your explanations make me feel like a neandertall confronted by homosapiens!
Don't assume you can't do this. It comes from spending hour after hour for years in classrooms learning this stuff. (I was going for an electrical engineering degree back in the days when you had to learn about rotating machinery to get the sheepskin that let you get a job designing a computer. B-) )
a capacitor has two terminals each connected to a large strip of aluminum foil.
two huge surface areas seperated by a layer of paper.
Or a layer of oil(or oil-soaked paper). Or of air. Or one plate is a sheet of aluminum foil, the other is a solution of chemicals, and the insulator between them is a layer of oxide on the foil created by electrochemical acton (the same process as "anodizing").
That last on is an "electrolytic" capacitor. You have to only charge it up a particular direction, because if you charge it the other way the insulation disolves and it shorts out. You can make several electrolyitcs in one can with the negative (solution) terminal connected togeter by putting multiple separate pieces of foil in it. A "non-polarized electrolytic" capacitor is such a device, with two equal plates in the solution, and the terminals connected to the plates (the "positive" terminals). It acts like two electrolytic capacitors in series, and the small asymmetric leakage charges the solution negative so the insulation is maintained.
when a current is applied to one surface it builds up like static electricity untill it reaches a certan level and then it unloads across the paper to the other strip of foil. the same thing happens in reverse when the next part of the ac wave is applied.
More like: If you push an electron in a wire onto one plate, its field repells an electron off the other plate and out the other wire. Push in two, get out two. Push in a billion, get out a billion.
Meanwhile, as you pile up electrons on one plate and atoms lose electrons on the other, you get an electric field pushing back against the current. This shows up as a voltage between the therminals. The more electrons you push in, the harder it fights back with a reverse voltage. (It's like cocking a spring.) The bigger the capacitor, the more electrons it takes to get a given amount of push-back. (A one-farad capacitor will push back with one volt once you've pushed one amp through it for one second.)
(Interestingly, the change in the field between the plates creates exactly the same magnetic field as if the electrons were actually crossing the boundary. So you can't tell if you've got a capacitor or a wire in a box by measuring the magnetic field from the current through it. B-) That has nothing to do with this circuit. But it's useful when you're designing antennas.)
it's able to do this with out a heat build up because of the large surface area.
Actually, it's able to do this without a heat buildup because there's negligible resistance in the wires and plates. As the field is like a spring, resistance is like friction, turning motion (of charge carriers) into heat while slowing them down. A good capacitor has very little resistance - either in the wiring, or from losses as charged particles move around in the dilectric insultion under the influence of the varying electric field.
The more often you reverse the voltage, the more current you drive throught the capacitor. ok It's an approximately linear increase at low frequencies but starts to peter out as the halfcycle period approaches and crosses the time constant
Right. Because a given voltage can only push so many electrons through before the capacitor is fighting back with that voltage and the electrons stop. So reverse it more offen and get more bursts of electrons, for a higher average current. (An amp is a coulumb of electrons per second moving past a particular point in a circuit.)
the idea of a time constant escapes me
The resistance is also fighting the curent. That's like fluid friction: The faster the current, the harder it fights back. After the voltage turns on, initially the current is high (with the voltage drop all across the resistor). As the capacitor charges up, its voltage-drop rises. The total voltage across the capacitor and resistor must equal the applied voltage, which means the voltage across the resistor has to drop, which means the current through the resistor goes down. The result is that the current starts out high and gradually drops, approaching zero as the capacitor approaches a volatge equal to the input. The closer the cap gets to the applied voltage, the slower it charges. It never QUITE gets there, but it quickly gets VERY close. (It's a "decaying exponential". The time constant tells you how fast it decays, i.e. how fast it approaches the limit of no current.)
If your freuqency is low, the capacitor gets essentially fully charged right away. It acts like an electron counter, letting a certain number of electrons through on every cycle. So the more cycles, the more electrons. The current is limited by the capacitor, not the resistor.
As your frequency increases, the capacitor misses getting a full charge by progressively more. The amount of electrons per cycle starts to fall off from the ideal. While the current still goes up with frequency, it starts to fall back from being in direct proportion. The current is limited by the combination of the capacitor and the resistor.
If your frequency is really high, the capacitor hardly gets any charge before it reverses. It never fights significantly, but acts like a short circuit. The current is limited by the resistor.
By telling you how fast the capacitor approaches the limit of charge equal to the applied voltage, the "time constant" also tells you the period (inverse of frequency) where the split between the two effects is equal. You need calculus to derive it, but once you have the formula (1/RC) it's trivial to compute.
the idea of a time constant escapes me,so after that, things get fogg
First year calculus (differentiation and integration) is enough to analyze linear circuits. "Linear" means "sum of reactions to multiple inputs equals sum of reactions to each input individually". (This often implies graphing how things react to inputs makes straight lines, hence "linear".) That means circuits made out of resistors, capacitors, inductors, transformers (as long as the core isn't saturating). (Diodes, transistors, saturable cores, etc. make things 'way complicated to get them dead on. But there are simple linear approximations that let you handle them easily if you're careful not to operate them where the approximations are too far off.)
It's also great for understanding forces in structures, strength of materials, motion of masses, vibration, designing to avoid cogging, and a lot of other stuff related to windmills (like the Betz Limit and other maximization problems.)
If you haven't taken it yet, first-year calculus requires beating your head against a couple of ideas for a couple months. But all of a sudden you "get it" and it's like the scales falling from your eyes. You walk around and see the math everywhere.
If you HAVE taken it, the currents in a linear circuit ARE calculus made real:
- Resistor: Voltage across it is proportional to current through it. Resistance is the proportionality constant.
- Capacitor: Voltage across it is proportional to the integral of current through it. Capacitance is the (inverse of the) proportionality constant.
- Inductor: Voltage across it is proportional to the derivative of current through it. Inductance is the proportionality constant.
Combine with a little matrix algebra for handling complex interconnects and laplace transforms (derivable from first-year calculus) for shortcutting computations, and you have the whole of "analysis of linear circuits" right there in your toolkit.
