7 ms·
"especially on a circuit with already a lot of load on it" That's basically my point: sure, plugging something in can blow a fuse, but either the device is fau
by manarth 10y ago
"especially on a circuit with already a lot of load on it"
That's basically my point: sure, plugging something in can blow a fuse, but either the device is faulty, the circuit itself is faulty, or the circuit is already near capacity.
If it's a laptop, a docking station, speakers, multiple monitors, all plugged into a socket extension, plugged into a single outlet - even then, I wouldn't expect it to trip the fuse unless the circuit was already well on the way to overload.
If a circuit is so close to overload, it could be tripped simply by plugging in a phone charger, so the super-expensive mega-powerful laptop isn't really relevant.
- autokad 10y agoi see your point, a laptop or computer alone is not likely to trip a fuse, but when you got a 20 amp capacity and you plug in a 5 amp laptop system, thats meaningful. after buying a new alien ware, I tripped a fuse that wasn't tripped in years with a 'normal pc'. one could argue it wasn't the pc's fault, as there was a lot of stuff on that circuit, but one could also just as easily say the opposite.
- SamReidHughes 10y agoMaybe it was that Eurocom 12 core Xeon two power adapter having laptop.
- dsacco 10y agoFor what it's worth, I had to draw two dedicated electrical circuits up to my home office because my custom PC was simply too much load on its own, let alone the four monitors and assorted other equipment I have. Granted, my computer is about $10,000 of hardware with four GPUs; my point is that I'm not sure it's poetic license. You can definitely build a computer, even a custom laptop, that would at least strain typical circuits and maybe trip them with other peripherals attached.
- Veratyr 10y agoI dunno, a standard residential US circuit is max 15A at 120V, or 1800W. A GTX980 is 165W and the highest powered Xeon I could find was 165W. A Xeon + 4 GPUs + power supply loss comes to around 1050W and between the motherboard, RAM, SSDs, HDDs, monitors, speakers, cooling and such I'm sure you can push it high enough that it causes issues but I really can't see that coming from a laptop, even with peripherals.
- seanp2k2 10y agoIf you can plug in a hairdryer or space heater or vacuum cleaner, it'll run any single-mobo desktop I can think of without issue. Pretty much every US dwelling is set up to handle those on most of the outlets (1200-1800 watts typically). 2000 watts is a LOT. Laptop, no freaking way. Consider even just the heat dissipation.
- Nomentatus 10y agoThe odds of tripping a circuit-breaker with a phone charger are near zero; the odds of tripping a circuit-breaker or fuse with a vacuum or electric heater are excellent. I live with weak circuits, and only those two items that can be "the last straw," here. (OK, maybe an older, honking PC I used to own.) For a laptop to be in that sort of range is pretty darn impressive. Something that won't happen again, as CPU and GPU circuits and voltage are now shrinking.
- simmons 10y agoI don't know what's going on here, but my old Mac Pro would always trip the AFCI circuit breaker when I turned it on, even with little else on the circuit. It must have been taking a big gulp of power that resembled an arc-fault. Perhaps some laptop power supplies could also have weird characteristics that trip such a breaker even without exceeding the current threshold, and be interpreted in storytelling as "blowing out the fuses."
- yardie 10y agoThe GFCI is not a circuit breaker. It measures the difference between the current going in and the current going out over the hot and neutral wire. If it's faulting either the circuit is faulty (your Pc) or the GFCI is faulty. If the current isn't flowing back out through neutral it means it's going directly to ground or it's going through a human, to ground. The current threshold is 50mA
- simmons 10y agoThis wasn't a GFCI, but an AFCI. According to Wikipedia, AFCI's "detect electrical current alternating at characteristic frequencies, usually around 100 kHz, known to be associated with wire arcing, which are sustained for more than a few milliseconds." I suppose the Mac Pro's power supply produced some of these indicators, since all was well on a non-AFCI circuit. I've heard other people having similar false positives with equipment on AFCI-protected circuits.
- meta_AU 10y agoMany isolated flyback AC/DC converters run around the 100kHz mark, so probably on the right track there.
- seanp2k2 10y agoMy Airocide (air purifier) would trip the bathroom outlet on turn-on at a place I lived. Never figured out why, but I thought that the thing was broken when I first got it.