Re: [Phys-l] Capacitance problem

[John Denker <jsd@av8n.com>]

On 03/29/2008 09:32 AM, Bob Sciamanda wrote:
> I have taken a different approach, concentrating on the CHARGE on each 
> capacitor plate.  For simplicity I assume ideal capacitors (no fringing), of 
> equal value.  Once the EMF is disconnected, the charge +Q is trapped on the 
> upper plate of C1.  For an ideal capacitor, a charge -Q will be attracted 
> onto the lower plate of C1 from wherever it is available.  Grounding this 
> lower plate does not affect this charge value, and the potential across C1 
> is still [unchanged] ....

A) Yes, that is essentially the definition of what we mean by "capacitor", 
i.e. it is the canonical derivation that justifies treating a capacitor
as a two-terminal black box that behaves according to the equation

   ΔV = Q / C                [1]


B) In engineering, it is not customary to re-do the physics each time, 
but rather to do the physics once, encapsulate it in equation [1],
and then rely on equation [1] for day-to-day use.


C) On special occasions we open the black box and look at the physics
again.  One example is the Thompson-Lampard calculable capacitor,
which is IMHO a brilliant piece of physics.

I get 1000 hits from
  http://www.google.com/search?q=thompson+lampard+capacitor
but I don't know of any good tutorials.  The best I've seen
are
  http://www.ecse.monash.edu.au/museum/dglcap.html
  http://www.ecse.monash.edu.au/museum/DGLCAP1.jpg
and of course
  http://www.nature.com/nature/journal/v177/n4515/abs/177888a0.html

This idea has been around for 50 years now.  Maybe in another
50 years it will make it into the textbooks.