resistance. Negative resistance is another function that cannot be realized by a passive device. It also uses the DC battery to be activated or to be "biased".
2) It is not a necessity but it is about nature. But, there is no causation between being an active device and being a nonlinear device (correlation is not causation). For instance, you can use a MOSFET in the linear region (triode) and it would still be an active device and it would still provide power amplification. Another example is CMOS at velocity saturation region gives a linear ID-VGS characteristic (check short channel effects in MOSFET). In general, devices have exponential (or square law->it is not entirely true for modern CMOS processes) characteristics due to diffusion mechanism (BJT, diode, subthreshold region MOS). This is why we often encounter nonlinear devices (even a simple resistor is not entirely linear). Briefly, nature is rarely linear (or the physical phenomenas). For a small signal (which is the case for sensor output) we can assume that these non-linear devices behave as linear, which is good for us. The output signal must include the same frequency as the input signal. Otherwise, we would hear some weird signals from the speaker. Although, Nonlinear processing includes a lot of functions (rectifier, limiter, automatic gain control, etc.), linear time-invariant systems and linear control are crucial in the real life. Thus, it is the general case that we use non-linear devices by linearizing them, exploiting the small input signal, or by applying negative feedback.
Let's summarize the differences between an active and a passive device.
1) Active devices need to be powered by an external DC source to function properly (We generally call it DC biasing). Passive devices can properly operate without any external source. For instance, a simple resistor or a capacitor does not need to be powered. But as the function-wise an active device injects power into the load (the efficiency may change).
2)Active devices can provide negative resistance while passive devices cannot.
3) Both of them can provide voltage and current gain. (Passive voltage boosting, transformers .etc)
4) Both of them can operate as a switch but a switch is a passive device because it does not need to be powered and does not inject power to the load.
5)Both of them may be linear or non-linear (for instance a diode is non-linear, the resistor is linear, MOS transistor at strong inversion region is nonlinear while MOS transistor at velocity saturation is region is linear).
We should not forget that there is no perfect linearity in nature, everything is nonlinear.
1) Active amplifying element has to show some non-linearity at some stage otherwise it would amplify signal to infinite power.
2) Passive usually implies perfectly linear so it cannot combine = create some interactions to convert two voltages/currents to pump energy from one to another.
However once you have some non-linearity, you can make some amplifier e.g. magnetic amplifier, varactor parametric amplifier, etc.
I will answer the second question first because it is more difficult and as far as I know it is not still answered. So, "Why is it necessary for the active elements to be non linear?"
My answer is simple and a little paradoxical: It is necessary that active elements are non-linear to make linear devices with them (by "non-linear" I mean "non-linear in general but linear in their working part"; for example, for transistors, this is the horizontal part of their output IV curve).
I will use a purely "geometrical" explanation of this phenomenon based on the graphical representation of the transistor amplifier operation. It consists of two intersecting curves, one of which is the so-called "load line" and the other is the transistor output IV curve. When the input voltage varies, the transistor curve moves to remain parallel to itself and the intersection (operating) point moves along the load line.
Obviously, in order to have a linear relationship between the input and output voltage, the horizontal part of the transistor output curve must be a straight line (so two straight lines will intersect). But notice that the whole transistor IV curve is non-linear (consists of an almost vertical and an almost horizontal part). It would be linear (starting from the coordinate origin) if the transistor was a linear (ohmic) resistor; but it is а non-linear current-stabilizing "resistor".
The actual "active elements" in the circuit are only the power sources. Tubes, transistors, etc. - only controls the distribution of energy from the power source - in accordance with the control signal. From this basic point of view, tubes, transistors, op amps (today always "made up" of some type of transistors), etc. are not actually active elements.
These elements are able to function only after setting a suitable working point. "By their physical nature" their properties are described (modeled) by non-linear functions.
In linear applications (amplifiers, filters, oscillators), we perform linearization (of properties) - in a small vicinity of the working point.
In non-linear applications (mixers, modulators, ...) we use the non-linear properties of tubes and transistors (different types of diodes). Or we construct more complex circuits - e.g. multipliers.
My insight is that we have three types of linearity - on the entire curve (represented by ohmic resistance), on a section of the curve (constant-current resistance), and at a point on the curve (differential resistance). The most desired by us is the second (using a large signal) but we are forced to use the third (small signal).
Passive elements, such as resistors, capacitors, and inductors, can only store, dissipate, or transform energy. They cannot amplify it. To amplify a signal, you need an active element, such as a transistor. Transistors are able to amplify signals because they can control the flow of current. This allows them to take a small input signal and produce a larger output signal.
Active elements are necessary for amplification because they can provide the gain that is required to increase the amplitude of a signal. Passive elements, on the other hand, cannot provide gain. They can only attenuate or pass signals.
Active elements are also necessary for amplification because they can provide the nonlinearities that are required to create new frequencies from the input signal. Passive elements can only produce signals that are at the same frequencies as the input signal. Here are some of the reasons why active elements must be nonlinear:
To create new frequencies: Passive elements can only produce signals that are at the same frequencies as the input signal. Active elements, on the other hand, can create new frequencies by mixing the input signal with its own signal. This is done by using the nonlinear properties of the active element.
To amplify signals without distortion: If an active element were linear, it would amplify all frequencies equally. This would result in distortion of the signal, as the different frequencies would be amplified by different amounts. Nonlinearity allows the active element to amplify different frequencies by different amounts, which helps to preserve the shape of the signal.
To provide gain: Gain is the ability to increase the amplitude of a signal. Passive elements cannot provide gain, as they can only attenuate or pass signals. Active elements, on the other hand, can provide gain by controlling the flow of current. This allows them to take a small input signal and produce a larger output signal.
Here are some of the most common active elements used in amplifiers:
Transistors: Transistors are the most common type of active element used in amplifiers. They are made from semiconductor materials, such as silicon, and they can be used to amplify signals in a variety of ways.
Vacuum tubes: Vacuum tubes were the first type of active element used in amplifiers. They are made from glass and they contain a vacuum. A heated filament in the vacuum tube creates electrons, which are then amplified by the tube.
Integrated circuits: Integrated circuits are made from semiconductor materials and they contain millions of transistors. They are used in a wide variety of electronic devices, including amplifiers.
Amplifiers are used in a wide variety of electronic devices, including radios, televisions, and amplifiers. They are essential for increasing the amplitude of signals, which is necessary for many applications.