Watch electricity hit a fork in the road at half a billion fps
December 6, 2023 9:04 AM Subscribe
fluid dynamics!
That's what I was thinking too, and it makes sense since individual electrons are mostly just talking to the atom next door, like molecules bouncing off each other in a liquid.
I didn't watch the whole thing, maybe they mention it, but I'm wondering if you can somehow get a standing wave going in these walled off circuits to behave as an oscillator, potentially a very high frequency one.
posted by AzraelBrown at 9:48 AM on December 6, 2023 [1 favorite]
That's what I was thinking too, and it makes sense since individual electrons are mostly just talking to the atom next door, like molecules bouncing off each other in a liquid.
I didn't watch the whole thing, maybe they mention it, but I'm wondering if you can somehow get a standing wave going in these walled off circuits to behave as an oscillator, potentially a very high frequency one.
posted by AzraelBrown at 9:48 AM on December 6, 2023 [1 favorite]
I didn't watch the whole thing, maybe they mention it, but I'm wondering if you can somehow get a standing wave going in these walled off circuits to behave as an oscillator, potentially a very high frequency one.
Starting around 10:45 of the video, he has an animation based on the data he collected. It looks like the oscillations in the open circuit damp down pretty quickly - in about 2000 nanoseconds - though he doesn't say what's doing the damping or where the energy is going.
However... he did make the wires in a way that slowed the electricity down somehow by using a twisted pair... so maybe that's where the energy is going? and maybe if it was a lower... whatever? ...wire you'd be more likely to get a standing wave? I have no idea.
(Would it have anything to do with the way that antennas can be tuned by varying their length?)
posted by clawsoon at 10:05 AM on December 6, 2023 [2 favorites]
Starting around 10:45 of the video, he has an animation based on the data he collected. It looks like the oscillations in the open circuit damp down pretty quickly - in about 2000 nanoseconds - though he doesn't say what's doing the damping or where the energy is going.
However... he did make the wires in a way that slowed the electricity down somehow by using a twisted pair... so maybe that's where the energy is going? and maybe if it was a lower... whatever? ...wire you'd be more likely to get a standing wave? I have no idea.
(Would it have anything to do with the way that antennas can be tuned by varying their length?)
posted by clawsoon at 10:05 AM on December 6, 2023 [2 favorites]
Yay, AlphaPhoenix! Now go watch his "Elections made to order" video! It will shake your belief in... wait, who am I kidding. It will confirm your worst fears.
posted by tigrrrlily at 10:34 AM on December 6, 2023 [3 favorites]
posted by tigrrrlily at 10:34 AM on December 6, 2023 [3 favorites]
There's no such thing as an open circuit, because that would be an oxymoron. This fact does not change simply because somebody who by his own admission started with no understanding of transmission lines chooses to ignore distributed capacitance and inductance and reason as if Ohm's Law were the be-all and end-all of electrical theory.
I'm wondering if you can somehow get a standing wave going in these walled off circuits to behave as an oscillator, potentially a very high frequency one
Welcome to antenna tuning.
posted by flabdablet at 11:15 AM on December 6, 2023 [6 favorites]
I'm wondering if you can somehow get a standing wave going in these walled off circuits to behave as an oscillator, potentially a very high frequency one
Welcome to antenna tuning.
posted by flabdablet at 11:15 AM on December 6, 2023 [6 favorites]
Note, however, that you’d have to keep adding energy to the system to maintain a standing wave, due to energy loss from resistance/impedance (like friction). Without energy being added, the system would tend (fairly quickly, as I understand it? But I have little hands-on experience with this) toward electrostatic equilibrium (no movement of electrons).
posted by eviemath at 11:44 AM on December 6, 2023 [1 favorite]
posted by eviemath at 11:44 AM on December 6, 2023 [1 favorite]
Everything I learn about how electricity works makes me want to learn more about it. This was a really cool video - Well presented but not over-produced, crunchy enough that I feel like I learned something beyond the flashy demonstration but flashy enough that the demonstration itself was compelling. I like this guy's style.
posted by Phobos the Space Potato at 11:45 AM on December 6, 2023 [2 favorites]
posted by Phobos the Space Potato at 11:45 AM on December 6, 2023 [2 favorites]
There's no such thing as an open circuit, because that would be an oxymoron.
It's a perfectly cromulent word. A Google search for 'site:ieee.org "open circuit"' (with verbatim results on) returns thousands of hits.
A Google scholar search for '"open circuit" "electrical engineering"' returns about 141,000 results.
A Google nGram search shows the term has stayed essentially equally frequent since the mid-1890s. Its overall trend follows the same trend as "electrical engineering".
And in any case it's a very useful term. If a circuit is broken, calling it an open circuit is much more meaningful than calling it, e.g. a "non-circuit" because almost everything in the universe is a non-circuit, whereas "open circuit" implies that is a proper circuit but for some break, often one that is readily capable of being closed. You might prefer the term "incomplete circuit" but another nGram search shows that is vastly less common than "open circuit", so I'm afraid you're fighting an uphill linguistic battle here.
posted by jedicus at 11:58 AM on December 6, 2023 [14 favorites]
It's a perfectly cromulent word. A Google search for 'site:ieee.org "open circuit"' (with verbatim results on) returns thousands of hits.
A Google scholar search for '"open circuit" "electrical engineering"' returns about 141,000 results.
A Google nGram search shows the term has stayed essentially equally frequent since the mid-1890s. Its overall trend follows the same trend as "electrical engineering".
And in any case it's a very useful term. If a circuit is broken, calling it an open circuit is much more meaningful than calling it, e.g. a "non-circuit" because almost everything in the universe is a non-circuit, whereas "open circuit" implies that is a proper circuit but for some break, often one that is readily capable of being closed. You might prefer the term "incomplete circuit" but another nGram search shows that is vastly less common than "open circuit", so I'm afraid you're fighting an uphill linguistic battle here.
posted by jedicus at 11:58 AM on December 6, 2023 [14 favorites]
Conventional wisdom about electricity is that unlike water, electricity does not follow the path of least resistance.
Instead, it follows all paths in inverse proportion to their resistance.
