Unlike a resistor, the amount of current through a diode will depend upon 'which way round' we apply the voltage.
When the voltage is applied this way round it tends to pull the free electrons and holes apart, and
increases the height of the energy barrier between the two sides of the diode. As a result it is almost
impossible for any electrons or holes to cross the depletion zone and the diode current produced
is virtually zero. A few lucky electrons and holes may happen to pick up a lot of thermal (kinetic) energy.
This gives them enough 'go' to cross the barrier, hence the reversed biassed current is not zero, just very, very small.
When the voltage is applied this way round it tends to push the electrons and holes towards the junction. It also reduces the
height of the energy barrier and reduces the width of the depletion zone. These effects make it easier for free electrons and
holes with modest amounts of thermal (kinetic) energy to cross the junction. As a result, we get a sizeable current through the
diode when we apply a forward bias voltage.
If you look up diodes in a physics book you'll probably find an explanation which finishes up telling you that current through a
diode varies exponentially with the applied voltage. The shape of this exponential curve depends upon various factors which
include a 'fiddle factor' called the saturation current.
There are two problems with this result. One is that the equation is fairly complicated and quite difficult to use for analysing some
circuits. The second problem is that this equation is usually wrong! The reason for this is that the actual current/voltage
relationship depends upon the detail of how the diode was made - the choice of materials, doping, etc. Being simple souls who
like a quiet life, electronic engineers deal with these problems by simplifying things and using whichever of the following three
models of the diode suits them.
Vd represents the height of the diode barrier when no voltage is applied. This means that the energy required
for an electron (or hole) to be able to cross the barrier is eVd.
The 'square law' model assumes that the current, when forward biassed, is proportional to the square of the applied voltage.
The 'corner' model assumes that the current is zero for any voltage below Vd but rises when we try to apply
a voltage greater than this. In effect, the diode is viewed as a sort of 'switch' which is open when we apply low or negative voltages
but which closes when we try to apply a voltage equal to or greater than Vd. This means that it is impossible to
get a voltage larger than this across the diode. If we try to apply a larger voltage one of two things happen; either the voltage source
gives all the current it can manage and the diode current sits at this value, or the current is so high that the power warming the diode
(= voltage x current) blows it up! In this way of looking at things, Vd is called the corner voltage.
The 'one way' model simplifies things even more by assuming that the corner voltage is so small that we can regard it as being zero!
Although this is a drastic simplification, for many practical purposes, it is the only thing we need to remember about a diode. That is, it
behaves like a switch which is open (no current) when we apply volts on way and closed (as much current as we like) when we try to
apply volts the other way.
Whichever simple model we adopt, the basic diode maxim is therefore:
Forward bias - all current, almost no volts.
Reverse bias - all volts, almost no current.