3 ms·

Comment #16 at that eevblog link is supposedly a direct quote from the manufacturer, and it’s pretty damning. Most panelboard manufacturers make a device that

by amluto 3y ago

Comment #16 at that eevblog link is supposedly a direct quote from the manufacturer, and it’s pretty damning. Most panelboard manufacturers make a device that can plug in to your main electrical panel just like a breaker and is probably much more effective due to being directly attached to the bus. (Not necessarily against EMPs. But it should do pretty well against a conducted surge due to nearby lightning or a utility switching transient. It won’t do very well if used in a non-main-panel, because they generally lack separate neutral and ground connections — a good surge protector anywhere other than a “service entrance” will separately protect line-line, line-neutral (2x), line-ground (2x), and neutral-ground, for a total of six individual transient voltage suppression devices.)

mindslight 3y ago
TVS-based surge suppressors seem like wishful thinking to me, especially given MOVs' finite lifespan. Series mode surge protection is the better approach. And regardless of the protection type, star power/ground and absolutely no wires going from any of the protected devices to other unprotected devices.
- amluto 3y ago
  Series mode seems to mean Zero Surge or maybe one of their rather sketchy-looking competitors. If these devices were so great, I would expect to see a real standard and certification for them. And this seems a bit dubious to me: > Unlike most surge protectors, Zero Surge products do not rely on the ground circuit for effective surge protection, so you can use them even in ungrounded outlets. So let’s see… they apparently pass ground straight through (as they should for safety!), and they’re some sort of filter on hot and/or neutral. (I found a sketch online suggesting that they are, in fact, just an inductor on the hot wire.). So suppose a surge happens and, by some miracle, the surge is just a voltage spike on hot relative to everything else, then, into a resistive load, current and hence voltage will be limited. But the load might not he resistive! Into a switched-off MOSFET or similar, the current needed to damage something is really quite low, and the only way this seems like it will work is if there’s enough capacitance somewhere to make the filter work. If there is any filtering on neutral, though, or if ground is contaminated, then I could imagine arcing between the protected circuit and ground. In any case, this all seems silly. If the service entrance is protected, a hot-only surge seems pretty unlikely. (Your house’s neutral and ground wiring is not that great, and surges can couple inductively from nearby lightning strikes or even result from failures inside the house.) I’ll stick with conventional surge protectors. (I do own a Zero Surge that I bought to test as an EMI filter. I had such bad EMI on one circuit that it made a computer buzz very audibly — this was caused by cheap LEDs dimmed by a mediocre TRIAC. The result: the Zero Surge buzzed even louder than the power supply!)
  - mindslight 3y ago
    I've got a few SurgeX units that I've taken apart. It's an inductor in line with the hot, and sense circuitry that shuts off a downstream relay when the input voltage rises. I don't remember sizing up the capacitance but surely there has to be some, and presumably they've done the math so that the voltage rise downstream of the inductor is slower than the relay switching speed. > a hot-only surge seems pretty unlikely What do you mean? N/G should be bonded and at the same voltage with exterior surges, assuming a dedicated circuit. Across L-N/G is the only place a surge can show up - barring parasitic coupling of the equipment to ground/environment, or erroneously connecting a wire to unprotected equipment. And compared to TVS ones that pass all the lines through, relying on their ability to shunt overvoltage/current, with the active element actually dissipating much of the surge power? The series mode of operation at least seems analytically tractable.
blincoln 3y ago
Yes, as I was reading their various descriptions of the device, I started to think "wait, is this just a whole-house surge protector?" ...and sure enough, they come right out and state "The EMP Shield is one of the World’s fastest whole home surge protectors" right on the main page of their site. Call me a pawn of big business, but I'd much rather pay $200 or less for a whole house surge protector made by a respected electrical equipment manufacturer than almost $400 for one made by a company that ladles grandiose claims about nuclear weapons over their product. I'm at a loss how they managed to convince anyone into thinking they were going to employ 1000 people, or do anything related to chip manufacturing.

by amluto 3y ago

mindslight 3y ago

TVS-based surge suppressors seem like wishful thinking to me, especially given MOVs' finite lifespan. Series mode surge protection is the better approach. And regardless of the protection type, star power/ground and absolutely no wires going from any of the protected devices to other unprotected devices.

amluto 3y ago
Series mode seems to mean Zero Surge or maybe one of their rather sketchy-looking competitors. If these devices were so great, I would expect to see a real standard and certification for them. And this seems a bit dubious to me: > Unlike most surge protectors, Zero Surge products do not rely on the ground circuit for effective surge protection, so you can use them even in ungrounded outlets. So let’s see… they apparently pass ground straight through (as they should for safety!), and they’re some sort of filter on hot and/or neutral. (I found a sketch online suggesting that they are, in fact, just an inductor on the hot wire.). So suppose a surge happens and, by some miracle, the surge is just a voltage spike on hot relative to everything else, then, into a resistive load, current and hence voltage will be limited. But the load might not he resistive! Into a switched-off MOSFET or similar, the current needed to damage something is really quite low, and the only way this seems like it will work is if there’s enough capacitance somewhere to make the filter work. If there is any filtering on neutral, though, or if ground is contaminated, then I could imagine arcing between the protected circuit and ground. In any case, this all seems silly. If the service entrance is protected, a hot-only surge seems pretty unlikely. (Your house’s neutral and ground wiring is not that great, and surges can couple inductively from nearby lightning strikes or even result from failures inside the house.) I’ll stick with conventional surge protectors. (I do own a Zero Surge that I bought to test as an EMI filter. I had such bad EMI on one circuit that it made a computer buzz very audibly — this was caused by cheap LEDs dimmed by a mediocre TRIAC. The result: the Zero Surge buzzed even louder than the power supply!)
- mindslight 3y ago
  I've got a few SurgeX units that I've taken apart. It's an inductor in line with the hot, and sense circuitry that shuts off a downstream relay when the input voltage rises. I don't remember sizing up the capacitance but surely there has to be some, and presumably they've done the math so that the voltage rise downstream of the inductor is slower than the relay switching speed. > a hot-only surge seems pretty unlikely What do you mean? N/G should be bonded and at the same voltage with exterior surges, assuming a dedicated circuit. Across L-N/G is the only place a surge can show up - barring parasitic coupling of the equipment to ground/environment, or erroneously connecting a wire to unprotected equipment. And compared to TVS ones that pass all the lines through, relying on their ability to shunt overvoltage/current, with the active element actually dissipating much of the surge power? The series mode of operation at least seems analytically tractable.

blincoln 3y ago

Yes, as I was reading their various descriptions of the device, I started to think "wait, is this just a whole-house surge protector?" ...and sure enough, they come right out and state "The EMP Shield is one of the World’s fastest whole home surge protectors" right on the main page of their site. Call me a pawn of big business, but I'd much rather pay $200 or less for a whole house surge protector made by a respected electrical equipment manufacturer than almost $400 for one made by a company that ladles grandiose claims about nuclear weapons over their product. I'm at a loss how they managed to convince anyone into thinking they were going to employ 1000 people, or do anything related to chip manufacturing.