Natural Science Forum / Physics / General Physics / July 2008

Actually no he doesn't. He wants to acquaint the students with the
principles of a potentiometer.  If he wanted them to learn to make
precision measurements of voltage he'd get them modern high precision
digital voltmeters!

I already said what I'd use.  Note the principle of a potentiometer.
A unknown source is measured by comparison with a variable precisely
known voltage.  The two sources are bucked against each other through
a sensitive detector (meter). When the two voltages are exactly equal
there is no current through the detector and the unknown voltage can
be read off of the calibrated source.  Why is this device useful?  Why
would it be better than some high precision digital voltmeter?  All
things students need to learn. The answer, of course, is that because
the voltage is measured at a point where no current flows, the
internal impedance of the unknown has no voltage drop (V = IR) and
provides no error to the measurement. A digital voltmeter, on the
other hand may have 10 or even 100 megs input impedance but how much
error that will produce depends on the internal impedance of the
source (students can calculate!).

The problem then is generating the known variable voltage. The old
school way is a voltage divider run by a big battery. That can be
switched precision resistors or the "slide wire" which is a way
convert the % division of a divider into distance which can then be
measured with some precision.  Note that while this source doesn't
need to supply any current the divider itself takes current to run. So
how does one produce a stable source of variable voltage? Well again
"old school" is to use the divider to measure a KNOWN voltage and then
ADJUST the current in the divider until when the divider is set to the
known value, the detector reads zero. In other words you CALIBRATE
it.  So where does that "known" voltage come from? Well Harold used to
use Weston standard cells. But these are so "old school" you can't get
them anymore!  So what to do?

You suggest a zener diode. Very BAD idea. Zener diodes are not very
precision voltage sources even in the "precision" ones! They are
temperature sensitive and current sensitive as well. Sure good enough
for some computer power supply reference, but we are talking
"precision" measurements here. So what do we do? Well, OK. We are
talking semantics here because what we do is take your precision zener
diode and then we surround it with all manner of regulation and
compensation. We make sure the current through it is never changing to
some high precision. We compensate for it's temperature coefficients.
We extract the reference voltage off it with a ultra high impedance
follower and then we have to make sure that "follower" itself does not
introduce gain changes with temperature and supply voltages. And if we
want REAL precision we put the whole damn thing in a precision oven
that keeps the temperature constant to a faction of a degree. And on
and on.

So what did I suggest? A voltage reference IC! And what is that? Well
it's your zener diode on a chip with a huge circuit around it that
does all the stuff I just talked about. They are cheap.  And they work
accurately enough for the student experiments. It's what is inside the
digital voltmeter to give it ITS reference voltage!  Of course if you
REALLY want to get serious about voltage measurements then you need
something like the Fluke instrument where they've tracked down ALL
sources of error to the nth decimal place.  But that is hardly needed
to show students what a potentiometer is all about.

My other suggestion was the mercury battery.  They will work pretty
well because they are reasonably stable when new, they can withstand
current being drawn (which standard cells don't like) and can be
individually measured for open circuit voltage before the experiments
and will pretty much retain that voltage unless abused.  No they
aren't "real" standard cells, and have a slightly higher voltage but
they aren't bad. They are reasonably stable with temperature in a room
temperature environment. Harold would have to judge if they are good
enough for his purposes.  They are a "poor man's" standard cell.  But
the ICs are better as batteries need replaced and cost money.  But
they are quick and dirty answers if your standard cells are all dead.
Note that you need a precision voltmeter to measure these batteries
before use. In a way this is sort of like the calibration of the
potentiometer in that we are relying on the stability of the batteries
rather than the absolute accuracy of the output as in a standard cell.
(Though they are calibrated as well)

Personally, if this were my class I'd have some voltage reference IC
supplies made on plexiglas so you could see the circuit. I'd give the
students the IC data sheet to acquaint them with "modern" voltage
references. I'd show them the old Standard cells for a historical
perspective and if the cells still more or less worked I'd have them
measure their voltage just for grins. And this is all added onto the
"potentiometer" concept. Yeah. make it ALL a learning experience. But
hey, it's not my class, and it's not my job. I've already said too
much here.