News for hackers. Stuff that matters.

A comment on DC Circuit Water Analogy (1998). Parent comment

Re:DC Circuit Water Analogy (1998) by tzs on Monday January 24, 2022 @05:31PM (#30064975)
> Anything involving the electrical energy not inside wires has no water analogue (ie the electromagnetic waves inside a transformer)

You don't even need fancy things like transformers for energy not in the wires to be important. Even a simple DC circuit consisting of a battery in series with a light bulb mostly involves an electromagnetic field to transfer energy from the battery to the light bulb, which mostly takes place outside the wires.

The function of the wires when it comes to energy transfer is to carry moving charge which creates the magnetic part of the electromagnetic field that actually carries the transferred energy.

Here's a pretty good explanation [1]. That video was from January 2019, and not controversial. Veritasium did a video on the topic in late 2021 that didn't really present things as well and ended up being quite controversial [2]. Other well-known YouTube channels such as EEVblog [3] and ElectroBOOM [4] responded.

There was also someone who bought a bunch of wire and did the experiment as described in the Veritasium video (although scaled down). That was discussed on HN and that discussion contains some very interesting links [5].

In particular the link to a talk by Rick Hartley in this subthread is very informative [6]. He talks about how most EMI problems in PCB designs are due to people not taking into account the the energy is not in the wires.

[1] https://www.youtube.com/watch?v=C7tQJ42nGno

[2] https://www.youtube.com/watch?v=bHIhgxav9LY

[3] https://www.youtube.com/watch?v=VQsoG45Y_00

[4] https://www.youtube.com/watch?v=iph500cPK28

[5] https://news.ycombinator.com/item?id=29598860

[6] https://news.ycombinator.com/item?id=29601273

[ Parent | on HN ]
Re:DC Circuit Water Analogy (1998) by gaze on Monday January 24, 2022 @07:09PM (#30066023)
The thing is that KVL and KCL hold, and KVL and KCL hold for both electrons and water. You just have to not think too deeply about what motivates the electrons to move. Current is indeed the mean velocity of carriers through some area, and voltage is mmm... well it's related to the electric field. It's just not exactly fluid pressure even though it behaves as such.

The right way to think about this if you _really_ care is a bit like the way that RF engineers think about dielectric resonators. The metal motivates the fields to be guided along pairs of conductors by the motion of charge carriers within. A metal will respond in a particular way to fields, which results in new fields, and so on and so forth in a way that results in the fields propagating along in a particular way.

A propagating pulse is a bit hard to think about but you can imagine some field configuration that results in the electrons in the metal slipping a bit, and those electrons move in a way that results in new fields which contain the new wavefront of the pulse. The electrons don't keep moving -- they move on average the distance of the mean free path and give their energy to the lattice as heat. This is why the drift velocity is just a meaningless number. The fields propagate at around c and the conductors guide them.

[ Parent | on HN ]