Re: Internal resistance

[Brian Whatcott <betwys1@SBCGLOBAL.NET>]

At 04:28 PM 4/1/2004, you wrote:
> On Thursday, April 1, 2004, at 04:56 PM, John Denker wrote:
>
> > Ludwik Kowalski wrote:
> >>> ...  r ... was found to decrease
> >>> monotonically from 1.4 ohms, at I=0.85 A,
> >>> to 1.0 ohms at 3.8A. ....  (V1-V2)/I
> >
> > The question is ill-posed.
> >
> > In the key formula, V1 is not measured "at"
> > 3.8A so the quotient (V1-V2)/I cannot be called
> > the resistance "at" (V1-V2)/I.
>
> The value of r was expected to be the same for
> all currents, in the range for which the power
> supply was designed. But it turned out to be
> different for different currents. Nowhere in our
> textbooks was this possibility mentioned.
> Many numerical problems would make no
> sense if r were current-dependent. What is
> wrong with testing this experimentally? What
> is wrong with asking how r depends on I?
> How to explain the observed dependance?
> Ludwik Kowalski


An interesting aspect of these observations,
is the aspect that you certainly  mention to students, but
perhaps didn't recognize here: the implied zero crossing
 of the  voltage drop versus load current graph.
It is not a given that the supply has zero [internal] current
with no output load connected.
To account for this possibility, you could work with two
load resistors.   To establish non-linearity, you would use
three load resistors. Cooling the supply between each
application, to account for temperature effects with current...



Brian Whatcott    Altus OK    Eureka!