Homemade Diodes

[AbbaRue]

I came upon the idea from a website I stumbled upon.
I just built the diodes to see how they work and found them useful.
I haven't tried many other tests with them, but I do remember reading that this
type of diode was used many years ago before SS diodes were invented.
I think they were even used for welding machines at one time.
Perhaps I should do more study on them.

[AbbaRue]

The bigger the aluminum and SS plates are the more capacitance you get,  naturally.
But I found that if the plates are to large the diode effect seems to stop.
Perhaps it just takes to long for enough oxide to form to get the effect working.
The capacitance if quite high for the size.  This is the principle behind electrolytic capacitors of coarse.
The thin oxide layer acts as the dielectric of the capacitor. 
Only electrolytic caps use aluminum hydroxide instead of baking soda and water.
At least the older ones did, maybe the newer ones use something else now.

[aethernut]

The diode doesn't work as a diode until a good coating of oxide forms on the aluminum.
So if you build one you have to slowly build up the voltage you apply so the oxide can build.

It seems that you could get a coating of oxide quickly by dipping the aluminum in a solution of lye.

A related topic is Rust Removal using Electrolysis--see antique-engines.com/electrol.asp  At that site they post a warning about using stainless steel electrodes--Hexavalent chromate can be produced which is poisonous hazardous waste. (The electrolyte turns yellow.)  see antique-engines.com/stainless-steel-electrodes.htm

[juice]

How would you make a homemade transistor? I guess you could manage it with mechanics or with fluids, just as you could manage a diode, but what about something in the spirit of this diode you describe? Is it possible?

[Koen1]

Hmmm... interesting idea...
I guess you could make something like a transistor,
or rather a type of "triode"...
Although I would have to think it over some more,
I would suggest one could add a thrid electrode as
"activator current" source, which could increase or
decrease the charge flow between the diode plates...
That would be a transistor-like function, in a way...
... right?