Magic capacitors!
in [Design]
|
From: John Larkin on 7 Aug 2010 15:48 On Sat, 07 Aug 2010 15:30:42 -0400, Phil Hobbs <pcdhSpamMeSenseless(a)electrooptical.net> wrote: >On 8/7/2010 11:17 AM, John Larkin wrote: >> On Sat, 7 Aug 2010 11:52:23 +0100, "markp"<map.nospam(a)f2s.com> wrote: >> ><sniiiiiiip> >>> >>> "markp"<map.nospam(a)f2s.com> wrote in message >>> news:8c3jbcFr8rU1(a)mid.individual.net... >>>> >>> It seems I owe John a BIG apology. Looking back at the posts, I see that >>> whenever he talks about charge being stored in a capacitor he is talking >>> about the convention of what is stored on a capacitor, which is actually >>> +/-q on the plates, related to voltage by the q=CV equation. >>> >>> John actually said in a post to me "We say that a capacitor stores charge, >>> the amount being C*V in coulombs, and it works. My whole point, which has >>> evoked such ranting, is that when you use this convention, be careful about >>> designing using the concept that (this kind of) charge is always conserved." >>> >>> He is right. I assumed the phrase 'this kind of' charge meant a different >>> type of charge that wasn't conserved, he actually meant be careful of using >>> the q=CV charge definition, which is actually +/-q on the plates. It was >>> actually ME who misinterpreted what was being said. The 'charge on a >>> capacitor' by this definition is not conserved. The total net charge is. >>> >>> So I acknowledge John actually really understands this, and I was in the >>> wrong to assume he meant net charge. >> >> Hey, I took two years of college physics, and got As. But as an >> engineer, it's not prudent to say that a capacitor has zero charge >> when it would actually knock you dead if you touched it. >> >> Whatever term a physicist uses for "the differential charge on a >> capacitor" or "the integral of all the current that has ever gone >> through a capacitor" or "the charge on one plate", circuit designers >> just call "charge", which happens to be C*V, in coulombs. I have no >> idea how a typical physicist describes this in everyday English. The >> few physicists I know wouldn't correct me for saying that a 15 pF cap >> charged to 4 volts stores 60 picocoulombs. >> >> Given that this is how EEs design electronics, one must be careful >> about basing conclusions on conservation of C*V. That's all I said. >> >> In the case of the ancient "connecting the capacitors" riddle, the >> explanation almost always includes the phrase "since charge is >> conserved..." and uses C*V as the definition of "charge." It works in >> this circuit. In some circuits it doesn't. Using an inductor, I can >> transfer all the energy from one cap into another of a different C >> value, and C*V will not be conserved. No electrons will be created or >> destroyed. >> >> Don't apologize. > >The occasional apology is a nice change from the Godzilla-Versus-Rodan >style of dialogue usually employed here. (Oh, no! There goes Tokyo!) >The very fact that Mark didn't actually keel over dead might be >reassuring to some of The Usual Suspects. > >Just recognize that we use the word "charge" in a way >> than a physicist might get legal-picky about. (Unless that physicist >> designs circuits, too.) > > >Most of us have actual work to do, and recognize the working vocabulary >of other fields. It's the Fourier transform sign conventions that >really screw up communications. ;) > Hey, Word thinks that rotating a picture 90 degrees is *clockwise* John
From: Grant on 7 Aug 2010 17:10 On Sat, 7 Aug 2010 16:43:38 +0100, "markp" <map.nospam(a)f2s.com> wrote: .... >I'm afraid a real physicist would, as far as their interpretation of what >'electrical charge' is, because to them this would require more charged >particles to be present on the capacitor than were before. If they know you >are using the electrical convention of 'charge stored on a capacitor', where >q=CV and the plates have +/-q on them, then maybe not. What has happened in >reality is you have taken charge (in the form of electrons) from one side of >the plate to the other, via the external circuit, in the process doing work. >The total number of electrons is the same before as it is afterwards. The >net storage of electrical charge in a capacitor is therefore zero. What you >have done, though, is created an electric field between the plates, and it >is the electric field that stores the energy Okay, so what do you call that electric field? In my mind, that's the charge? I accept your argument about the equal but opposite quantity of electrons, but in moving 'charge' over to define that is what loses the point of energy stored, no? > (equal to the work done needed >to move the electrons from one plate to the other in the first place) - the >belt you get is due to that energy discharging (and hence causing a current And here you quite happily talk about 'discharging', you cannot have it both ways? >to flow, which moves the electrons in the plate with the abundance of >electrons back into the plate with the depletions of electrons). > >> >> Given that this is how EEs design electronics, one must be careful >> about basing conclusions on conservation of C*V. That's all I said. >> >> In the case of the ancient "connecting the capacitors" riddle, the >> explanation