Re: A Simple Lab Demo of Resistor Noise EMF.

["John S. Denker" <jsd@MONMOUTH.COM>]

At 11:52 AM 3/19/01 -0500, Michael Edmiston wrote:
> ... if signals of millivolts are being seen when torching a
> several-hundred-kohm resistor, this cannot be thermal noise.

Right.

> As other have
> stated, it might be junction potentials caused by different temperatures at
> the junctions (i.e. a thermocouple, i.e. Seebeck effect).

That's been suggested a few times now...  That's the very first hypothesis
that came to my mind yesterday, but I immediately rejected it.  That's
because every digital multimeter I've seen in the last umpteen years is
blind to DC on the AC scales, and the thermocouple effects should be nearly
DC, i.e. changing so slowly as to be below the meter's cutoff.  The meter
might respond to thermocouple effects if/when there is a large dT/dt, but
if that's what's going on I would have expected Brian to notice it and
report it.

Maybe radio shack voltmeters are sensitive to DC even on their AC scales.
 -- It is easy to check for this "feature":  in AC mode, put a 1.5v
battery across the leads!
 -- Conversely, if this "feature" is causing problems, there is an easy
remedy:  put a blocking capacitor in series with the hot resistor.  Choose
a capacitance that corresponds to the input impedance of the meter, to give
an RC time of a few milliseconds. (I'm assuming the high end of the meter's
passband is 10kHz or higher).