Monday, April 26, 2021

Transistors: UNIZOR.COM - Physics4Teens - Electromagnetism - Semiconduct...

New lecture is released to UNIZOR.COM.
It's about transistors and the principles of using semiconductors in amplifying an analogue signal or working as ON/OFF switch.
UNIZOR.COM - Physics 4 Teens - Electromagnetism - Semiconductors in Electronics - Junction Transistors

Notes to a video lecture on

Bipolar Junction Transistors

Below is a schematic representation of a so-called n-p-n bipolar junction transistor - one of the most popular types.

Functionally, it's playing a role similar to triodes discussed in one of the previous lectures. It can amplify a signal and it can work as On/Off switch.

To understand how it works, we will "build" it, step by step, gradually introducing new components.

As the first step in this process, imagine the device above without a narrow layer of p-type semiconductor in the middle and wiring connecting it to a battery (that is, without components marked in blue color).

Then this device would represent a solid n-type semiconductor built, let's assume, from silicon foundation with added atoms of phosphorus that have excess of one valence electron for each atom, which does not fit into a crystalline structure of silicon.

This n-type semiconductor is attached to a battery through two electrodes.
Let's assume that the difference in electric potential (voltage) between electrodes is sufficient to attract extra valence electrons of phosphorus (held only by electrostatic attraction to a nucleus, but not participating in covalent bonds) toward a positive electrode, but it's not strong enough to rip off other valence electrons participating in both the covalent bonds and electrostatic attraction to a nucleus, as electrons of silicon atoms are.

When such voltage is present between the electrodes, there is a small current going through a semiconductor, as the extra valence electrons from phosphorus atoms will be attracted to a positive electrode, rendering these phosphorus atoms positively charged, while the negative electrode will compensate the loss of electrons in the body of a semiconductor.
The current, obviously, depends on density of phosphorus atoms within silicon foundation and voltage applied.

The next step is to split the solid n-type semiconductor in two halves and put a layer of p-type semiconductor (let's assume its silicon with boron additive) in between two halves of n-type semiconductor (thin blue layer on the picture above represents this p-type semiconductor). Let's not connected this p-type semiconductor to a battery yet.

This will stop the current and here is why.
As before, the positive electrode will attract extra valence electrons of phosphorus atoms from the n-type semiconductor (the right side on the picture above).

The other side of n-type semiconductor, connected to a negative electrode, will have its extra valence electrons of phosphorus penetrating the n-p junction into p-type semiconductor. Both random moving of these free electrons and repelling off the negative electrode are the factors in this process. Crossed the n-p junction, they will be captured by holes in the covalent bonds of p-type semiconductor.

After some time there will be a saturation of the covalent bonds in the p-type semiconductor and the flow of electrons from the n-type to p-type stops, as the barrier of extra electrons in the p-type layer would repel electrons from the n-type connected to a negative electrode.

At this time the covalent bonds of the p-type layer will be filled, which renders it to be negatively charged. Extra electrons around boron nuclei will be held only by covalent bonds, but not attracted to a nucleus.
The n-type connected to a negative electrode will be neutral, as the electrons crossed to the p-type layer through n-p junction will be compensated from the negative electrode.
The n-type connected to a positive electrode will be positively charged, as its free valence electrons, not participating in covalent bonds, will be consumed by a positive electrode.

Now we will introduce the last component of this device - connect the p-type layer to a positive electrode of another battery and the negative electrode of this battery connect to the n-type semiconductor that already has a connection to a negative electrode of another battery.
Thus this n-type semiconductor will be connected to negative electrodes of both batteries.

What's important now is that the electrons, crossed into p-type layer through n-p junction, making it negatively charged and making a barrier on the migration of new electrons from the n-type semiconductor connected to a negative electrode, will be attracted by a positive electrode of a new battery. These electrons are held only by covalent bonds inside the p-type layer, not the electrostatic attraction to a nucleus. Positive electrode of a new battery will attract some of them from the place where they acted as a barrier at the n-p junction, the barrier weakens, and new electrons from the n-type semiconductor could penetrate the n-p junction barrier into p-type layer.

The process does not stop here. Since the barrier is weakened, the electrons from n-type semiconductor connected to a negative electrode continue migrating to p-type layer. That's why this n-type semiconductor is called emitter. Some of these electrons go further through p-n junction between a p-type layer and the other n-type semiconductor connected to a positive electrode, establishing an electric current between two initial electrodes. That's why that other n-type semiconductor is called collector. The semiconductor of p-type, making the layer between emitter and collector is called base.

By changing the voltages on two batteries involved and by changing the amounts of additives into emitter, collector and base we can control the current between emitter and collector.

Under some conditions the n-p-n bipolar junction transistor, whose principles of work are described above, can act as an amplifier of the signal between emitter and base into a stronger signal between emitter and collector.

Under some other conditions n-p-n bipolar junction transistor can act as the On/Off switch, opening or closing a circuit between emitter and collector by applying some voltage between emitter and base.

There are other ways to connect emitter, collector, base and batteries that we will not consider, as our purpose is to introduce a concept, rather than going into details of implementation. The development of contemporary transistors, their theoretical and technological aspects took a lot of efforts and time, so now we have a pretty advanced devices. But the principles of their work are still the same, those we demonstrated on the example presented above.

No comments: