Because an inductor in series, or a capacitor in parallel, will both work as a low-pass filter. They'll be OK for filtering out spikes when your target wave shape is flat, i.e. DC (and capacitors are used this way inside the computer all over, to filter the power going into various ICs: see "bypass capacitor"), but neither one will work very well filtering AC. Unless you had enough resistance there to get the proper RC time constant to make the normal shape of the output be a sine wave. But that resistance will hinder the circuit's ability to quickly react to surges, too, because of the higher time constant.
And an inductor in parallel, or a capacitor in series, won't filter the spike at all. (Well, hardly anyway.) Plus, neither of them have a fixed total-energy lifetime, while surge suppressors do. So that lifetime *must* be from some other component. (Also, using just a capacitor or just an inductor will really mess up the power factor. The power factor is 1 when total inductance == total capacitance (or at least, that's how I remember it; that could be wrong), and goes down from 1 when you add either without the other. (The power factor is the cosine of the phase angle, and is also equal to the ratio between resistance and impedance. The phase angle is modified by adding inductance or capacitance. A ratio of 1:1 (inductance:capacitance) gives a "pure resistor" type circuit, and a phase angle of zero, and a power factor of 1. Any other ratio (of R:Z) gives a phase angle that gets close to pi/4 or -pi/4, and a power factor that goes to about 0.7 (cos(pi/4), or sqrt(2)/2).)
A MOV, on the other hand, has a high impedance when the voltage across it is below its threshold, and a much lower impedance when the voltage goes above this threshold. So if you take the 120V (RMS) AC signal, figure that the peak-to-peak voltage is actually closer to 170V (multiply by sqrt(2), since it's a sine wave), then you get a MOV with a threshold somewhere around 170V, and just hook it up between the hot and ground lines. (And probably also neutral and ground.) When the voltage is above the threshold, it gets shunted to ground by the lower impedance in the MOV. (Of course the problem is that a MOV can only take a fixed total amount of energy before it becomes useless.)
Anyway, I don't think this sort of a setup changes the behavior of the circuit at all -- if you have multiple MOVs hooked up in parallel, that just means you can sink more total current. (Though I don't know what happens when you get more total current. They may take equal amounts, which would rather defeat the purpose of hooking them up like that. Hmm.)
(Unfortunately I have to get going. I'll be back later tonight, hopefully.) |