I’m not sure how far we can push that rubric when we're dealing with extremely high resistance and very low currents, since among other things, electrons are not infinitely divisible.
posted by jamjam at 12:09 PM on December 6, 2023
Instead, it follows all paths in inverse proportion to their resistance.
I’m not sure how far we can push that rubric when we're dealing with extremely high resistance and very low currents, since among other things, electrons are not infinitely divisible.
posted by jamjam at 12:09 PM on December 6, 2023
And we don’t necessarily understand current flow as well as we thought we did, anyway.
posted by jamjam at 12:21 PM on December 6, 2023
posted by jamjam at 12:21 PM on December 6, 2023
Conventional wisdom about electricity is that unlike water, electricity does not follow the path of least resistance.
What? In both situations, "current" flows whenever there is a "pressure" difference - actual fluid pressure difference for fluids, voltage difference for electricity - and a path to flow on. Current continues to flow til the difference is eliminated.
Homework assignment: resistance vs reactance
Our frame of reference dictates how closely we look at the details. If we're concerned about RF generation, transients causing interference, momentary peaks that overload or damage components like switches or semiconductors, and other EE minutae, then this video's level of analysis is useful. If we just need to connect up a light through a switch, it's not very useful.
I don't need to grasp the general theory of relativity to load rocks onto a pickup.
posted by Artful Codger at 12:44 PM on December 6, 2023 [1 favorite]
What? In both situations, "current" flows whenever there is a "pressure" difference - actual fluid pressure difference for fluids, voltage difference for electricity - and a path to flow on. Current continues to flow til the difference is eliminated.
Homework assignment: resistance vs reactance
Our frame of reference dictates how closely we look at the details. If we're concerned about RF generation, transients causing interference, momentary peaks that overload or damage components like switches or semiconductors, and other EE minutae, then this video's level of analysis is useful. If we just need to connect up a light through a switch, it's not very useful.
I don't need to grasp the general theory of relativity to load rocks onto a pickup.
posted by Artful Codger at 12:44 PM on December 6, 2023 [1 favorite]
you’d have to keep adding energy to the system to maintain a standing wave
The first electronics project I ever built as a kid (as opposed to the first electrical project) was a crystal set.
This is a receiver for AM radio, and it has just four components: a capacitor and an inductor that together form a resonant "tank" circuit, a diode to recover the modulating audio-frequency signal from the radio-frequency carrier resonating in the tank, and a high sensitivity earpiece. Ground one end of the tank, connect the other to a loose bit of wire to function as an antenna, and now we're listening to parentally unapproved radio stations under the blankets at night.
The capacitor is a variable type, allowing tank circuit to be tuned to resonate at the carrier frequency of the transmitter whose signal you're trying to pick up. This does a good enough job of being stimulated only by one particular carrier frequency out of the mish-mash delivered by the antenna to deliver a surprisingly listenable result. The really nifty thing is that all of the energy sloshing around in the tank, some of which which ultimately manifests as sound coming out of the earpiece, is delivered by the radio waves picked up by the antenna; the circuit has no other energy supply.
you're fighting an uphill linguistic battle here
It's my new hobby.
electrons are not infinitely divisible
They don't have to be for the purposes of creating the kinds of phenomena to which Ohm's Law can usefully be applied, because although the electron does indeed carry an indivisible quantum of charge, in any conductor the charge carriers will be in travelling-wave states and therefore won't have precise locations.
posted by flabdablet at 1:02 PM on December 6, 2023 [4 favorites]
The first electronics project I ever built as a kid (as opposed to the first electrical project) was a crystal set.
This is a receiver for AM radio, and it has just four components: a capacitor and an inductor that together form a resonant "tank" circuit, a diode to recover the modulating audio-frequency signal from the radio-frequency carrier resonating in the tank, and a high sensitivity earpiece. Ground one end of the tank, connect the other to a loose bit of wire to function as an antenna, and now we're listening to parentally unapproved radio stations under the blankets at night.
The capacitor is a variable type, allowing tank circuit to be tuned to resonate at the carrier frequency of the transmitter whose signal you're trying to pick up. This does a good enough job of being stimulated only by one particular carrier frequency out of the mish-mash delivered by the antenna to deliver a surprisingly listenable result. The really nifty thing is that all of the energy sloshing around in the tank, some of which which ultimately manifests as sound coming out of the earpiece, is delivered by the radio waves picked up by the antenna; the circuit has no other energy supply.
you're fighting an uphill linguistic battle here
It's my new hobby.
electrons are not infinitely divisible
They don't have to be for the purposes of creating the kinds of phenomena to which Ohm's Law can usefully be applied, because although the electron does indeed carry an indivisible quantum of charge, in any conductor the charge carriers will be in travelling-wave states and therefore won't have precise locations.
posted by flabdablet at 1:02 PM on December 6, 2023 [4 favorites]
flabdablet: the circuit has no other energy supply.
But, for clarity, the radio waves are continuously supplying additional energy, correct?
They don't have to be for the purposes of creating the kinds of phenomena to which Ohm's Law can usefully be applied, because although the electron does indeed carry an indivisible quantum of charge, in any conductor the charge carriers will be in travelling-wave states and therefore won't have precise locations.
Do I remember correctly from the video that the electrons are travelling a trillion times slower than the electricity, or something like that?
posted by clawsoon at 1:07 PM on December 6, 2023
But, for clarity, the radio waves are continuously supplying additional energy, correct?
They don't have to be for the purposes of creating the kinds of phenomena to which Ohm's Law can usefully be applied, because although the electron does indeed carry an indivisible quantum of charge, in any conductor the charge carriers will be in travelling-wave states and therefore won't have precise locations.
Do I remember correctly from the video that the electrons are travelling a trillion times slower than the electricity, or something like that?
posted by clawsoon at 1:07 PM on December 6, 2023
Artful Codger: I don't need to grasp the general theory of relativity to load rocks onto a pickup.
Although it would help if you wanted to load those rocks verrrrrrrrrrry quickly.
posted by clawsoon at 1:09 PM on December 6, 2023 [4 favorites]
Although it would help if you wanted to load those rocks verrrrrrrrrrry quickly.
posted by clawsoon at 1:09 PM on December 6, 2023 [4 favorites]
oh snap.
