On Mon, 17 Nov 2003 12:30:05 -0800, Ryan Press <ryan@PRESSLAB.US> wrote:

ryan:

i agree with all you have said.

after re-reading the many posts on this topic, i realized that perhaps the
confusion could be cleared up by noting:

1. the equation R = V/I is the mathematical definition of resistance, and
is valid irrespective of whether a material is ohmic or non-ohmic.

2. ohmic materials, however, obey the equation I = V/R, with R constant;
i.e., ohm's law.

here is a beautiful application of ohm's law.

suppose an incandescent light bulb is rated at 10 W @ 10 VDC and we want
to know the output at 11 VDC. assuming that the filament obey's ohm's law
for current values over at least +/- 10 % from nominal, we know at once
that the current increases by the same factor as voltage, or 10%.
therefore, writing power as VI, the power increases to (1.1)^2 = 1.21 of
its former value, or an increase of 21%. note that because ohm's law was
valid, it was unnecessary to know the (constant) resistance of the
filament.

on the other hand, we know that the resistance of a hot filament may
increase by an order of magnitude or more from its cold value, and over
that range, the filament is obviously not ohmic.

daniel

>This has been discussed ad nauseam on other forums.  Okay, in an attempt
>to curb these OT posts I thought I'd throw my $0.02 in.  When Mr. Ohm
>created his "law" he didn't know about reactances and nonlinearities.  He
>was using metals which had quite a linear response.  So in the strict
>sense Ohm's Law only applies to something with a linear relationship.
>Even then I can't think of anything with a perfect linear relationship,
>hence the quotes around "law".
>
>But I think it is perfectly reasonable to approximate part of a nonlinear
>device using Ohm's Law.  This is very useful and I use it all the time.
>It works great for most everything if you add a couple constraints, say in
>time or in voltage.
>
>So by strict definition I would say Ohm's Law is meant to be used on
>linear devices (simple resistors).  Even then it's not perfect but it
>works pretty darned well.  And I'll continue to use Ohm's Law to
>approximate a nonlinear device and that will work pretty darned well too.
>
>Ryan