| General > General Technical Chat |
| Does a capacitor charges smooth, or in stairs? |
| << < (4/16) > >> |
| Nominal Animal:
--- Quote from: RoGeorge on June 06, 2020, 09:24:14 pm --- ;D --- End quote --- I for one am not booing you, RoGeorge; I am only saying it is not an intrinsic property of all capacitors (and because of wording, that makes me disagree with your statement/poll). (This is also at the core of my recent one-sided arguments about learning things, even if those feel like rants or useless whining to others.) Let me repeat an example I've already mentioned recently elsewhere in this forum. At least here, in school, kids are taught that electrons orbit atomic nuclei. The core problem is the definition of "orbit". (In physics and chemistry, they're not called "orbits", they're called "orbitals", "orbital shells", and so on.) Yes, the electron does have properties that are analogous to rotating around a center: angular momentum, something like an orbital radius, and even spin (analogous to "direction of rotation"). But the fact is that those electrons aren't actually moving, they're delocalized around the nucleus. This delocalization can be described using quantum mechanics, so well that the best models we have (most accurate with respect to real-world measurable properties) of how molecules behave and react, are based on the quantum mechanical modeling of just the outermost interacting electrons! No fitting to real-world data, just pure math and some physical constants, and out pops our best predictions of what certain molecules are like, and what their properties are. What does it matter, then? The orbit model gives an intuitive grasp of the properties – radius and angular momentum in particular; and even spin! – and those are what even physicists and chemists work with, so there's no harm, right? Wrong. When a charged particle like an electron is deflected, accelerated, or decelerated by another charged particle, like an atomic nucleus or another electron, it radiates energy. We call this phenomenon Brehmsstrahlung, or "braking radiation". It happens for all charged particles, and is fundamentally due to conservation of energy. So, it is fundamental to charged particles, including electrons. The problem occurs when one tries to integrate the two, Brehmsstrahlung and electron orbits. I fear that most people simply decide that somehow an electron orbiting an atomic nucleus, or flitting about in a metal lattice, is just an exception to Brehmsstrahlung, give a quick :-//, and ignore the dichotomy. That is magical thinking, and leads to the inability of correctly choosing which "model" applies in which situation – so those people can recite information, but not apply it in real world to solve a problem or predict or estimate physical phenomena. But, there is no dichotomy, only an incorrect description and thus incorrect understanding. It would have been not that hard to explain correctly in the first place: delocalization looks like the electron is "blurred" across its "orbital", which looks more like a cloud than a ring; but if you measure it, it is like putting a stick into the spokes of a fast-turning bicycle wheel: you'll see one quite clearly. The properties these delocalized electrons have, are almost exactly like if the electron was orbiting the nucleus like a moon orbiting a planet, and that's why we use that analog; but the actual physical description is much weirder, and involves quantum mechanics. Trying to fix the mislearning later requires "un-learning" – or at minimum, accepting you were taught an incorrect thing because someone thought that teaching the actual thing was too hard. Now, in the capacitor case, the staircase effect is not due to any intrinsic property of capacitors, but due to capacitors being part of an electrical circuit. You can extend this into the capacitors themselves, if you model a long skinny capacitor charged or discharged at one end, in which case the capacitor itself becomes part of the circuit – transmission line, really. The model breaks down if you have either very short transmission lines, or if the transmission "line" does not have a clearly defined length, so that the change in the electromagnetic field (in the transmission line) does not have a clear wavefront anymore. I suppose the reason for emphasizing the transmission line model is to break through some preconceptions the students might have, and make sure they understand that the transmission line model applies to even circuits on a circuit board – even if in practice it only matters enough to worry about when delivering power via relatively long transmission lines. That it is not something that applies to, well, actual transmission lines, but is a model that accurately describes what happens in circuits. The laws that we use every day to describe electrical circuits or networks – Kirchhoff's laws, Ohm's law, Norton's theorem, Thévenin's theorem – are models that do describe the behaviour of the circuit or network, but themselves are the result of more fundamental descriptions of charges; they describe the behaviour, not what is nor why. The underlying reality is much, much weirder, but these models give us tools to work with such systems without getting too bogged down into the weirdness. That's basically what my own field, molecular dynamics simulations with classical potential models is, "classical" meaning just "non-quantum-mechanic". We can do QM for maybe a system with a thousand electrons (but that may take a week to a month, depending on how large a computing cluster you can grab), but that's just a pretty small molecule – and we need repeated boundary conditions, too, which can be an issue if we want surface phenomena. With less reliable/accurate but much simpler classical interaction models we can model systems with hundreds of millions to billions of atoms, getting into things like corrosion and damage resistance, defect migration in the large scale (and self-healing materials), ion implantation, sputtering, and so on, and get very useful real-world results. |