(now back to reminiscing about crystal radios)
posted by Artful Codger at 1:20 PM on December 6, 2023
(now back to reminiscing about crystal radios)
posted by Artful Codger at 1:20 PM on December 6, 2023
the radio waves are continuously supplying additional energy, correct?
Correct.
Do I remember correctly from the video that the electrons are travelling a trillion times slower than the electricity
Yes.
The electricity-as-water and wires-as-plumbing analogy is useful to get a basic understanding of what's going on, but taking it too seriously rapidly becomes misleading.
Electrical systems are all about transferring energy from an electrical source to an electrical sink; plumbing is more about transferring bulk water.
Thinking of the speed of propagation of an electrical signal down a transmission line as loosely analogous to the speed of sound in a pipe full of water is a good first step on the way to disentangling the two use cases and refining the idea of electric current away from some vague notion of bulk electron delivery.
It's all about the energy, not so much about the charge. A lightning bolt that blasts a tree trunk to splinters does so while transferring maybe a thirtieth of the charge that a typical smartphone battery can hold.
posted by flabdablet at 1:33 PM on December 6, 2023 [1 favorite]
Correct.
Do I remember correctly from the video that the electrons are travelling a trillion times slower than the electricity
Yes.
The electricity-as-water and wires-as-plumbing analogy is useful to get a basic understanding of what's going on, but taking it too seriously rapidly becomes misleading.
Electrical systems are all about transferring energy from an electrical source to an electrical sink; plumbing is more about transferring bulk water.
Thinking of the speed of propagation of an electrical signal down a transmission line as loosely analogous to the speed of sound in a pipe full of water is a good first step on the way to disentangling the two use cases and refining the idea of electric current away from some vague notion of bulk electron delivery.
It's all about the energy, not so much about the charge. A lightning bolt that blasts a tree trunk to splinters does so while transferring maybe a thirtieth of the charge that a typical smartphone battery can hold.
posted by flabdablet at 1:33 PM on December 6, 2023 [1 favorite]
I don't need to grasp the general theory of relativity to load rocks onto a pickup.
Perhaps, but you do rely on QM effects. Things have sharp edges, rather than spongily mushing into each other because of QM.
posted by bonehead at 2:37 PM on December 6, 2023
Perhaps, but you do rely on QM effects. Things have sharp edges, rather than spongily mushing into each other because of QM.
posted by bonehead at 2:37 PM on December 6, 2023
This is very cool! I don't pretend to understand the science, but it does help me understand some of the back magic that is electricity.
posted by dg at 3:04 PM on December 6, 2023
posted by dg at 3:04 PM on December 6, 2023
0.4 pF/cm. https://en.wikipedia.org/wiki/Gimmick_capacitor
posted by Chef Flamboyardee at 3:23 PM on December 6, 2023 [5 favorites]
posted by Chef Flamboyardee at 3:23 PM on December 6, 2023 [5 favorites]
> Conventional wisdom about electricity is that unlike water, electricity does not follow the path of least resistance.
> Instead, it follows all paths in inverse proportion to their resistance.
I can assure you from bitter experience that if you sit in a boat with multiple holes in it, water flows into all of them, inversely proportional to whatever means of resistance you can find to shove into each.
posted by automatronic at 4:25 PM on December 6, 2023 [5 favorites]
> Instead, it follows all paths in inverse proportion to their resistance.
I can assure you from bitter experience that if you sit in a boat with multiple holes in it, water flows into all of them, inversely proportional to whatever means of resistance you can find to shove into each.
posted by automatronic at 4:25 PM on December 6, 2023 [5 favorites]
posted by jedicus at 11:58 AM on December 6
This thread's tiny little hbomberguy moment right there.
Okay, yes, it's not that. But it is a pretty cromulent explainer of the situation.
posted by hippybear at 4:45 PM on December 6, 2023
This thread's tiny little hbomberguy moment right there.
Okay, yes, it's not that. But it is a pretty cromulent explainer of the situation.
posted by hippybear at 4:45 PM on December 6, 2023
In water terms, the initial wave of electricity in this video is making me think of a tsunami, or the leading wave of a flash flood in a dry riverbed.
posted by clawsoon at 5:31 PM on December 6, 2023
posted by clawsoon at 5:31 PM on December 6, 2023
There's no such thing as an open circuit, because that would be an oxymoron. This fact does not change simply because somebody who by his own admission started with no understanding of transmission lines chooses to ignore distributed capacitance and inductance and reason as if Ohm's Law were the be-all and end-all of electrical theory.
I think you're misinterpreting when he says ‘didn't properly understand’ in the phrase “I didn't properly understand the concept of line impedance for example until I had built the apparatus that I planned to use for this video and used it to watch waves of electricity traveling around in wires”.
He doesn't mean ‘I'd never seen the theory before’, he means ‘I'd never modelled and measured it for myself before’. Here's a making of video where he covers his experiments about that in more detail.
posted by ambrosen at 8:36 PM on December 6, 2023 [1 favorite]
I think you're misinterpreting when he says ‘didn't properly understand’ in the phrase “I didn't properly understand the concept of line impedance for example until I had built the apparatus that I planned to use for this video and used it to watch waves of electricity traveling around in wires”.
He doesn't mean ‘I'd never seen the theory before’, he means ‘I'd never modelled and measured it for myself before’. Here's a making of video where he covers his experiments about that in more detail.
posted by ambrosen at 8:36 PM on December 6, 2023 [1 favorite]
I have always wondered about this, so now I am very happy.
posted by Well I never at 8:40 PM on December 6, 2023
posted by Well I never at 8:40 PM on December 6, 2023
Kathy Loves Physics (another YT channel worth your subscription) posted a deep dive today into the historical people and experiments that discovered all of this behaviour a century or two ago:
https://youtu.be/MyzhyhN2038?si=HNhY_UoeepAzVfwe
posted by Popular Ethics at 10:42 PM on December 6, 2023 [1 favorite]
https://youtu.be/MyzhyhN2038?si=HNhY_UoeepAzVfwe
posted by Popular Ethics at 10:42 PM on December 6, 2023 [1 favorite]
Yes, "open circuit" is indeed a term in common use. But it is an oxymoron by inspection, so whenever you do see it in use then it should always be questioned, even if only briefly.
