For as long as I have been aware of such things, I've been fascinated by systems that feed back on themselves; or, to put it another way, govern their own action. I recall how delighted I was when, as a kid, I discovered how a flush toilet tank mechanism worked -- also, the home heating thermostat and the speed governor on a diesel engine. Later on, I learned about the beneficial effects of negative feedback in electronic systems. More leads to less, less leads to more -- I thought it was a great idea.
One specific example is really striking. That is a simple circuit composed of nothing more than two complementary junction field-effect transistors (JFETs) forming a two-terminal passive network (i.e., no internal power source) which exhibits the unusual property of negative resistance. The neat thing about a negative resistor is that it can, within certain limitations, be used to cancel out the resistive losses which are a part of any normal circuit. In particular, it can be used to make one of the simplest, widest ranging, most reliable oscillators you ever saw. I'll describe details in a moment.
This circuit has been around awhile and a schematic is shown here. Notice that it's made up of a P-channel and an N-channel JFET and nothing else. NO resistors, capacitors or inductors -- and, as mentioned above, no power source, either. It's known as the "lambda" circuit (or lambda diode) because its configuration resembles that of the Greek letter.
In order to understand how it works, we first have to look at the transfer characteristics of a typical junction FET. In other words, we need to see how the drain current, Id, behaves as the gate-to-source voltage , Vgs, changes.
Here's the transfer curve for a typical N-channel JFET. The curve for a P-channel JFET has a similar shape, but the polarity of Vgs and Id are opposite to those of the N-channel device. The major thing to notice is that the magnitude of Id is high when Vgs is near zero and falls off as Vgs gets more negative for a N-channel JFET (or more positive for a P-channel JFET). Putting it another way, the drain current of an N-channel JFET becomes less as the source voltage becomes more positive than the gate. Likewise, the drain current of a P-channel FET falls off as the source voltage becomes more negative than the gate.
This is all you need to know to understand the negative resistance action of the lambda diode circuit. To illustrate: Suppose you connect an external voltage source across a lambda diode circuit you have built, as shown here. Now, slowly increase the voltage, starting from zero volts, while monitoring the resulting current into the lambda diode on the milliammeter. Initally, of course, all voltages are zero, including Vgs of both transistors. But, since the external voltage is zero, no current can flow and Id is zero.
As the external voltage increases, current starts to flow through both JFETs. You can think of them as two variable resistors, forming a voltage divider, thus putting the source voltage of both devices, Vs, somewhere between the two terminal voltages. But this can't go on for long, since the gates are tied to the external voltage source and are being pulled in opposite directions, in such a way as to turn the JFETs OFF. This is the self-governing action I mentioned in the beginning of this article. It's also where the negative resistance comes in. Now, the more the external voltage increases, the LESS the JFETs conduct and current through the lambda diode decreases. Just the opposite to a normal, positive resistance!
And there you have it. You can build the simple circuit shown above, as I did, and if you plot the applied voltage against the resulting current (the V-I curve), you will see something like this. The negative resistance region is where the curve has a downward slope, and extends from roughly 3 to 7 volts. The value of the negative resistance is determined simply by measuring the slope of the curve and in this case, is about -0.5 mA per volt, or -2000 ohms.
Put the lambda diode in series with an L-C resonant circuit, so that the negative resistance cancels out the losses in the L-C network, and it will oscillate. This becomes a very useful device to have around the work bench/ham shack.
The oscillator circuit shown here
There's a lot more to know about negative resistance and the lambda diode, and they are fun to experiment with. In this article, I've only just scratched the surface of the subject. For a more detailed analysis as well as a complete schematic and construction guide for a dip meter, Click Here to visit the web site of Lloyd Butler, VK5BR. I have bult one along the lines he describes and I must say it works extremely well.
Another insightful paper on the lambda diode ("Exploring negative resistance: the lambda diode" by Samuel Dick) appeared in the January 1992 issue of Elektor Electronics magazine.