# UJT Negative Resistance and Ohm's Law

A Unijunction Transistor has a negative resistance region where voltage drops with an increase in current. This was explained to me as a situation where "Ohm's Law is broken". This doesn't sit well with me.

Can anyone explain how this negative resistance and inverse voltage-current relationship works in terms of Ohm's law?

• The Wiki page on the UJT has a nice description; UJT Device Operation. What more would you like? Or, what do you feel you need that is missing there? – jonk Feb 1 '18 at 5:48
• I think maybe I might just be misunderstanding. Is this 'resistance' value just a proxy for the interaction between voltage and current through the device? From what I can tell, at any given point R still equals V/I, so Ohm's Law is in fact being satisfied. – Trevor Feb 1 '18 at 14:14
• There are three perspectives: (1) A DC operating point perspective; (2) a dynamic perspective about the operating point; and (3) the overall curves that describe the totality of a complex behavior -- effectively, the integral of the derivative given a known starting point. (1) and (3) always behave as from Ohm's law. But (2) can appear to violate it. In short, whether or not the claim is true depends on assumptions. I don't get worried about it. I know KCL and KVL always work and their use depends on Ohm's law. So, that's how I see it. It always works. – jonk Feb 1 '18 at 16:57
• The UJT has a positive-feedback region. – analogsystemsrf Feb 2 '18 at 3:58

## 1 Answer

Ohm's law describes the proportionality of current to applied voltage in an ideal resistive material.

Even real passive parts do not behave in exact accordance (for example, a resistor may have a measurable voltage coefficient).

Things like diodes are very nonlinear, so Ohm's law does not apply (though for very small changes in a known current it may be useful to apply it in the form of differential or dynamic resistance). The differential resistance of most types of diodes is positive, with some exceptions.

Differential resistance is just $R_d=\frac{\Delta V}{\Delta I}$ for small changes about a bias point.

Negative (differential) resistance is similar- except the sign of the differential resistance is negative. It should be obvious that this is only possible for a region of applied voltages, otherwise we could extract unlimited energy from the device.