If connecting a power supply to some arrangement of components results in interesting behaviour then the making of that connection has created a circuit, regardless of whether or not it happens to be one that's open with respect to direct current. And wires are components, regardless of the fact that in many cases their capacitance, inductance and even resistance can reasonably be disregarded.
Gimmick_capacitor
Exactly. The run of twisted pair that's open at one end doesn't make its branch of the circuit under test open unless you restrict yourself to using steady-state (i.e. direct current) analysis, which is completely the wrong set of tools to bring to bear when trying to answer time-sensitive questions such as how long it takes for current to start flowing from a battery after a switch contact closes.
When timing is what you care about, you need to be able to measure and/or model the reactive effects of inductance and capacitance as well as the resistive effects modelled by Ohm's Law. Assuming that no current can flow through a gimmick capacitor simply because its resistance is arbitrarily high is incorrect; an impediment, not an aid, to understanding.
posted by flabdablet at 3:30 AM on December 7, 2023 [2 favorites]
If connecting a power supply to some arrangement of components results in interesting behaviour then the making of that connection has created a circuit, regardless of whether or not it happens to be one that's open with respect to direct current. And wires are components, regardless of the fact that in many cases their capacitance, inductance and even resistance can reasonably be disregarded.
Gimmick_capacitor
Exactly. The run of twisted pair that's open at one end doesn't make its branch of the circuit under test open unless you restrict yourself to using steady-state (i.e. direct current) analysis, which is completely the wrong set of tools to bring to bear when trying to answer time-sensitive questions such as how long it takes for current to start flowing from a battery after a switch contact closes.
When timing is what you care about, you need to be able to measure and/or model the reactive effects of inductance and capacitance as well as the resistive effects modelled by Ohm's Law. Assuming that no current can flow through a gimmick capacitor simply because its resistance is arbitrarily high is incorrect; an impediment, not an aid, to understanding.
posted by flabdablet at 3:30 AM on December 7, 2023 [2 favorites]
I'm trying to learn from your comment, flabdablet, but I don't have the background to understand it properly. Are you saying that something qualitatively different would happen if the twisted pair was replaced with a single straight wire - i.e. a completely different pattern of electrical flow that can't be mapped to his result with a set of simple transformations - or would it merely result in a quantitative change - i.e. the same basic pattern, just faster/bigger/slower/smaller?
I have the vague impression that everything is a capacitor, so I'm guessing that it'd just be a quantitative change, but's that's just a guess on my part.
The run of twisted pair that's open at one end doesn't make its branch of the circuit under test open
Wait, did you just indirectly say "open circuit"? :-)
posted by clawsoon at 4:42 AM on December 7, 2023
I have the vague impression that everything is a capacitor, so I'm guessing that it'd just be a quantitative change, but's that's just a guess on my part.
The run of twisted pair that's open at one end doesn't make its branch of the circuit under test open
Wait, did you just indirectly say "open circuit"? :-)
posted by clawsoon at 4:42 AM on December 7, 2023
Just so we're clear, flabdalet, is this a problem with his terminology, or are you saying that the entire video is misleading?
posted by ambrosen at 5:04 AM on December 7, 2023 [1 favorite]
posted by ambrosen at 5:04 AM on December 7, 2023 [1 favorite]
I have the vague impression that everything is a capacitor
Probably closer to the mark to say that everything is a transmission line consisting of distributed resistive, capacitative and inductive elements, some of which we can often get away with ignoring especially when working with DC.
are you saying that the entire video is misleading?
I'm saying that I think presenting the idea of current flowing in a transmission line that's open at the undriven end as some kind of paradox is unhelpful, because it creates confusion where none need be.
The only way the paradox works as a paradox is if the audience is led to expect it to be reasonable to think of long pieces of closely coupled wire as essentially equivalent to isolated zero-ohm conductors of negligible length, which they are absolutely not.
I also note that he talks a lot about current flow while what he's actually measuring is potential difference (voltage).
The question of what voltages one should expect to be able to measure at various points in such a circuit, and at what times, is a good question; but I think a clearer understanding of it would be had by starting with an exploration at how impulses behave in more physically accessible transmission media.
Watch a pile of slinky wave demos, paying attention to the way that pulses applied at one end get reflected from the far end for both fixed and free terminations.
If you're serious, get hold of three slinkies and build yourself a physical analogue to AlphaPhoenix's "fork in the road" twisted pairs, where the far end of one of the forks is fixed and the other is free. Play with applying step changes to the position of the slinky at the "battery" end, and pay attention to how the ensuing waves bounce around and interact. Then come back to the section of this video where he's complaining about how weird his initial scope traces look. You too might find yourself fighting an urge to shout "well why would you expect anything else to happen?" at the screen. As a bonus, you'll probably get a good head start at understanding how time-domain reflectometers work.
AlphaPhoenix does show some water trough models that do something similar, but I think they're less useful as intuition pumps than the slinky spring models because they foreground the distracting idea that what we're trying to understand here is primarily some kind of flow as opposed to a somewhat more abstract propagating energy transfer.
posted by flabdablet at 5:51 AM on December 7, 2023 [3 favorites]
Probably closer to the mark to say that everything is a transmission line consisting of distributed resistive, capacitative and inductive elements, some of which we can often get away with ignoring especially when working with DC.
are you saying that the entire video is misleading?
I'm saying that I think presenting the idea of current flowing in a transmission line that's open at the undriven end as some kind of paradox is unhelpful, because it creates confusion where none need be.
The only way the paradox works as a paradox is if the audience is led to expect it to be reasonable to think of long pieces of closely coupled wire as essentially equivalent to isolated zero-ohm conductors of negligible length, which they are absolutely not.
I also note that he talks a lot about current flow while what he's actually measuring is potential difference (voltage).
The question of what voltages one should expect to be able to measure at various points in such a circuit, and at what times, is a good question; but I think a clearer understanding of it would be had by starting with an exploration at how impulses behave in more physically accessible transmission media.