| Labrat101:
Hi, Nominal Animal That was very Good explanation . Very true and totally correct. I did not want to go into that sort of explanation as it maybe above some of the younger reader that may read this forum. There are Many factors in transmission lines etc . There is one factor that will always be .. is our constant Mr. "Time" how ever we look at it will effect. And I think our friend "RoGeorge" has some interesting ideas . and it gives him a Job. I remember that Dave made some videos a while back on similar . A square orbit would be more interesting . as it would allow 1 to cut corners (joking of course) |
| Someone:
--- Quote from: bdunham7 on June 06, 2020, 03:56:25 pm ---That all looks theoretical. Can you show us an actual example of this phenomenon, like a scope capture or something? :box: --- End quote --- Image sensors with sub-single-photon noise. |
| StillTrying:
If I use 2 1/2 meters of coax as the capacitor I can see the steps. LOL Charged and discharged through a 390R resistor. |
| RoGeorge:
--- Quote from: bdunham7 on June 06, 2020, 03:56:25 pm ---That all looks theoretical. Can you show us an actual example of this phenomenon, like a scope capture or something? :box: --- End quote --- I will love to see an oscilloscope screen capture with some stair steps or at least some undulating exponential, where the undulations can be proved as depending of the capacitor's geometry. I tried once with an 100MHz oscilloscope and failed miserably. If anybody have the time and the equipment to experiment, please share the results. A clean experiment setup is much harder in real life than it is on paper. That's why I never blame those that like to simulate. However, even if I'm aware of a simulation's advantages, I'm not a big fan of it. The best "comments wisdom quote" I can remember (related with my feelings about simulation) is about a software that was simulating robots (mainly a physics engine). Some random dude commented "Simulated robots, simulated fun!" ;D --- Quote from: Nominal Animal on June 07, 2020, 06:45:21 am --- --- Quote from: RoGeorge on June 06, 2020, 09:24:14 pm --- ;D --- End quote --- I for one am not booing you, RoGeorge; I am only saying it is not an intrinsic property of all capacitors (and because of wording, that makes me disagree with your statement/poll). (This is also at the core of my recent one-sided arguments about learning things, even if those feel like rants or useless whining to others.) Let me repeat an example I've already mentioned recently elsewhere in this forum. At least here, in school, kids are taught that electrons orbit atomic nuclei. The core problem is the definition of "orbit". (In physics and chemistry, they're not called "orbits", they're called "orbitals", "orbital shells", and so on.) Yes, the electron does have properties that are analogous to rotating around a center: angular momentum, something like an orbital radius, and even spin (analogous to "direction of rotation"). But the fact is that those electrons aren't actually moving, they're delocalized around the nucleus. This delocalization can be described using quantum mechanics, so well that the best models we have (most accurate with respect to real-world measurable properties) of how molecules behave and react, are based on the quantum mechanical modeling of just the outermost interacting electrons! No fitting to real-world data, just pure math and some physical constants, and out pops our best predictions of what certain molecules are like, and what their properties are. What does it matter, then? The orbit model gives an intuitive grasp of the properties – radius and angular momentum in particular; and even spin! – and those are what even physicists and chemists work with, so there's no harm, right? Wrong. When a charged particle like an electron is deflected, accelerated, or decelerated by another charged particle, like an atomic nucleus or another electron, it radiates energy. We call this phenomenon Brehmsstrahlung, or "braking radiation". It happens for all charged particles, and is fundamentally due to conservation of energy. So, it is fundamental to charged particles, including electrons. The problem occurs when one tries to integrate the two, Brehmsstrahlung and electron orbits. I fear that most people simply decide that somehow an electron orbiting an atomic nucleus, or flitting about in a metal lattice, is just an exception to Brehmsstrahlung, give a quick :-//, and ignore the dichotomy. That is magical thinking, and leads to the inability of correctly choosing which "model" applies in which situation – so those people can recite information, but not apply it in real world to solve a problem or predict or estimate physical phenomena. But, there is no dichotomy, only an incorrect description and thus incorrect understanding. It would have been not that hard to explain correctly in the first place: delocalization looks like the electron is "blurred" across its "orbital", which looks more like a cloud than a ring; but if you measure it, it is like putting a stick into the spokes of a fast-turning bicycle wheel: you'll see one quite clearly. The properties these delocalized electrons have, are almost exactly like if the electron was orbiting the nucleus like a moon