[Date Prev][Date Next][Thread Prev][Thread Next] - [Date Index][Thread Index][Author Index]

LED as Noise Source

Bob Bruninga wrote:

> On Mon, 27 Nov 2000 Mike73@aol.com wrote:
> >
> > If you're looking to check out your S band receiver, here's a good way to
> > test it...  You will see a nifty little harmonic generator made out of a
> > 10K resistor and 1SS85 diode....
> WHile putting an LED on the roof to indicate that my preamp was powered
> and ON, I wondered about what kind of broadband noise is genereated by a
> forward biased LED less than 1 inch from the input circuit of my preamp.
> As I remember it, a forwared biased diodes makes a good noise generator..?

Which begs the question:  What isn't an RF noise generator???

Remember Johnson noise?  Or the term noise temperature?  The former is often
applied to resistors but a forward biassed diode (reversed biassed as well for
that matter) carries current and there is a voltage drop across it.  Ohms law
says the ratio of the two is resistance.  Therefore the diode has a resistance
albeit at DC.  It also has an AC resistance/impedance, which is more
complicated,  but the V over I DC resistance concept serves as an example of
how to think about components in a different light.  Johnson noise is a
relation that predicts the magnitude of an AC signal across a resistance based
upon the temperature and the bandwidth over which you measure.  The latter,
noise temperature, is routinely addressed in LNAs.  Donít confuse noise
temperature with noise figure.  Noise temperature is a different way of
referring to Johnson noise.  A source is said to have a specific noise
temperature (in degrees Kelvin) given a resistance and bandwidth.  Noise
temperature can be used to express the amount of noise produced by a
semiconductor (transistor, diode, IC... whatever) since a semiconductor or a
circuit, at one point or another is a source or sink with some sort of
impedance.  What is a transistor?  It is a couple of diodes put together in a
fancy way.  NPN, PNP whatever way you slice it you have a couple diode
junctions.  If noise temperature applies to a transistor, which is a couple of
diodes, it will apply to a single diode.

You could look at it the other way.  Forward or reverse biassed Diodes are
sources with some non zero resistance.  They therefore have some sort of noise
temperature or noise power.  An LNA is often made up of transistors, which are
themselves made up of diode junctions so there is a source and an impedance in
there somewhere... presto RF noise source.

Diodes are of coarse square law devices and so generate lots of harmonics.
Depending on the diode type and its packaging it will be more or less broad
banded and undoubtedly emit RF.  As you know an LED is a diode so it will be a
source of RF noise.  What are the range of frequencies produced?  I don't
remember my device physics well enough.  Perhaps more important is how the
inherent noise of a diode junction is coupled into other circuits and how much
noise power is there to be coupled.

The junction, as far as its dimensions are concerned, is probably a poor
radiator.  On the other hand the leads to and from the device may be good
radiators or transmission lines.  The diode is the RF source the leads could
be transmission lines or radiators (antennas) to your sensitive circuits.
LEDs typically have a hefty amount of current going through them and almost
twice the voltage drop as a typical silicon diode, so there is potentially a
fair amount of noise power available to be picked up by sensitive circuitry.

It may have been Zack Lau who wrote about quieting LEDs and diodes in his
column in QEX at one point or maybe an HP or NEC app note... whatever.
Putting ferrite beads on diode leads was one of the suggested therapies.  If
you think of things as distributed components you can find more places for
possible RF noise generation than you can shake a stick at.  If you think of
things as lumped components you can overlook possible noise sources and
paths.  The trick is to use enough of both strategies to solve the problem.  I
keep this in mind whenever I am trying to diagnose a problem.  Another arrow
in your debugging quiver is the one I mentioned before.  Remember to try to
think outside the box.  Find a different way to look at the problem, sink or
source, conductance or resistance, admittance or susceptance,  small signal or
S parameter models, where does the charge go? and above all limits.  Applying
the extremes is a great reality check.

> Bob
> ----
> Via the amsat-bb mailing list at AMSAT.ORG courtesy of AMSAT-NA.
> To unsubscribe, send "unsubscribe amsat-bb" to Majordomo@amsat.org

Bronson Crothers
5764 Saw. Res. Center
University of Maine
Orono  ME  04469-5764

Phone: 207 581 2252
Fax:.....207 581 2255


-... ...-

Email: bronson@eece.maine.edu

Via the amsat-bb mailing list at AMSAT.ORG courtesy of AMSAT-NA.
To unsubscribe, send "unsubscribe amsat-bb" to Majordomo@amsat.org