Watch a pile of slinky wave demos, paying attention to the way that pulses applied at one end get reflected from the far end for both fixed and free terminations.
If you're serious, get hold of three slinkies and build yourself a physical analogue to AlphaPhoenix's "fork in the road" twisted pairs, where the far end of one of the forks is fixed and the other is free. Play with applying step changes to the position of the slinky at the "battery" end, and pay attention to how the ensuing waves bounce around and interact. Then come back to the section of this video where he's complaining about how weird his initial scope traces look. You too might find yourself fighting an urge to shout "well why would you expect anything else to happen?" at the screen. As a bonus, you'll probably get a good head start at understanding how time-domain reflectometers work.
AlphaPhoenix does show some water trough models that do something similar, but I think they're less useful as intuition pumps than the slinky spring models because they foreground the distracting idea that what we're trying to understand here is primarily some kind of flow as opposed to a somewhat more abstract propagating energy transfer.
posted by flabdablet at 5:51 AM on December 7, 2023 [3 favorites]
And yes, fluid dynamics does include similar-looking wave propagation patterns - but if wave propagation in electrical transmission lines is what you're interested in understanding, quite a lot of the rest of fluid dynamics is going to be more misleading than helpful.
Maxwell's equations are far more tractable than Navier-Stokes.
posted by flabdablet at 6:00 AM on December 7, 2023
Maxwell's equations are far more tractable than Navier-Stokes.
posted by flabdablet at 6:00 AM on December 7, 2023
isolated zero-ohm conductors
Are you talking about superconductors, or about a useful simplifying idealization of normal wires?
posted by clawsoon at 6:02 AM on December 7, 2023
Are you talking about superconductors, or about a useful simplifying idealization of normal wires?
posted by clawsoon at 6:02 AM on December 7, 2023
The second one.
posted by flabdablet at 6:06 AM on December 7, 2023 [1 favorite]
posted by flabdablet at 6:06 AM on December 7, 2023 [1 favorite]
Any decent length of superconductor is a transmission line as well, just one whose distributed series resistance component is actually rather than approximately zero. All the inductance and capacitance is still there.
posted by flabdablet at 6:06 AM on December 7, 2023 [1 favorite]
posted by flabdablet at 6:06 AM on December 7, 2023 [1 favorite]
I tell my students every semester that if I could erase a single phrase from their brains so we could start over with the correct context, it would be "the path of least resistance." If I get two, Ohm's Law is next on the list.
The measurements that start around 11:00 are a really lovely way of showing transmission line behavior that we typically ask people to accept via bulk results (eg you can see the effects of reflections) rather than direct point-by-point measurement. On the other hand his conflation of voltage (which is what his scope is measuring) and current (which you can only infer via a possibly-complicated process) is really irritating from the standpoint of someone who is trying to clear up misconceptions about electricity.
On balance I'm glad this exists but I really wish folks who did this kind of thing would put it in the broader context that EVERY model we have to describe physical systems has a "works most of the time" Easy Mode and a "things got interesting" Hard Mode - if you put two kids on a teeter-totter on the playground, it's going to act just like the 1:1 lever you studied in high school physics. If you put a house on the same teeter-totter, all of a sudden you had better include the effects of soil fracture, beam bending, column buckling, etc or you'll have absolutely no way to predict what's going to happen. All of those effects were present with the kids too - just in amounts so small you were able to round them all off to zero.
In electrical systems the bad news is that it's really easy to get into that interesting domain by going really fast, and the good news is that most of the time you still don't care because all the interesting stuff is over before your finger left the switch. If you design circuits for Nvidia it's a different story but knowing how to use the interesting tools is why you work for Nvidia.
A lot of the kerfuffle about stuff like this reminds me of the day a professor I teach with accidentally torpedoed an entire lecture by mentioning in passing "there's no such thing as DC" because (a) you turned it on sometime and (b) you'd have to observe it forever in order to distinguish it from a very very slow sinusoid. True, but that's a very hard distinction to absorb if you're already working hard to grapple with the fundamentals.
posted by range at 7:34 AM on December 7, 2023 [6 favorites]
The measurements that start around 11:00 are a really lovely way of showing transmission line behavior that we typically ask people to accept via bulk results (eg you can see the effects of reflections) rather than direct point-by-point measurement. On the other hand his conflation of voltage (which is what his scope is measuring) and current (which you can only infer via a possibly-complicated process) is really irritating from the standpoint of someone who is trying to clear up misconceptions about electricity.
On balance I'm glad this exists but I really wish folks who did this kind of thing would put it in the broader context that EVERY model we have to describe physical systems has a "works most of the time" Easy Mode and a "things got interesting" Hard Mode - if you put two kids on a teeter-totter on the playground, it's going to act just like the 1:1 lever you studied in high school physics. If you put a house on the same teeter-totter, all of a sudden you had better include the effects of soil fracture, beam bending, column buckling, etc or you'll have absolutely no way to predict what's going to happen. All of those effects were present with the kids too - just in amounts so small you were able to round them all off to zero.
In electrical systems the bad news is that it's really easy to get into that interesting domain by going really fast, and the good news is that most of the time you still don't care because all the interesting stuff is over before your finger left the switch. If you design circuits for Nvidia it's a different story but knowing how to use the interesting tools is why you work for Nvidia.
A lot of the kerfuffle about stuff like this reminds me of the day a professor I teach with accidentally torpedoed an entire lecture by mentioning in passing "there's no such thing as DC" because (a) you turned it on sometime and (b) you'd have to observe it forever in order to distinguish it from a very very slow sinusoid. True, but that's a very hard distinction to absorb if you're already working hard to grapple with the fundamentals.
posted by range at 7:34 AM on December 7, 2023 [6 favorites]
almost everything in the universe is a non-circuit
Anyone who has ever had to set foot in an EMC lab wishes this were even close to true. That's the place that all your electronics has to go to check that it doesn't interfere with other electronics. Where if you are a young and naive electronics engineer, you get reminded that:
Note the one thing crossed off this list. Because superconductors are, incredibly, a thing! If you spend a truly ridiculous amount of money cooling weird shit to insane temperatures, you can get actual honest to god zero resistance. Congratulations, you show-off, you still get to live with the rest.
posted by automatronic at 7:56 AM on December 7, 2023 [8 favorites]
Anyone who has ever had to set foot in an EMC lab wishes this were even close to true. That's the place that all your electronics has to go to check that it doesn't interfere with other electronics. Where if you are a young and naive electronics engineer, you get reminded that:
- An antenna is not an open circuit.
