Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

well, the truth turned out to be quite complicated; see also https://www.youtube.com/watch?v=2Vrhk5OjBP8 and https://www.youtube.com/watch?v=2AXv49dDQJw. the result is more complicated than just a single time delay, and the details of the experimental setup (like the wire diameter! and load impedance relative to transmission-line pulse impedance) matter a lot. in alphaphoenix's first experiment, putting 5 volts into a kilometer of wire, he got 0.2 volts across his resistor immediately, then after about 1.6μs, a jump up to about 2 volts, and then a gradual rise to 1.7 volts, some overshoot peaking after another 1.6μs, and then settling back down to the 1.7-volt level

(derek's results in the linked video, which incidentally links to one of the two i linked above, were quantitatively different but qualitatively similar)

the really unintuitive thing about this i think is that people think of electrical energy as flowing inside wires, when actually almost all of it flows around the wires, as veritasium explained quite ably. this is something people doing high-speed pcb layout have to deal with a lot in order to avoid emi problems

as i understand it, derek has a ph.d. in physics, or actually in physics education research https://youtube.fandom.com/wiki/Veritasium. that doesn't mean he knows everything about physics, but generally my experience with people in ph.d. programs is that they're good at listening to counterarguments and admitting when they're wrong, and also seeking out experts before publishing

(i think the veritasium video does contain a minor error in that it says electrons collide with metal ions, which as i understand it is not exactly how ohmic resistance works—the 'electrons' moving through the lattice are not exactly electrons but virtual particles similar to phonons or plasmons, and so the things they scatter off of are not individual ions—but possibly derek knows this and was intentionally simplifying, or possibly my understanding is wrong. i mean, i don't have a ph.d. in anything, much less solid-state quantum physics! derek certainly knows the electrons traveling through the wire aren't point-like particles bouncing around like billiard balls, and that the ions aren't red spheres with plus signs on them, despite depicting them that way.)



>i think the veritasium video does contain a minor error in that it says electrons collide with metal ions, which as i understand it is not exactly how ohmic resistance works—the 'electrons' moving through the lattice are not exactly electrons but virtual particles similar to phonons or plasmons,

electrons are moving through the wire. The virtual particle thing is a misunderstanding on your part I think.

I think your talking about current. Current flows in the opposite direction of electron drift velocity. In the simplified model they use these things called negative and positive charges flowing through the wire and current is defined as the movement of positive charge. These "charges" are of course virtual in nature because it's not what's actually happens.

What actually happens is negative charge is moving and positive charge (protons) are frozen.

And yes of course derek knows that it's a wave traveling through the lattice.


while of course that is correct, that's not what i'm talking about

there is a good introductory presentation in https://www.feynmanlectures.caltech.edu/III_13.html but note that it assumes some previous familiarity with quantum mechanics. §13.6 explains that mostly what the electrons (really propagating waves of quantum probability amplitude for there to be an extra electron) are scattering off of is imperfections in the lattice. but that can't be the whole story or all perfect crystals would be superconductors




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: