>>Chris wells wrote: ...isn't there some value in preventing the nasty flash over of a MOV failure?
A positive Temperature Coefficient Thermistor (PTCT) will prevent catastrophic failure of a circuit when a MOV or other component shorts. However, it is often unable to protect the MOV from overheating. The following analysis explains why. A MOV is responsive to voltage, temperature, and history in complex ways, and has a fairly high dynamic impedance. Likewise, a PTC thermistor has complex properties. But for this evaluation, we+ll work with idealized components. Once the principle is understood, you can plug real values into a SPICE simulation and get similar results. Assume a series circuit containing an ideal MOV and an ideal PTC thermistor, as defined below: MOV: threshold voltage of 500V, max dissipation of 5W or 50 joules (watt-seconds). Then for any voltage below 500V, current and dissipation are zero. For I > 0, dissipation = P = I x E = 500I . Total energy dissipated = Integral 500I(dt). PTC thermistor: threshold dissipation of 1 watt or 10 joules. Below threshold, resistance is small, say, 10 ohms. Once threshold is reached, resistance changes to maintain 1W dissipation. Total transient dissipation up the threshold = 10J = Integral (I)(I)(10 ohms)dt. Transient Analysis: In this example, in order for the thermistor to switch into it+s high impedance state before the MOV overheats, it must receive 10 J before the MOV receives 50J: the dissipation RATE of the thermistor must be at least 1/5 that of the MOV. Assuming a square wave of amplitude Vmax (V constant) and differentiating both sides, 5(I)(I)(10 ohms)=500(I), so I=10A. Vmax = Vmov + Vptct = 500V + ( 10A)(10 ohms) = 600V. (It turns out t = 10 ms for this case.) Conclusion: For any voltage between 500V and 600V the thermistor cannot save the MOV! Pick any real MOV and any real PTC thermistor. Unless your MOV is huge and your thermistor very tiny, you+ll always be able to find a voltage just above the MOV voltage where the MOV will be destroyed before the thermistor heats. >>Chris Wells wrote: Lastly - what do you know about the acceptance of MOVs in Europe and in that light how would you deal with Surge? I have heard some contentions in particular when placed line to ground. What would you do: >>1) Use robust front end filter and limiting impedances (resistors, caps and inductors) with no MOVs? >>2) Put Gas tubes in series with MOVs. (this sounds like a mess) >>3) Fuse each MOV. line to line and line to ground. This way the fuses could prevent a short failure from becoming dangerous and then the fuse could be removed for specific applications that do not allow MOVs to ground. 1.) When designing with MOVs, I assume they+ll fail shorted. I use them to limit circuit damage, but not for safety compliance. They must be inside some protection. 2.) I+ve used MOVs in series with spark gaps, line to line. The MOVs were rated at above the line voltage, allowing the spark gap to quench before the zero crossing. The spark gap was too. If the MOV shorted, transients would latch the spark gap on, ocasionally blowing the mains fuse. Not an elegant solution, but it worked. 3.) I have not used MOVs to ground, can't say.