orbiting a planet, and that's why we use that analog; but the actual physical description is much weirder, and involves quantum mechanics. Trying to fix the mislearning later requires "un-learning" – or at minimum, accepting you were taught an incorrect thing because someone thought that teaching the actual thing was too hard. Now, in the capacitor case, the staircase effect is not due to any intrinsic property of capacitors, but due to capacitors being part of an electrical circuit. You can extend this into the capacitors themselves, if you model a long skinny capacitor charged or discharged at one end, in which case the capacitor itself becomes part of the circuit – transmission line, really. The model breaks down if you have either very short transmission lines, or if the transmission "line" does not have a clearly defined length, so that the change in the electromagnetic field (in the transmission line) does not have a clear wavefront anymore. I suppose the reason for emphasizing the transmission line model is to break through some preconceptions the students might have, and make sure they understand that the transmission line model applies to even circuits on a circuit board – even if in practice it only matters enough to worry about when delivering power via relatively long transmission lines. That it is not something that applies to, well, actual transmission lines, but is a model that accurately describes what happens in circuits. The laws that we use every day to describe electrical circuits or networks – Kirchhoff's laws, Ohm's law, Norton's theorem, Thévenin's theorem – are models that do describe the behaviour of the circuit or network, but themselves are the result of more fundamental descriptions of charges; they describe the behaviour, not what is nor why. The underlying reality is much, much weirder, but these models give us tools to work with such systems without getting too bogged down into the weirdness. That's basically what my own field, molecular dynamics simulations with classical potential models is, "classical" meaning just "non-quantum-mechanic". We can do QM for maybe a system with a thousand electrons (but that may take a week to a month, depending on how large a computing cluster you can grab), but that's just a pretty small molecule – and we need repeated boundary conditions, too, which can be an issue if we want surface phenomena. With less reliable/accurate but much simpler classical interaction models we can model systems with hundreds of millions to billions of atoms, getting into things like corrosion and damage resistance, defect migration in the large scale (and self-healing materials), ion implantation, sputtering, and so on, and get very useful real-world results. --- End quote --- Would love to chat on each paragraph, but I'm a slow typer. - I know nobody is booing, and I know that is not some fundamental capacitor's property, just steering the hype - about your rant regarding the educational style, maybe it's OK to make things easy to comprehend, at first, then unlearn that later in order to achieve greater levels of detail, simply because one can not just suddenly jump to the most complicated description of reality - even if the jump to the most complex explanation we have for now would be comprehensible, never forget that there is no ultimate explanation. Don't fall for "my theory can make so many correct predictions, therefore it must be the correct one". Remember the Greek legends about seasons changing? Those myths were able to predict seasons, and when to start working the land. Does that makes them true? No. Another example: the religious ideas about the Earth being the center of the Universe apparently was good enough to predict some planets trajectories, and many other religious/social traits. Those ideas look pretty barbaric now, isn't it? So don't count of "predictability implies the ultimate truth". After all, prophets are still dead-wrong in their belief, even if their prediction became true. - let's set aside that capacitor transmission line, you want to talk about the simulation you are developing. Put you Viking helmet on (I've just read that Finns are not Vikings :-\ ) because my respect for most of the quantum theory (and other latest century physics) might be pretty low, I mean far, far away from the quantum praising we see in the media nowadays. Won't go into it, because having a talk around such ideas will require to "unlearn" as you said, or at least to put aside for a while a lot of ideas we assume unconditionally correct right now just because "they can predict season changes and correct months for seeding and harvesting" as in the Greek mythology about Persephone. - OK then, but do you have some other better theory to replace what we have now? Nope. I'm just generally suspicious against any claims or theories, and since I don't have something better to completely replace the shoddy ones we have now, instead of trying to demolish what others are doing right now, let's try to build on what we have: Can your software run a simulation where electrons are replaced by muons? Asking because if we replace electrons with muons, then the same matter will become much heavier. If it's heavier, then it will move slower (for a given temperature). If it moves slower, then it will stay for longer in the proximity of another piece of matter, or atom/molecule whatever. If it can stay for long enough and close enough, some tunneling phenomena will be expected to be seen, therefore "tunneling cold fusion", and that'll be your Nobel prize. ;D (Just to be clear, not my original idea.) |
| Navigation |
| Message Index |
| Next page |
| Previous page |