- Every conductive object is an antenna, to some degree.
- Every varying current that gets onto an antenna is transmitted, to some degree.
- Every antenna receives signals from every other antenna in the universe, to some degree.
- A capacitor is not an open circuit.
- Every varying voltage that gets onto one side of a capacitor shows up on the other side, to some degree.
- Every pair of conductive objects in the universe is a capacitor, to some degree.
- An inductor is not a short circuit.
- Every conductive object is an inductor, to some degree.
- Every varying current that gets into an inductor creates a magnetic field, to some degree.
- Every magnetic field induces current in every other conductive object in the universe, to some degree.
- A resistor is not a short circuit.
Every object in the universe is a resistor, to some degree.- Every voltage across a resistance creates a current, to some degree.
Note the one thing crossed off this list. Because superconductors are, incredibly, a thing! If you spend a truly ridiculous amount of money cooling weird shit to insane temperatures, you can get actual honest to god zero resistance. Congratulations, you show-off, you still get to live with the rest.
posted by automatronic at 7:56 AM on December 7, 2023 [8 favorites]
Have anybody ever tried creating a zero-capacitance or zero-inductance substance analogous to zero-resistance superconductors? (Is that even a sensible question to ask?)
posted by clawsoon at 8:04 AM on December 7, 2023
posted by clawsoon at 8:04 AM on December 7, 2023
Inductance and capacitance ultimately arise from the fact that electromagnetism is a single unified force - I'm not aware of a temperature below which they split in the same way that the electromagnetic and weak forces are split at normal human temperatures. I'm pretty sure you will see inductive and capacitive effects in any system that moves charges around (if not in lumped-parameter form then in continuous Maxwell-equation-style effects) but maybe that just represents a lack of imagination on my part.
posted by range at 8:21 AM on December 7, 2023 [2 favorites]
posted by range at 8:21 AM on December 7, 2023 [2 favorites]
Also, inductance and capacitance are not so much properties of the substance with the charge carriers in it, as the properties of the shapes and relative arrangements of such substances.
Inductance arises as a consequence of the fact that making a charge move creates a magnetic field, interacting with the flip side of that relationship where changing a magnetic field applies motive forces to charges within the field. These two effects combine in such a way that once a charge is in motion it tends to keep that state of motion, because any change in the way a charge is moving causes a change in the local magnetic field that opposes the change in the charge's motion.
You can minimize inductance by arranging for much of the magnetic field associated with any current to be cancelled by having an equal current flowing in the opposite direction, physically very close to the first one. For example, the inductance of a twisted pair or coaxial cable carrying a balanced signal, where the current in one of the conductors is always the exact opposite of the current in the other, will be much lower than that of a pair of loose wires carrying the same balanced signal.
Capacitance arises as a consequence of the fact that charge imbalances of opposite sign attract one another. If there's a conductor between two regions, each of which has a charge imbalance opposite to that of the other - for example, one region has a net positive charge while the other has a net negative - then that attraction will cause the charges to move, and current will flow until the resulting net charge is equal in both regions. But if there's a dielectric - an insulator - between the two regions, and yet they're close enough to exert any appreciable attractive force on each other, then they will just sit there doing that, and an electric field will persist between the regions.
Both the magnetic field created around a current and the electric field created by a standing charge imbalance (i.e. a potential difference, or voltage) are stores of potential energy.
You can add to the potential energy stored in an inductor by supplying additional electrical energy, thereby forcing the current flowing through the inductor to increase, thereby making its magnetic field stronger. That increase in field strength is the increase in potential energy, and making that increase happen costs the driving current that same amount of electrical energy.
You can get the stored potential energy back out again by arranging for your current to carry electrical energy away to somewhere else, thereby allowing the magnetic field to collapse over time. That collapse is the reduction of potential energy stored in the magnetic field, and that reduction returns electrical energy to the current.
If the conducting part of your inductor is itself a superconductor, so that current can flow in it without dissipating energy in the form of resistance-induced heating, then you can simply connect your inductor's terminals together to form a complete circuit and it will store energy indefinitely. Run a second conductor through the same magnetic field zone, and force a current through that so that the field gets built, and then turn the driving current off, and it will just show up inside the superconducting loop, accompanied by its corresponding magnetic field. You can keep on doing that until the magnetic field gets strong enough to bust the superconductivity, at which point magnetic potential energy starts being converted to electrical energy which starts being dissipated as resistance-induced heating, which makes the superconductor even less superconducting and the whole thing collapses. Depending how much energy you'd managed to pump into the field first, this might even let out the magic smoke.
As a side note: one way to think about permanent magnets and how they fucking work is to conceptualize them as vast crowds of aligned, superconducting looped inductors at particle scale, each with its own quantum of stored magnetic potential energy; you don't need to do huge amounts of violence to the idea of a tiny looped current to turn it into the idea of an individual electron's spin. What you can't do with an electron's spin, though, is turn it into an actual current in a bigger loop in order to collapse its associated magnetic field and extract the potential energy from that. Extracting potential energy from a permanent magnet's field requires physically reducing the size of that field and makes the extracted energy come in kinetic form, not electrical.
Heading on over to the electric field aspect: you can add to the potential energy stored in a capacitor just by forcing a current to flow through it. Because there's no conductive path between the capacitor's plates, forcing a current through it necessarily involves charge being stored in the plates, thereby increasing the strength of the electric field between them. That increase in field strength is the increase in stored potential energy. You can keep on doing that until the electric field gets intense enough either to damage the dielectric (or for truly teeny tiny capacitors, allow electrons to tunnel across it).
Storing energy in or removing energy from a reactive component (inductor or capacitor) is a change in energy, and change in energy over time is power, and electrical power is just voltage across multiplied by current through. You can spot whether the component is storing or releasing energy by looking at the direction of the current through compared to the voltage across: if current is going into the positive terminal then the component is storing energy, but if current is coming out of the positive terminal then it's releasing it.
