5 ms·
Have the commenters here extolling the virtues of these "obliterators" actually built one? Because they don't do much. Younger me made one ~20 years ago (actual
by exmadscientist 4y ago
Have the commenters here extolling the virtues of these "obliterators" actually built one? Because they don't do much. Younger me made one ~20 years ago (actually around the time this page first came out!) and was very disappointed in how little it really did. Nothing happened to the hated DSL modem I wished destroyed! So what gives? It took me a while to figure it out, but these days I'm a professional EE and I have some insight....
Turns out the magnetics on an Ethernet port are good at their jobs. Like, really good. Not quite telecoms lightning-strike-survival good, but up there. Mains 60Hz may as well be DC as far as these things are concerned, and they're explicitly designed to block DC. Sure, for some configurations of magnetics you can blow them up if you get the pin connections just right, but that requires you to do a lot more than just "connect the transmit pins of the RJ-45 to HOT on 110VAC and the recieve pins to Common". Because that doesn't work. And even when it does work, it just overheats and pops the magnetics, opening them. Nothing happens inside the device proper.
If you want to cause some real fun you need a hi-pot tester plus a ground bonder. Those can do things that shouldn't happen.
- humanistbot 4y ago> If you want to cause some real fun you need a hi-pot tester plus a ground bonder. Those can do things that shouldn't happen. You posted this knowing someone would beg you to say more, right? What a tease. Fine. Say more.
- exmadscientist 4y agoI mean, it's 5,000 volts at 100mA for the "high current" hi-pot or 20,000 volts at 10mA for the "high voltage" one or 40A AC @ 8V AC for the ground bonder. They do about what you'd expect: components pop, things arc, things arc a lot at 20kV, things smoke and then melt if you shove 40A through them. It's more interesting when you break the tester itself. But then you have to fix the tester. These are not complicated machines (they're basically just transformers)... there's some complexity in regulating the outputs and more in making the measurements but most of the hard stuff is in having the tester itself survive the horrid abuse it's visiting on everything else around it.
- kevin_thibedeau 4y agoI lost two Ethernet ports a few months ago from a lightning strike on a neighboring house. The devices still work otherwise and other Ethernet ports connected to the switch were unharmed. They were both on a PoE switch which may have been partially at fault (though it still works too).
- jrockway 4y agoA friend and I made one of these in high school and plugged it into everything we could find because we were complete idiots. Literally nothing broke. My iMac Rev B's CRT did go crazy while the device was plugged into the Ethernet port, but everything was fine afterwards. Why would I try to break my only computer? I have no fucking clue. Kids are dumb.
- bluedino 4y agoI fried a device or two plugging a T1 into the ethernet ports. I didn't think they'd be protected from 120V
- gorkish 4y agoA friend of mine was having a lot of really really weird network issues at one of their location in another city (B2B supply company). Ports randomly going up and down on various equipment over the network. Appeared to be an electric issue, so we did various testing, ground bonding, etc. Added redundant equipment after that failed to be reliable. Having little else to offer in the way of suggestions, I made a joke that someone was using an etherkiller. Damned if I didn't end up being right. They had a clerk who was embezzling by creating false customer returns. Every time she made one she plugged the network into the mains. Her boyfriend apparently made this thing and told her that it would 'fuzz the system' or some other kind of idiotic idea. What it actually did was eventually start a small fire in the wiring closet. So, yeah the effect of one of these things is gonna be highly dependent on the environment. The galvanic isolation of Ethernet is not any sort of guarantee. Though I can confirm that the Cisco gear stood up to unreasonable abuse, it was unequivocally damaged.