When I was working in detector research in the Electronics Division of
Los Alamos National Lab, we specifically used non-ohmic to refer to
non-linear current-versus-voltage data caused by an interplay between
the voltage and the material. A good example of this would be the
junction of a silicon diode. The diode is almost non-conducting in the
reverse direction until we hit the breakdown voltage. In the forward
direction the device does not conduct until we arrive at a potential
difference of about 0.6 volts. At this point conduction begins, but it
does not rise linearly as the voltage rises.
The explanation for this is that there is a depletion zone formed (no
charge carriers) at the junction. The depletion zone prohibits
conduction through the junction. Reverse voltage only tends to widen
the depletion zone (further depleting it) preventing conduction.
Forward voltage reduces the depletion zone, and the depletion zone is
"gone" with a forward voltage of about 0.6 volts.
This description shows how the material is exhibiting gross property
changes as a result of voltage changes, i.e. electric field changes, in
the junction area. That is non-ohmic behavior. In our detector work we
needed non-ohmic behavior at some points in our detectors, but we needed
ohmic behavior at other points (such as the lead wires to the detector).
We had to follow special techniques to be sure to get ohmic contacts
with silicon and germanium and other semiconductor materials.
Now compare this to the filament of a tungsten bulb. Nothing remotely
similar to a diode is taking place. The filament is simply heating up
and the resistivity is changing because of a temperature effect, not
because of a voltage effect. Therefore the tungsten filament is ohmic
even though it shows non-linearity in the current-versus-voltage graph.
The non-linearity is not caused by the voltage. If we keep the
temperature constant, the current-versus-voltage plot is linear.
In my sophomore physics class I have the students do an experiment with
a 1/4-watt carbon-film resistor. They plot current versus voltage on a
300-ohm resistor up to 20 V across the device. At this point the device
is dissipating about 1.3 watts (five times its rated value) and it is so
hot it begins to smell because the epoxy paint is beginning to burn.
The plot students get has a slight bit of "upward" curvature. (The
curvature is slight and students might not notice it if they don't lay a
ruler along the points or view it with their line-of-sight parallel to
the paper.
Then the students use the same resistor with a heat sink. This time the
current-versus-voltage plot is very linear.
I say this to the students... The carbon-film resistor is an ohmic
device. However, it can produce a non-linear current-versus-voltage
plot if we do not control the temperature. That doesn't mean the device
is non-ohmic, it just means that when we observe non-linear behavior it
is the failure to regulate the temperature that caused the non-linear
behavior rather than the increased voltage that caused the non-linear
behavior. If we control the temperature we can see the device is ohmic.
We can also see that the current deviation with temperature is more
current at higher temperature. This demonstrates the temperature
coefficient for the resistivity is negative. That helps us realize this
is a carbon resistor as opposed to a nichrome resistor because carbon
has a negative temperature coefficient for resistivity and nichrome has
a positive temperature coefficient.
The tungsten bulb is like the resistor, except its temperature change
during operation is considerably larger.
Michael D. Edmiston, Ph.D.
Professor of Chemistry and Physics
Bluffton College
Bluffton, OH 45817
(419)-358-3270
edmiston@bluffton.edu