If a capacitor or inductor is part of a circuit but isn't having energy stored in or removed from it, then the instantaneous voltage across it multiplied by the instantaneous current through it will therefore necessarily be zero. For a perfect inductor, that happens when current through is constant, and voltage across is zero; for a perfect capacitor, it happens when voltage across is constant, and current through is zero.
In practice, any real inductor is best modelled as a perfect inductor with a low-value resistor in series, and any real capacitor is best modelled as a perfect capacitor with a high-value resistor in parallel.
Ohm's Law describes the relationship between resistance, voltage and current under steady-state conditions for perfect resistors: V = IR, where V is voltage across the component in volts, I is current through it in amps, and R is its resistance in ohms. And note that for resistors, current always flows into the positive terminal. This is because resistors don't store the energy you supply them with, they just dissipate it as heat, so there's no getting it out again.
There are similar laws describing the relationship between instantaneous voltage, instantaneous current, capacitance and inductance for reactive components. For inductors, we have
V = LdI/dt where V is voltage across the component in volts, L is inductance in henries, and dI/dt is the rate of change of current in amps per second.
Note that this law implies that the faster you try to change the current flowing in an inductor, the higher the voltage across it will be. This is the fundamental principle behind car ignition coils: the points opening causes a very sudden change in current flowing in the primary side of the coil, which responds by generating a secondary-side voltage spike that's high enough to strip the outer electrons off the gas between the spark plug electrodes and turn it into a conductive plasma.
For capacitors, the law is
I = CdV/dt where I is current through in amps, C is capacitance in farads, and dV/dt is rate of change of voltage across in volts per second.
And this law implies that the faster you try to change the voltage across a capacitor, the more current it will draw or supply. Deliberately shorting out a capacitor that's been charged to any decent voltage makes some quite entertaining arcs, and discharging a large one can even weld your screwdriver to its terminals.
The main thing to understand about the voltage and current relationships for reactive components is that they are time sensitive. Conversely, any investigation of time-sensitive behaviour for any circuit requires considering not only its assorted ohmic resistances but its reactive aspects as well.
posted by flabdablet at 11:20 AM on December 7, 2023 [4 favorites]
Inductance arises as a consequence of the fact that making a charge move creates a magnetic field, interacting with the flip side of that relationship where changing a magnetic field applies motive forces to charges within the field. These two effects combine in such a way that once a charge is in motion it tends to keep that state of motion, because any change in the way a charge is moving causes a change in the local magnetic field that opposes the change in the charge's motion.
You can minimize inductance by arranging for much of the magnetic field associated with any current to be cancelled by having an equal current flowing in the opposite direction, physically very close to the first one. For example, the inductance of a twisted pair or coaxial cable carrying a balanced signal, where the current in one of the conductors is always the exact opposite of the current in the other, will be much lower than that of a pair of loose wires carrying the same balanced signal.
Capacitance arises as a consequence of the fact that charge imbalances of opposite sign attract one another. If there's a conductor between two regions, each of which has a charge imbalance opposite to that of the other - for example, one region has a net positive charge while the other has a net negative - then that attraction will cause the charges to move, and current will flow until the resulting net charge is equal in both regions. But if there's a dielectric - an insulator - between the two regions, and yet they're close enough to exert any appreciable attractive force on each other, then they will just sit there doing that, and an electric field will persist between the regions.
Both the magnetic field created around a current and the electric field created by a standing charge imbalance (i.e. a potential difference, or voltage) are stores of potential energy.
You can add to the potential energy stored in an inductor by supplying additional electrical energy, thereby forcing the current flowing through the inductor to increase, thereby making its magnetic field stronger. That increase in field strength is the increase in potential energy, and making that increase happen costs the driving current that same amount of electrical energy.
You can get the stored potential energy back out again by arranging for your current to carry electrical energy away to somewhere else, thereby allowing the magnetic field to collapse over time. That collapse is the reduction of potential energy stored in the magnetic field, and that reduction returns electrical energy to the current.
If the conducting part of your inductor is itself a superconductor, so that current can flow in it without dissipating energy in the form of resistance-induced heating, then you can simply connect your inductor's terminals together to form a complete circuit and it will store energy indefinitely. Run a second conductor through the same magnetic field zone, and force a current through that so that the field gets built, and then turn the driving current off, and it will just show up inside the superconducting loop, accompanied by its corresponding magnetic field. You can keep on doing that until the magnetic field gets strong enough to bust the superconductivity, at which point magnetic potential energy starts being converted to electrical energy which starts being dissipated as resistance-induced heating, which makes the superconductor even less superconducting and the whole thing collapses. Depending how much energy you'd managed to pump into the field first, this might even let out the magic smoke.
As a side note: one way to think about permanent magnets and how they fucking work is to conceptualize them as vast crowds of aligned, superconducting looped inductors at particle scale, each with its own quantum of stored magnetic potential energy; you don't need to do huge amounts of violence to the idea of a tiny looped current to turn it into the idea of an individual electron's spin. What you can't do with an electron's spin, though, is turn it into an actual current in a bigger loop in order to collapse its associated magnetic field and extract the potential energy from that. Extracting potential energy from a permanent magnet's field requires physically reducing the size of that field and makes the extracted energy come in kinetic form, not electrical.
Heading on over to the electric field aspect: you can add to the potential energy stored in a capacitor just by forcing a current to flow through it. Because there's no conductive path between the capacitor's plates, forcing a current through it necessarily involves charge being stored in the plates, thereby increasing the strength of the electric field between them. That increase in field strength is the increase in stored potential energy. You can keep on doing that until the electric field gets intense enough either to damage the dielectric (or for truly teeny tiny capacitors, allow electrons to tunnel across it).
Storing energy in or removing energy from a reactive component (inductor or capacitor) is a change in energy, and change in energy over time is power, and electrical power is just voltage across multiplied by current through. You can spot whether the component is storing or releasing energy by looking at the direction of the current through compared to the voltage across: if current is going into the positive terminal then the component is storing energy, but if current is coming out of the positive terminal then it's releasing it.
