Is leakage a certainty?

[ViolenceOnTheRadio]

Do all caps and diodes experience a degree of leakage?

[R.G.]

Yes. Silicon in general is 1/1000th of germanium, but all semiconductors leak to some extent.

[PRR]

The universe sucks. Everything. So everything leaks.

If you want a low-low-LOW leakage cap, try special grades of glass. (You might think "vacuum", but you have to support your electrodes somehow.)

Diodes are sure to leak. They are lively little crystals. There is a current available with ZERO voltage across the diode.

In audio, "conventional" caps are low-leakage enough for almost any purpose. Diode leakage is part of the game of making stuff work with real-world parts; with modern Si diodes we may generally overlook diode leakage.

We can even get away with "high" leakage. Electrolytic caps leak 100 times worse than paper/oil or poly-plastic caps, but 100 times less than the working current in many parts of audio circuits.

[stringsthings]

Does anyone manufacture diodes from SiGe?  ( i wonder if they would leak more/less than Si )

http://www.aip.org/tip/INPHFA/vol-8/iss-3/p22.pdf

The universe sucks. Everything. So everything leaks.

[mac]

If you want a low-low-LOW leakage cap, try special grades of glass. (You might think "vacuum", but you have to support your electrodes somehow.)

...and vacuum is just an ideal thought because if you take all the air inside a closed box, the walls are going to evaporate, so no vacuum 

oh, BTW leakage is an "uncertainty" 

mac

[stringsthings]

...and vacuum is just an ideal thought because if you take all the air inside a closed box, the walls are going to evaporate, so no vacuum 

oh, BTW leakage is an "uncertainty" 

mac

was that an attempt at humour?

[Ben N]

To the original question:

Depends! 

[brett]

Hi
original question: how much leakage is important to you?
All transistors eventually fail (even if you don't use them, they won't last a million years).  But I'm not worried about that !  Have I ever worried about the amount that Si diodes leak?  Not that I can remember.

RE:  and vacuum is just an ideal thought because if you take all the air inside a closed box, the walls are going to evaporate, so no vacuum 
no.  The physics says a definite no.  Otherwise rocks in space (including Earth) would "evaporate".
cheers

[alanlan]

Do all caps and diodes experience a degree of leakage?

It's useful to read the datasheets and ask yourself if the amount of leakage in question would pose a problem for your application.

[mac]

brett,
talking about certainty, or uncertainty in this case, as we are dealing with electrons that obey quantum physics, there is always a probability for a particle to pass through a finite potential barrier. please refer to the step potential (in english IIRC). this means that in the example i gave of a closed box, a particle near the surface of the internal wall can leave the surface. much like what happens in our tubes. of course we help electrons to jump between plates by heating on of them. but this happens anyway
at a very slow rate to be noticed by our "classical" eyes.
as Rg said any part leak, and everything evaporates.
when we say "vacuum" tube, RG can help here, there is no such vacuum, only 1/10.000.000 IIRC of atmospheric pressure obtained by a mechanical+diffusion pump. if you divide the Avogadro number by 10.000.000 you still get a large number of particles.
even in deep space you find some H2 mollecules per m3.
the most radical idea that everything evaporates was given by Stephen Hawkind in 1974 who said that a proton has a probability of leaving the extreme forces of a black hole!!! the explanation is based on the uncertainty principle.
coming back to the original post, leakage in a transistor is due to uncertainty. models are obtained by using quantum statistical mechanics.

mac

[nick d]

             STEPHEN HAWKWIND ??
             
               "Space Is Deep" 
                 
                 Do not panic..........

[Ronsonic]

             STEPHEN HAWKWIND ??
             
               "Space Is Deep" 
                 
                 Do not panic..........

Sure Stephen Hawkwind, played a lot of powerful psychedelic / metal music back in the day with lots of lyrics about math and science.