That's interesting. Does it imply that Al electrolytics can withstand short bursts of reverse voltage, as long as the average is substantially the correct polarity?
If so, I wonder how long they'd withstand a reverse - seconds, minutes - I suppose it would be stated as volt seconds per farad?
Perhaps in the sense of being partially conductive. But not, I'm pretty sure, in the sense of being able to form a rectifying junction, much less a BJT or FET.
(I wonder if you could make a copper-oxide BJT -- I know it'll make a rectifying junction.)
It's the inexpensive option if you want 45 kV working voltage; heck, the 'lossy' ceramics are excellent at energy storage. The problem is, scaling down to low operating voltage,
stored_energy = V^2 C/2
At low V you need large capacity values which are made using nanostructure objects (supercaps, wet-slug tantalum, aluminum electrolytics and MLC caps). Letting a bit of leakage current build/rebuild your insulating layer is a big reliability and fabrication-cost win. MLC capacitors have to have plate separation of a few microns, with good uniformity, over a LOT of area. The electrolytics just need to generate an oxidized patina on the surfaces to accomplish the same thing.
Hmm, probably just charge. Should be independent of voltage rating, but proportional to capacitance, so, you'd have a ratio of charge to capacitance....... which is a voltage, but not the rated voltage. :o)
The amount of "voltage" necessary to reform the cap, may depend on the rated voltage as well (higher rating = thicker oxide to form / reform / deform). In which case, dividing by that gives a dimensionless "electrolytic reform" voltage gain!
Not counting chemical quirks, like non-stoichiometric compositions leading to uncontrollable doping (AFAIK, it's unreasonable to make N-type ZnO, because of excess oxygen?), and the different kinds of bandgap (direct vs. indirect), and some other things I forget -- it's just a matter of temperature. A semiconductor is any old insulator with a conveniently low bandgap, sufficient to allow a usefully small amount of thermally freed charge carriers.
I want to say, transistors were first discovered in Cu2O, late 19th century. I can't seem to find a reference to it, though. Lillenfeld made some interesting patents in the 20s, but they were poorly understood, and technically impossible, until years later (it took two score until FETs entered commercial use).
Without advanced materials science, and theoretical knowledge to back it up (you say it has to be HOW PURE!?), such observations were lucky coincidences, easily written off as shoddy work.
It's just as well; no existing infrastructure could have used them. ESD/surge from telegraph lines would've quickly toasted a point-contact transistor, if they even lasted long enough (between mechanical creep and atmospheric contamination) to perform any useful amplification in the first place.
Bipolar implies need for medication to achieve a semblance of sanity.
NON-polar, on the other hand,is what you mean. Nonpolar electrolytics have been around since before 1940 and it seems the lifetime (degradation due to heat,etc) has been comparable to polarized electrolytics. At least up into the'80s when i stopped seeing them. Did no repair work on speaker cross-overs,where they were "common". There is no reason that a nonpolar polymer aluminum capacitor would be worse. In fact, i would expect better longevity and reliability. Availability may be rather low due to a small market like speaker cross-overs. But, there is no reason you cannot use the age old trick of back-to-back caps.
When I was in IBM, we used tantalums by the millions (perhaps a thousand per system). The only problems we ran into were reverse insertion but that could be spectacular and occur years later. Our design ground rules called for a voltage de-rating of 2-3 times (no more), not much different that ceramics (though for different reasons;-). We didn't use aluminums. These were 100K hour applications.
if you can control peak current they're fine. The problem is many have thought current was under control and found later due to tant failures that it wasn't.
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.