| Topic: |
Science > Physics |
| User: |
"Dave" |
| Date: |
28 Jul 2005 02:31:12 PM |
| Object: |
There noise - where does it tail off? |
The noise power from a conductor is
Power = k T B
were k is Boltzmann's constant
T absolute temperature
B Bandwidth
Taken literally, that means if you integrate over all frequencies, the
power is infinite. I have seen several "quotes" such as "the noise is
flat to at least 50GHz", and another will say 5000GHz, but have never
seen either
a) A reason it is not flat to an infinite frequency - I guess there is a
limit due to the finite speed of electrons, but I am not sure. Can
anyone explain?
b) Any properly refereed journal where someone has measured it, or
derived something that puts a limit on it.
I just seem to find quotes in books or applications notes about noise,
with nothing to back up what they say.
Anyone got either an explanation of why the noise power falls off at
very high frequencies, and any decent references from peer reviewed
journals?
.
|
|
| User: "Zigoteau" |
|
| Title: Re: There noise - where does it tail off? |
29 Jul 2005 11:34:47 AM |
|
|
Hi, Dave,
The noise power from a conductor is
Power = k T B
were k is Boltzmann's constant
T absolute temperature
B Bandwidth
You've left a couple of things out, but there's the ghost of a true
statement in there.
Taken literally, that means if you integrate over all frequencies, the
power is infinite. I have seen several "quotes" such as "the noise is
flat to at least 50GHz", and another will say 5000GHz, but have never
seen either
a) A reason it is not flat to an infinite frequency - I guess there is a
limit due to the finite speed of electrons, but I am not sure. Can
anyone explain?
For systems at room temperature, it remains that value up to
frequencies of a THz or so, more precisely kT/h. Beyond that frequency
it drops off rapidly to zero. Photons are non-conserved bosons, and you
need the appropriate form of Bose-Einstein statistics.
b) Any properly refereed journal where someone has measured it, or
derived something that puts a limit on it.
It's textbook stuff. Look up Johnson noise in a textbook or on the
internet.
I just seem to find quotes in books or applications notes about noise,
with nothing to back up what they say.
You're reading the wrong books.
Anyone got either an explanation of why the noise power falls off at
very high frequencies, and any decent references from peer reviewed
journals?
How about:
http://www.lpa.ens.fr/mesolpa/publication/julien.pdf
Since there are no transistors yet capable of handling kT/h=6 THz, the
falloff at higher frequencies is academic and will be for a while.
However transistors have been recently reported with fTs of 0.5 THz, so
it might not be much longer before it will be necessary to take these
quantum effects seriously into account.
Cheers,
Zigoteau.
.
|
|
|
|
| User: "Dave" |
|
| Title: Re: There noise - where does it tail off? |
28 Jul 2005 02:59:21 PM |
|
|
you have to back up a few steps in the derivation of that. there is a
simplification step needed to reduce it to that simple form. to get from
the energy per mode in the line which takes the form:
hbar*omega/(e**(hbar*omega/k_B*T)-1) to the
k_B*T*delta_f you have to assume that you are working with circuits in the
'classical limit', meaning they aren't very small... this places a
restriction that hbar*omega<<k_B*T which of course limits the frequency
range over which you can use that equation.
"Dave" <nospam@nowhere.com> wrote in message news:42e93280@212.67.96.135...
The noise power from a conductor is
Power = k T B
were k is Boltzmann's constant
T absolute temperature
B Bandwidth
Taken literally, that means if you integrate over all frequencies, the
power is infinite. I have seen several "quotes" such as "the noise is flat
to at least 50GHz", and another will say 5000GHz, but have never seen
either
a) A reason it is not flat to an infinite frequency - I guess there is a
limit due to the finite speed of electrons, but I am not sure. Can anyone
explain?
b) Any properly refereed journal where someone has measured it, or derived
something that puts a limit on it.
I just seem to find quotes in books or applications notes about noise,
with nothing to back up what they say.
Anyone got either an explanation of why the noise power falls off at very
high frequencies, and any decent references from peer reviewed journals?
.
|
|
|
|
|
Related Articles |
|
|