almost always includes the phrase "since charge is >> conserved..." and uses C*V as the definition of "charge." It works in >> this circuit. In some circuits it doesn't. Using an inductor, I can >> transfer all the energy from one cap into another of a different C >> value, and C*V will not be conserved. No electrons will be created or >> destroyed. >> >> Don't apologize. Just recognize that we use the word "charge" in a way >> than a physicist might get legal-picky about. (Unless that physicist >> designs circuits, too.) > >Well that's my point, a physicist *will* get picky, unless they understand >you are using the electrical convention of q=CV, and they (and you!) >understand that the capacitor has +/-q on its plates. Sure, a capacitor's not going to do much with the other plate isolated, but it does do a little, because a body may carry charge relative to the environment. Grant. > >> >> John >> >
From: Bill Beaty on 8 Aug 2010 23:14 On Aug 7, 12:30 pm, Phil Hobbs <pcdhSpamMeSensel...(a)electrooptical.net> wrote: > On 8/7/2010 11:17 AM, John Larkin wrote: > > > Whatever term a physicist uses for "the differential charge on a > > capacitor" or "the integral of all the current that has ever gone > > through a capacitor" or "the charge on one plate", circuit designers > > just call "charge", which happens to be C*V, in coulombs. I have no > > idea how a typical physicist describes this in everyday English. The > > few physicists I know wouldn't correct me for saying that a 15 pF cap > > charged to 4 volts stores 60 picocoulombs. Physicists make this same mistake, and so do their undergrad textbooks. A few years ago this "capacitors store charge" error was discussed in detail on PHYS-L forum. If it wasn't such a widespread error and barrier to understanding, it wouldn't have attracted any discussion. As for us techs and engineers, in early grades we were all taught power supplies and lowpass filters, where one capacitor plate is grounded and ignored. (Well, that's what happened to me, and that's where I got the misconception.) Or perhaps we started out with Leyden jars, where the outer foil is grasped by a grounded human, and only the inner foil is charged and dangerous. If capacitors only have one terminal, then obviously they can store positive or negative net charge. And therefore a capacitor is charged with coulombs injected into that one terminal. At least in my own case, the misconception was caused by "simplified one-wire capacitors." But later in EE classes I sort of figured it out: charge *always goes through* all two-terminal components. Charge goes through resistors and inductors, but it also goes through capacitors in the form of displacement current in the dielectric which is exactly equal to the currents in both capacitor terminals. Mathematically, capacitors are like resistors, but where the voltage across the capacitor terminals is the time integral of the current plus K. So capacitors store... "integrated current?" Yeah, I guess. But since the path for current in circuitry is always a network of loops, no charge builds up anywhere. It's not charge that gets stored, it's potential energy. So now I imagine capacitors as being mechanical alarm clock windup springs, where the internal spring gets wound by a leather belt passing over the little wheel. If you remove all resistance and let the little wheel spin freely, that's a short circuit. And then I carefully avoid telling kids that "capacitors store leather." > > Just recognize that we use the word "charge" in a way > > > than a physicist might get legal-picky about. (Unless that physicist > > designs circuits, too.) We're also using it in a way that might greatly confuse all literal- minded beginners who are trying like hell to figure out basic circuitry. The problem isn't with nitpickers, the problem is the same one described by CF Bohren: "Lest you think that I am quibbling over minor points of language, I note that in my experience many of the misconceptions people harbor have their origins in imprecise language... Precise language is needed in science, not to please pedants but to avoid absorbing nonsense that will take years, if ever, to purge from our minds." Bohren's own teaching philosophy leans towards misconception-removal rather than just "teaching true facts." Others: "The ill and unfit choice of words wonderfully obstructs the understanding." - Francis Bacon "Many errors, of a truth, consist merely in the application of the wrong names of things." -Spinoza "(language) becomes ugly and inaccurate because our thoughts are foolish, but the slovenliness of our language makes it easier for us to have foolish thoughts." - George Orwell "The search for the MOT JUSTE is not a pedantic fad but a vital necessity. Words are our precision tools. Imprecision engenders ambiguity and hours are wasted in removing verbal misunderstandings before the argument of substance can begin." (from Roget's Thesaurus Webpage) ((((((((((((((((((((((( ( ( (o) ) ) ))))))))))))))))))))))) William J. Beaty Research Engineer beaty a chem washington edu UW Chem Dept, Bagley Hall RM74 billb a eskimo com Box 351700, Seattle, WA 98195-1700 ph206-762-3818 http://staff.washington.edu/wbeaty/
