From a technical point of view, measuring the resistance of something is very simple. You put a voltage over the subject you want to measure the resistance of, measure the current flowing through it and apply Ohm’s law (R = U / I - resistance is voltage divided by current) to get the resistance. For our purposes it’s even easier: we don’t actually care about the real resistance values, just the changes to it over time. Since the voltage over a certain subject is dependant on its resistance (with an input voltage that remains the same), all we need to measure is voltage: when the resistance of our subject (the hand palm) goes down (sweaty palm), the voltage over it will go down too. So all we’d have to do is use a voltage meter and record the measurements over the time spanning the interrogation of our ’suspect’.

Now the technical part may be easy but not quite *that* easy. Because the resistance of skin is very big, the resulting voltages are very small. In order to get a meaningfull measurement we’ll have to amplify the signal. Fortunately, amplifiying signals can be done with a single electronic component: the transistor. Explaining the full workings of a transistor would lead too far here; I will refer interested reader to Wikipedia or an introductory book on electronics such as the classic ‘The Art of Electronics’ (by Horowitz and Hill).

So theoretically it could suffice to use a single transistor to amplify the signal enough to be measurable with simple equipment. Because of this low requirement, the lie detector circuit is a popular introductory circuit. A number of variations can be found, and most of the differences between them aren’t even in the part that does the actual amplification but more in the way the change in signal is visualized. Some circuits use leds (little lights that light up green or red, depending on whether the subject’s hand is surpassing a certain adjustable treshold of resistance), others use a speaker (sound with higher pitch as the resistance goes up), some use normal measuring equipment such as a multimeter for their readings.

Here are a number of circuits that can be found online and downloaded for free:

* The one-transistor design

This one is the simplest of them all. In its barest form it’s a transistor, two elektrode pads, a battery and a multimeter. The deluxe version has a led and a dial to calibrate the subject’s ‘normal’ skin resistance.

* With zero-calibration

This one is more or less the same as the ‘deluxe’ version above, with the addition of a capacitor that removes an alledged 50/60 hz ‘mains hum’, induced by the electrical grid.

* Lego version

Here’s one that is no more than a Lego programmable computer hooked up to some aluminum strips.

* LED version

This circuit has two leds: a green one that lights up when the detected resistance is below a certain treshold and a red one that lights up when its above the treshold. Red lights up = you’re lying. More components but the functioning is the same as the ones above.

Inspection of all of these circuits show that the ‘meat’ is always in one transistor, usually a general purpose low power npn one. For my tests, I will use the ’single transistor’ one with a fixed input voltage of 4.5 volts. I’ll have to add some specific elements to the circuit to let it safely work with my measuring equipment, but that’s for another post.