 |
First, lets consider the dynamic behavior of a simple
pn-junction. |
|
 | Whereas the small signal behavior
at low frequencies is exactly that of the example given
before, the large signal response in practice is simply
what we (and everybody else) calls the switching
behavior of the diode. In other words, we are looking at either - Suddenly switching
the voltage from 0 V to some forward value Uf , or
- Suddenly
switching the current from 0 A to some constant forward value If
.
|
 | In order to
have some idea of what we want to find out, here are some typical results of experiments with
modulated inputs. First, we look at the small signal behavior
by plotting the amplitude of the current output versus
the amplitude of the voltage input as a function of the
frequency. What we will find looks like this: |
|
| | |
|
|  | While
real curves may look quite different, we always will find that with increasing frequency the
output amplitude will eventually come down or go up. Often, the frequency where the output
value is 3 db below or above its low frequency value is identified as some limit frequency
fl, giving us a time constant t = 1/f
l which we want to understand in terms of materials properties. |
 |
Now we look at the switching
behavior for the voltage, i.e. we suddenly switch off the voltage across the diode
from 0 V (or some reverse voltage) to some forward voltage, e.g. 0,7 V. |
|  |
This is experimentally easy to do and to measure. What we obtain looks like
this: |
| |
|
|
 |
We have pronounced current "transients" describable by one (or possibly more) time constants;
and we want to know how these time constants relate to materials properties of the pn-junction. |
 |
Next, we switch
the current "suddenly". The question is how to do this. In formal electrical
terms, we just take a "constant current source" and turn it on. In reality this
might be a voltage source with a resistor RCS much larger than anything
else in the circuit. The current than is simply U/ RCS and switching is
obtained by switching the voltage which now could be rather large. |
|  |
If we do the experiment, we will observe the following
behavior: |
| |
|
|
 |
Again, we have transients
in the diode behavior, this time the voltage is affected. But we also have a kind of
transients in the current - we will not be able to switch it off abruptly, but for some time
a reverse current flow will be observed |
|  |
Since the current pulse was "made" by a (rectangular) voltage pulse and the external
voltage is zero after the switching, we are forced to conclude that the diode acts for some
time as a voltage/current source after the primary voltage
was turned off. More details can be found
in an advanced module |
 | Once
more we measure some time constants and we must ask ourselves how they relate to the time
constants of the voltage switching experiment and to material parameters. |
 |
Accepting all these "experimental" findings,
we even must conclude that there might be several time
constants which, moreover, might depend on the particulars of the experiment, e.g. the working
point, or the amplitudes of the input signal. |
|
 |
This does not just look
a bit involved; it really is. And the "ideal"
pn-junction diode is about the most simple device we have. |
 |
We are going to look at it in some detail in the
next subchapters, even so we are not particuarly interets in plain pn-junctions. But
this will help to develop some basic understanding of the underlying processes and make the
dicussion of more involved devices easier. |
© H. Föll (Semiconductors - Script)