Years and years ago I asked my Dad what MFD meant on a capacitor. He explained it meant micro farad. All the capacitors I saw were rated in micro farads or less than 1 micro farad. So I asked my Dad why they were all such small numbers, why none were rated in whole farads. He told me a 1 farad capacitor would be huge, about the size of a chair. Now I see little caps rated at 1 farad. Isn't the capacitance a function of surface area? How are the little caps made to acvhieve such high values? Thanks, Eric
those little caps do not have high voltage. they are low voltage caps using dielectric material that is very thin based. also over the years, different materials have been used to create dense capacitors, the Tantalum is one example.
--
"I am never wrong, once i thought i was, but i was mistaken"
Real Programmers Do things like this.
http://webpages.charter.net/jamie_5
But large value capacitors in general have shrunk.
I can remember getting 10,000uF capacitor at about 18v circa 1974, and it was about the size of a soda can. It was quite the find for the time.
Now you can get them in much smaller packages, with even larger values.
I suspect the improvements that allow more compact large value capacitors is driven by solid state devices. In the old days of tube devices, the only thing you'd need really large values for were timing (and even then "large" is relative) because they were all high impedance and thus could get by with smaller value capacitors. That's when you'd see cartoons about those almost imaginary 1F capacitors. When solid state devices came along, their lower impedance meant that larger values were needed. Nobody would put 10,000uF in a tube power supply, it would be way overkill since a far lesser value would provide enough filtering at the current levels needed (though you certainly needed to worry about enough voltage rating). But with solid state, 10,000uF might even be on the low side. They needed high current at low voltages, which drove the need for larger capacitors for the power supplies and bypass capacitors, which put larger values in smaller packages, and basically gave us values we never would have seen decades ago.
You're right, the "supercaps" are generally for very limited purposes. But I think they must be byproducts of getting 10,000uF capacitors that aren't the size of soda cans.
Capacitance is a function of surface area, dielectric constant and distance between plates.
Using new techniques we can get a lot more effective surface area in a much smaller space thatn before (with the added bonus the places are closer). The dielectric constants for ceramics are also quite high.
That does not mean these things are a panacea; high-k dielectrics have microphonic effects which can make them unsuitable for high end audio (ask Graham and others about this), and their capacitance can vary greatly with temperature *and* applied voltage. There are still places where older styles are optimal for certain things (although it's a while since I saw a polypropylene or mylar device - perhaps because I simply haven't been looking for them). I do remember using oil capacitors many years ago in a magnetron circuit.
So-called supercaps, especially those designed for use as memory backup have very high ESR which makes them unsuitable as filter caps (and many other uses).
So it's a matter of what one needs; ceramic devices are good for decoupling (too good in some circunstances, amusingly) but not necessarily good for other applications.
The classic formula for a capacitor is C = kA/d for a parallel plate device where k = dielectric constant (relative constant x 8.854 E-12), A is the area of a plate and d the distance between them. Dimensions in metres (square metres for area, of course), result in farads.
your dad was right... but new technologies have come on the scene since your dad's day. Some use a very thin layer of a sort of metallic foam dilectric that has a likewise thin layer of metal atomised onto both surfaces. The micro-construction permits (comparatively) huge values in very tiny containers and other construction features - even the shape of the electrodes inside the can - can greatly affective how good the cap is.
The problem is, they aren't very tolerant of misuse - short-circuit them with even a tiny charge and they die, put more than the rated voltage across them and they die, reverse the polarity and they die. I think most are static sensitive too. This is because the construction means the actual "plates" and dilectric are only a few hundred (or less) molecules thick and so will burn through incredibly easily.
Ordinary electrolytics can be a little tolerant of any of the situations above but these "mega-MFD" caps are quite tetchy beasts. They are mostly used for memory retention in MOS circuits so you don't use them as you would a "normal" cap, i.e.supply smoothing or coupling an AC signal.
The tiny currents (maybe only a few peco-Amps) drawn by modern MOS based memories in standby mean that a 1F cap can hold the contents for days if not weeks - obviously this depends on the memory devices and the circuit config. Bearing in mind the frailties outlined above, the power supply to these devices needs careful design to ensure you don't aggravate them - inrush current particularly, must be limited.
I light of Lord Garth's post... I feel I should point out that the technology has obviosly moved on a bit more than I was aware - using a
350F cap instead of a replacement D cell... nice! Obviously that one can expect a bit rougher treatment that the fretful memory retention caps that I have killed :o(
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.