From: John Larkin on 8 Aug 2010 23:33 On Sun, 8 Aug 2010 20:14:28 -0700 (PDT), Bill Beaty <billb(a)eskimo.com> wrote: >On Aug 7, 12:30�pm, Phil Hobbs ><pcdhSpamMeSensel...(a)electrooptical.net> wrote: >> On 8/7/2010 11:17 AM, John Larkin wrote: >> >> > Whatever term a physicist uses for "the differential charge on a >> > capacitor" or "the integral of all the current that has ever gone >> > through a capacitor" or "the charge on one plate", circuit designers >> > just call "charge", which happens to be C*V, in coulombs. I have no >> > idea how a typical physicist describes this in everyday English. The >> > few physicists I know wouldn't correct me for saying that a 15 pF cap >> > charged to 4 volts stores 60 picocoulombs. > > >Physicists make this same mistake, and so do their undergrad >textbooks. A few years ago this "capacitors store charge" error was >discussed in detail on PHYS-L forum. If it wasn't such a widespread >error and barrier to understanding, it wouldn't have attracted any >discussion. > >As for us techs and engineers, in early grades we were all taught >power supplies and lowpass filters, where one capacitor plate is >grounded and ignored. (Well, that's what happened to me, and that's >where I got the misconception.) Or perhaps we started out with Leyden >jars, where the outer foil is grasped by a grounded human, and only >the inner foil is charged and dangerous. If capacitors only have >one terminal, then obviously they can store positive or negative net >charge. And therefore a capacitor is charged with coulombs injected >into that one terminal. At least in my own case, the misconception >was caused by "simplified one-wire capacitors." > >But later in EE classes I sort of figured it out: charge *always goes >through* all two-terminal components. Charge goes through resistors >and inductors, but it also goes through capacitors in the form of >displacement current in the dielectric which is exactly equal to the >currents in both capacitor terminals. Mathematically, capacitors >are like resistors, but where the voltage across the capacitor >terminals is the time integral of the current plus K. > >So capacitors store... "integrated current?" Yeah, I guess. But >since the path for current in circuitry is always a network of loops, >no charge builds up anywhere. It's not charge that gets stored, it's >potential energy. Cool. You must design integrators and linear ramps and timers and power supplies using only capacitor energies in joules. That will make the math much more interesting. Do you think Spice uses joules for its internal representation of capacitor states? No wonder it's so slow. John
From: Bill Beaty on 9 Aug 2010 00:49
On Aug 8, 8:33 pm, John Larkin <jjlar...(a)highNOTlandTHIStechnologyPART.com> wrote: > Cool. You must design integrators and linear ramps and timers and > power supplies using only capacitor energies in joules. That will make > the math much more interesting. Lol. Nah, I just choose to recognize that "time integral of I" is a concept distinct from "charge stored on one capacitor plate." Even though their values are the same, and even though the former is the cause of the latter, they aren't measured the same. Why try to count the excess charges on a capacitor's internal plates when we can just measure the current in the lead wires? Besides endless Newsgroup fights, really this stuff is only important when teaching basic physics/electronics to newbies and when writing electronics textbooks. And further, it's only important if we've decided to avoid filling students' heads with misconceptions like single-wire capacitors or "capacitors store charge." Some educators say things like "what's good enough for me is good enough for them," and so proceed to infect their students with their own muddled thinking. This might work for most purposes, and the misconceptions in question might not act as very large learning barriers. But over decades and generations it's a "game of telephone." It's a recipe for filling textbooks and classrooms with increasing mistakes. Why not instead reverse the trend and turn your students into physicists who happen to specialize in electronics? It's not that hard ...just identify common misconceptions which violate basic physics rules, then avoid spreading those misconceptions to students. "Capacitors store charge" makes no sense to a student with a gut-level understanding of charge conservation. Or conversely, any student who truly believes that capacitors store charge, might forever afterward have troubles with basic physics. Remove the contradiction, and "Aha!" everything suddenly connects together in your brain and makes perfect sense. Ideas like "capacitors store charge" are bad because they prevent the wonderful Aha. That's why they need to be taken seriously as errors, and not just labeled as "nitpicking." > Do you think Spice uses joules for its internal representation of > capacitor states? And you're certain that it doesn't just take the time integral of the current in the capacitor leads? ((((((((((((((((((((((( ( ( (o) ) ) ))))))))))))))))))))))) William J. Beaty Research Engineer beaty a chem washington edu UW Chem Dept, Bagley Hall RM74 billb a eskimo com Box 351700, Seattle, WA 98195-1700 ph206-762-3818 http://staff.washington.edu/wbeaty/ |