Before we start, it is not really necessary to know what goes on inside a transistor or diode to use it. All a service engineer or design engineer really needs to know about is what the external connections, are and what voltages/currents need to be applied to make it operate as we want.
However, a basic understanding about the inner workings does make the subsequent material easier to follow. I have tried to simplify the description below as much as possible, and have left out anything that is not necessary, so please do not regard it as a fully detailed description.
A diode consists of a piece of semiconductor material, generally silicon or germanium. This is divided into two distinct regions by adding impurities. One region has a surplus of electrons and is known as "N-type". The other region has an electron deficiency and is known as the "P-type". The spaces where electrons should be in the P-type are known as "holes". The division between the two regions is well defined. An electrical connection is made to each region.
It might be assumed that the surplus of electrons on one side would cross the divide and fill the holes on the other side. A few do, but the current resulting from this electron movement causes a small voltage difference between the N and P regions (known as "barrier potential"), which opposes further electron movement.
If a battery is connected to the junction, such that the positive terminal is connected to the P-region and the negative to the N-region as figure (a) above, the electrons in the N-type material will be attracted towards the positive side of the battery via the P-region. Similarly the holes in the P-type material will be attracted towards the battery negative via the N-type. If the voltage from the battery is sufficient to overcome the effect of the barrier potential, current will flow.
If the battery is connected the other way round as figure (b) above, the opposite effect occurs. The holes and electrons are attracted away from their relevant sides of the barrier. This has the effect of reinforcing the barrier, blocking the flow of current.
Thus a diode acts like a one-way valve, allowing current to flow from the P-region to the N-region, but not in the opposite direction. The P-region connection is known as the Anode and the N-region connection is the Cathode. These terms are the same as those used for a diode valve, and the device performs the same basic function.
The voltage required across the junction to cause a current to flow depends on the type of semiconductor material used. For germanium it is about 0.2V to 0.3V and for silicon it is about 0.6V to 0.7V.
A transistor is constructed in a similar manner to a diode, except there are three regions. Depending on the type of transistor there may be two N-type regions with a P-type region sandwiched between them (known as an NPN transistor), or two P-type regions with an N-type region between them (PNP transistor). For this description we will assume an NPN transistor, but the operation of a PNP transistor is basically the same if you reverse the polarity.
The two N-regions are known as the Collector and Emitter, and the central P-region is known as the Base.
If a battery is connected with the positive terminal to the Base and negative to the Emitter (B1 in fig (a) above), current will flow in the same manner as described for diode operation above.
Now assume another battery is connected with the positive to the Collector and the negative to the Emitter. This battery has a higher voltage than the one connected to the Base (B2 in fig (a) above). If you think in terms of the transistor containing two separate barriers and acting like two diodes, then the Collector-Base N-P barrier would act like a reverse biased diode, so no current would flow.
However there is some interaction between the two barriers due to an effect I have not mentioned yet. When a barrier is forward biased and passing current, there is a small region surrounding the barrier, where the surplus electrons from one side fill the holes on the other side. In this region there are no surplus electrons on one side and no surplus holes on the other side. This area is known as the "depletion region" because the numbers of surplus holes and electrons are depleted when compared to the no-current state.
In our transistor we have current flowing from Base to Emitter, so we have a depletion region around this junction. If we make the Base very small, this depletion region will take up the whole of the Base. In this state the Collector-Base junction is no longer a N-P junction, instead it is a N-nothing junction. It therefore no longer acts as a barrier to the Collector current, so current can flow from the Collector to the Emitter, via the Base, unimpeded.
If the battery connected to the Base (B1) is removed, the depletion region filling the Base is removed. We are then back to having a N-P Collector-Base barrier so no current flows from Collector to Emitter.
By varying the current flowing into the Base, we can vary how depleted the depletion region is, which in turn varies how much current can flow from Collector to Emitter. The effect is approximately linear over the area between no current and maximum current.
The Collector current is considerably larger than the Base current. Depending on the type of transistor, the collector current could be anything between ten times and several hundred times the Base current. By applying a small current variation to the Base we get a much larger Collector current variation. The device therefore acts as an amplifier.
Fig (b) above shows the same thing as fig (a) using the conventional transistor symbol.
The diagram and description above suggests that the Emitter and Collector are similar N-type regions, and it might be assumed that the connections could be reversed with no effect. In practice this is not the case. The transistor will not work correctly if the connections are swapped - it will have very low gain and may be damaged.
Other transistor types
The description above concentrated on the bipolar
transistor. There are a number of other devices that are also described as transistors,
including FETs, MOSFETs and unijunction transistors. However the only type of
transistor you are likely to encounter in a vintage transistor radio is the
bipolar transistor. References to "transistors" in this website section
(and anywhere else in connection with vintage radios) mean bipolar transistors.