# Difference between ohmic and non ohmic resistors

1. Mar 12, 2016

### netsky

I was reading about ohmic and non-ohmicresistors, but I'm unsure as to what thedifference between them is. I understand that ohmic devices have a constantresistance, meaning the voltage-current graph produced for it will show a straight line and that non-ohmic devices have a changing resistance, leading to a curved voltage-current graph.

What I don't understand is why this happens. Non-ohmic resistors have changingresistance because as the voltage increases, the electrons transfer more energy to the atoms of the conductor, meaning there are greater vibrations leading to an increase in the temperature and resistance. But why is this not the case with ohmic devices? Is there not any increase in the vibrations of atoms in ohmic devices? Everywhere I've read, I've found that ohmic devices have a "limited temperature range", but that doesn't make sense to me. Surely any increase in voltage will lead to an increase in temperature, so a limit on the temperature is not possible.

2. Mar 12, 2016

### Staff: Mentor

You are misinterpreting that. A "limited temperature range" means that the resistor is ohmic only within a certain temperature range. For a high enough temperature, any resistor will start behaving in a non-ohmic way.

3. Mar 12, 2016

### netsky

Ohhh, so basically every resistor is non-ohmic, but can be ohmic for a certain temperature range in which any change in temperature caused by a change in voltage, will not result in a change in resistance?

4. Mar 12, 2016

### davenn

will not result in a SIGNIFICANT change in resistance

keep in mind as well that it is the current flowing through the resistor, not the voltage across it, that is the primary cause of the change in temperature
( yes, it is a cause and effect situation, but it is the current that induces the heating)

Dave

5. Mar 12, 2016

### netsky

Alright, thank you both very much for the clarification!

6. Mar 12, 2016

### davenn

you are welcome

Remember resistors have a tolerance value, eg 1%, 5%, 10%, being the 3 common ones
This takes into account manufacturing process errors, material the resistor is made of (metal film are usually always 1% where as carbon can be 5% or 10%)
and I suspect, if I dig into resistor spec's deep enough, I would find that a temperature range is also taken into account

Dave

7. Mar 13, 2016

### LvW

Netsky: You always should be aware that in electronics NOTHING is ideal.
* Each resistor has a temperature dependence and parasitic properties (capacitive, inductive)
* Something similar applies to other parts: Capacitors and inductors have parasitic resistive properties.
* More than that, no formula (gain, input/output resistance,...) is correct by 100%. Each formula contains simplifications and is neglecting some minor influences.

Knowing this, it is one of the primary tasks of a good engineer to decide for which applications he can use simplified expressions as well as idealized parts properties.
This decision is to be made taking into account parts tolerances and application-specific accuracy requirements. In this context, the operating frtequency range plays a major role.
Always remember: Each design should not as good/exact as possible but "only" as good/exact as necessary.

8. Mar 16, 2016

### sophiecentaur

Why don't people actually read Ohm's Law? It only applies for a metal at constant temperature. Even a lamp filament will 'obey Ohm's Law' if you can manage to keep the filament at a constant temperature but, of course, its resistance varies over a 10:1 [Edit 1:10] range as it warms up from room temperature to 2500K.
It is sloppy to use the term "Ohmic" in all cases because people get the wrong idea. The resistivity of some metals does not vary a lot over a practical range of temperatures so that's what to choose when you need to rely on the resistance, marked on the component.
As stated earlier - no electronic component is idea. We are lucky that the manufacturers have managed to produced devices that are usually 'near enough' to treat them as ideal in most circuits. Early devices were nothing like 'near enough'.