Questions about resistor noise ???

[jalbers@bsu.edu]

From my reading about electronics, I have been made aware that
resistors are sources for noise in circuits. It is kind of a downer
finding out that a very basic building block (the resistor) will
probably be a source for trouble later on if I was ever to get really
serious about amplifier design. Right now, I am at the crawling
stage. I want to do some of my own experiments and have some very
elementary questions.

1. In a simple DC circuit (stable power source, and carbon resistor
for example), will the resistor create noise? Looking at the fromula
Vrms = SQRT[4KRTB] where B is the bandwidth, since it is DC would the
B=0 making Vrms = 0?

2. I want to do some experiments to try to see noise from a
resistor. I have a function generator and a scope and some basic
components. I am looking for some sort of simple circuits both DC and
AC that will let me play around with resistor noise. I guess my
biggest concern is that the more complicated the circuit is, the more
chances of noise being created somewhere else that I don't know about.

Any help would be greatly appreciated. Thanks

 

[John Larkin]

On Wed, 17 Sep 2008 13:11:06 -0700 (PDT), "jalbers@bsu.edu"
<jalbers@bsu.edu> wrote:

From my reading about electronics, I have been made aware that
resistors are sources for noise in circuits. It is kind of a downer
finding out that a very basic building block (the resistor) will
probably be a source for trouble later on if I was ever to get really
serious about amplifier design. Right now, I am at the crawling
stage. I want to do some of my own experiments and have some very
elementary questions.

1. In a simple DC circuit (stable power source, and carbon resistor
for example), will the resistor create noise? Looking at the fromula
Vrms = SQRT[4KRTB] where B is the bandwidth, since it is DC would the
B=0 making Vrms = 0?

Resistors make Johnson noise, per the formula, with no DC bias ("power
source") present. And yes, if the measurement bandwidth is zero, the
noise is zero. But it will take literally forever to make the
measurement.

"Excess noise" is additional noise that some very bad resistors make
if there's DC across them.

2. I want to do some experiments to try to see noise from a
resistor. I have a function generator and a scope and some basic
components. I am looking for some sort of simple circuits both DC and
AC that will let me play around with resistor noise. I guess my
biggest concern is that the more complicated the circuit is, the more
chances of noise being created somewhere else that I don't know about.

You'll need an amplifier with very low noise, in order to see the
noise generated by a resistor. You can do the math, based on the
Johnson noise equation for a resistor and the noise specs of an opamp.
One gotcha is that opamps have both voltage noise and current noise.

A good low-noise fet opamp could be connected to a fairly high-value
resistor (kilohms maybe) and you could heat/cool the resistor and see
the noise change. 10K ohms in a 10K bandwidth at 500K temperature
makes about 16 nV/root Hz, about 1.6 uV RMS, which are usable numbers.

Or maybe a 50 ohm resistor into an LT1028? Blowtorch?

John

 

[Nemo]

2. I want to do some experiments to try to see noise from a
resistor.

Different types of resistor have different levels of noise. The old
fashioned carbon resistors you mentioned are probably the worst for
noise, which is to say, the best for your experiment. They are still
used in specialist applications where you might get a power surge, as
they can take big pulses of power (overloads) for a short time. The more
modern resistors like metal film, metal glaze (ie most surface mount
ones) are less noisy. I don't know about wire wound resistors as I have
rarely used power resistors. I believe the quietest type commonly
available is the "thin film" type, but as they cost about $1 each they
aren't used much.

Run the resistor warm and with a high current for max noise.

Of course now one of the real experts will say "actually the quietest
type, at least at low frequencies, is made by a Tibetan company that
employs mystics to chew them out of Yak hide, they are used in robots
used to stalk and shear sheep in the Australian outback. Sheep are very
wary creatures so low noise stalking is essential. Now for high
frequencies, you need your gold film resonant cavity devices, formed by
plating a betel nut which has been hollowed out by trained ants..."
--
Nemo

 

[whit3rd]

On Sep 17, 1:11 pm, "jalb...@bsu.edu" <jalb...@bsu.edu> wrote:

From my reading about electronics, I have been made aware that
resistors are sources for noise in circuits.  It is kind of a downer
finding out that a very basic building block (the resistor) will
probably be a source for trouble later on if I was ever to get really
serious about amplifier design.  Right now, I am at the crawling
stage.  I want to do some of my own experiments and have some very
elementary questions.

1.  In a simple DC circuit (stable power source, and carbon resistor
for example), will the resistor create noise?

Yes, of course. But, a CARBON resistor is also slightly microphonic,
and you might have more acoustic pickup than Johnson noise.
More to the point, 'stable' power sources are usually noisy, too.
In fact, just about every item in your field of view that has any kind
of electrical properties, has thermal fluctuations in the electric
field or magnetic field or ... whatever, that become noise when the
circuit model includes them.

There's thermal noise in the sounds you hear (it's dominant in the
hear-the-ocean-in-a-sea-shell case), and it isn't usually important.
Things that are BIGGER than the noise usually dominate.
Be aware of the resistor noise, just in case it IS important some day,
that's all you need to do.