Metal Oxide Voyeurism
Remember a few slides ago when I was wondering why the new model's protection had higher transient voltage ratings? Prying shrink-wrapped MOVs revealed those I could read this way to be GNR 20D201K, just like the old models, which were rated at 330V. Looks like the changes are simply a tweak in how they calculate and report ratings rather than component swaps.
MOV Ratings On The Line
Metal Oxide Varistors typically have four different voltage ratings:
- the nominal operating voltage below which the MOV does nothing
- the varistor voltage at which the MOV current reaches 1mADC – this is the rating that ends up in the model number: the 201 in 20D201K stands for 20x101 volts or about 200V
- the clamping voltage when the MOV is conducting its nominal rated current and repeatable pulse energy which the MOV is expected to be able to survive thousands of times if given a chance to cool down between pulses
- sometimes, the surge voltage when conducting its absolute maximum surge current which the MOV should be able to clamp but might not survive
In the GNR 20D201K's case, those would be 170VDC, 205V, 340V at 100A and 140 joules per pulse, and unspecified surge voltage at either 10kA one-shot or 6.5kA two-shots, respectively. This means connected equipment should not actually see much more than 340V unless the surge currents exceed 100A per MOV. Perhaps a schematic would make that easier to see.
The MOV Connection
If you had trouble imagining how those MOVs fit together based on my description over the last few slides, here is a simplified picture of what I meant. Orange circles highlight 20D201K MOVs sharing a thermal shut-off and the remaining two devices are unprotected 20D471K. The top two MOVs correspond to the pair shrink-tubed to the live-side thermal fuse, while the bottom four correspond to the two other heat-shrunk MOV pairs and their respective thermal shut-off. Crossing connections on the main MOVs lets them leverage the inductors to dampen transients and reduce the burden on individual devices.
Since the MOVs have a combined nominal rating of about 1.5kJ, other design considerations must be behind the packaging's 3kJ claim. Those likely include losses in the power cord, voltage drop across the inductors and “over-driving” the MOVs at the expense of potentially shorter lifespan. Since APC offers a lifetime warranty, the company is obviously not expecting many failures from this.
For comparison's sake, here is the simplified schematic of the old design. As you can see, six out of the old model's nine MOVs are located directly across mains lines where they have to eat the brunt of mains-side transients with no help from inductors to soften the blows.
The new design may have one less MOV than the old design, but they are being used in a more refined manner, which should enhance their effectiveness.
Less Is More
While the two models have nearly identical parts lists, a more clever use of MOVs in the new SurgeArrest should give it an advantage for surge protection. The millimeter-wide traces still bother me, though.
Aside from the loss of switched outlets and perhaps a regression in terms of aesthetics, I would have to call the new model a winner in terms of functionality. Its much less accident-prone switch is quite welcome, and the sliding outlet shutters are also less troublesome to deal with when you do not really need or want them. The boxier and flatter design is also less likely to cause fitting issues, and the more industrial look may appeal to fans of function over form.