If a capacitor or inductor is part of a circuit but isn't having energy stored in or removed from it, then the instantaneous voltage across it multiplied by the instantaneous current through it will therefore necessarily be zero. For a perfect inductor, that happens when current through is constant, and voltage across is zero; for a perfect capacitor, it happens when voltage across is constant, and current through is zero.
In practice, any real inductor is best modelled as a perfect inductor with a low-value resistor in series, and any real capacitor is best modelled as a perfect capacitor with a high-value resistor in parallel.
Ohm's Law describes the relationship between resistance, voltage and current under steady-state conditions for perfect resistors: V = IR, where V is voltage across the component in volts, I is current through it in amps, and R is its resistance in ohms. And note that for resistors, current always flows into the positive terminal. This is because resistors don't store the energy you supply them with, they just dissipate it as heat, so there's no getting it out again.
There are similar laws describing the relationship between instantaneous voltage, instantaneous current, capacitance and inductance for reactive components. For inductors, we have
V = LdI/dt where V is voltage across the component in volts, L is inductance in henries, and dI/dt is the rate of change of current in amps per second.
Note that this law implies that the faster you try to change the current flowing in an inductor, the higher the voltage across it will be. This is the fundamental principle behind car ignition coils: the points opening causes a very sudden change in current flowing in the primary side of the coil, which responds by generating a secondary-side voltage spike that's high enough to strip the outer electrons off the gas between the spark plug electrodes and turn it into a conductive plasma.
For capacitors, the law is
I = CdV/dt where I is current through in amps, C is capacitance in farads, and dV/dt is rate of change of voltage across in volts per second.
And this law implies that the faster you try to change the voltage across a capacitor, the more current it will draw or supply. Deliberately shorting out a capacitor that's been charged to any decent voltage makes some quite entertaining arcs, and discharging a large one can even weld your screwdriver to its terminals.
The main thing to understand about the voltage and current relationships for reactive components is that they are time sensitive. Conversely, any investigation of time-sensitive behaviour for any circuit requires considering not only its assorted ohmic resistances but its reactive aspects as well.
posted by flabdablet at 11:20 AM on December 7, 2023 [4 favorites]
All these years, I didn't know that flabdablet had a slot where if you put in a quarter you'd get an electricity education.
posted by hippybear at 11:31 AM on December 7, 2023 [3 favorites]
posted by hippybear at 11:31 AM on December 7, 2023 [3 favorites]
Additional formula for energy stored in a capacitor:
E = ½CV2, where E is energy in joules, C is capacitance in farads, and V is the potential difference between the plates in volts.
For energy stored in an inductor:
E = ½LI2, where E is energy in joules, L is inductance in henries, and I is the current through the inductor in amps.
Note also that unless you do have access to actual superconductors it's generally harder to approximate a perfect conductor than a perfect insulator, so in practice if you want to store energy in a component for appreciable amounts of time, you're better off with a capacitor than an inductor.
posted by flabdablet at 11:33 AM on December 7, 2023
E = ½CV2, where E is energy in joules, C is capacitance in farads, and V is the potential difference between the plates in volts.
For energy stored in an inductor:
E = ½LI2, where E is energy in joules, L is inductance in henries, and I is the current through the inductor in amps.
Note also that unless you do have access to actual superconductors it's generally harder to approximate a perfect conductor than a perfect insulator, so in practice if you want to store energy in a component for appreciable amounts of time, you're better off with a capacitor than an inductor.
posted by flabdablet at 11:33 AM on December 7, 2023
I didn't know that flabdablet had a slot where if you put in a quarter you'd get an electricity education
Wait til you see what happens when you put in twenty bucks!
posted by flabdablet at 11:35 AM on December 7, 2023 [3 favorites]
Wait til you see what happens when you put in twenty bucks!
posted by flabdablet at 11:35 AM on December 7, 2023 [3 favorites]
Wait til you see what happens when you put in twenty bucks!
Well, $20 will buy you some things while in town, Would this be the same as that?
posted by hippybear at 12:42 PM on December 7, 2023 [2 favorites]
Well, $20 will buy you some things while in town, Would this be the same as that?
posted by hippybear at 12:42 PM on December 7, 2023 [2 favorites]
Good Lord. I feel like I just took a high school course on YouTube, and then moved immediately to a college course here on Mefi. Fladdeblet, range, and automatronic thank you VERY MUCH for your contributions to this post!
posted by Frayed Knot at 12:43 PM on December 7, 2023 [2 favorites]
posted by Frayed Knot at 12:43 PM on December 7, 2023 [2 favorites]
It's a good video. I had a similar "school of hard knocks" lesson in transmission lines while building some fast electronics as a postdoc, with signals that needed to survive a few hundred meters of cable.
A cable ending with a zero-ohm short must have a reflected voltage signal with the opposite sign, to maintain zero voltage. Meanwhile an open terminus must generate a reflected (voltage) wave with the same sign, to keep the current across the gap at zero. Somewhere between these is a resistance which generates no reflection at all. That resistance matches the impedance of the cable.
Part of his noise is that his handmade cables don't have uniform twist, and therefore don't have uniform impedance. Impedance changes give you reflections from the middle of the cable, which show up in practice as distorted signals.
posted by fantabulous timewaster at 2:09 PM on December 7, 2023 [6 favorites]
A cable ending with a zero-ohm short must have a reflected voltage signal with the opposite sign, to maintain zero voltage. Meanwhile an open terminus must generate a reflected (voltage) wave with the same sign, to keep the current across the gap at zero. Somewhere between these is a resistance which generates no reflection at all. That resistance matches the impedance of the cable.
Part of his noise is that his handmade cables don't have uniform twist, and therefore don't have uniform impedance. Impedance changes give you reflections from the middle of the cable, which show up in practice as distorted signals.
posted by fantabulous timewaster at 2:09 PM on December 7, 2023 [6 favorites]
« Older There's no Rudolph, he's just one of them urban... | Blue chat bubbles for everyone Newer »
This thread has been archived and is closed to new comments
posted by grumpybear69 at 9:14 AM on December 6, 2023 [5 favorites]