Re: how does light cause interference phenomena?

From: Tom Roberts (tjroberts_at_lucent.com)
Date: 09/27/04 

Date: Mon, 27 Sep 2004 03:08:26 GMT

John Kennaugh wrote:
> If you design an oscillator and you want to produce as pure a sinewave
> as possible you make the overall gain = 1.

An "oscillator" with gain=1 won't reliably oscillate at all.

> Any more gain and the amplitude will continue to
> increase until the waveform distorts.

You need to control the amplitude, not the gain.

> If less than 1 it won't oscillate
> at all.

Yes. Which is why a circuit designed for gain=1 won't work reliably,
because components have tolerances....

>> Yes, a filament produces light via a random process: thermionic
>> emission. But all sources of any type of signal are "band limited
>> noise" -- the differences are in the bandwidth, not the principle.
> The output of a crystal (or any) oscillator is not a random process

Sure it is. All electronic oscillators amplify noise in a feedback loop
with high gain and limited bandwidth (determined by the
frequency-determining element). Just think about how such an oscillator
starts up and you'll realize that startup is always from random noise.
Once started, a well-designed oscillator will have the desired signal
dominate the circuit's operation so that out-of-frequency noise has no
chance.

BTW a laser starts up with random noise, too. But only noise within the
bandwidth (and direction) gets amplified.

>>> This would suggest that light from a filament is not in fact random
>>> but in short bursts of coherent light with no fixed phase
>>> relationship between one bust and the next but a fixed relationship
>>> within the burst.
>>
>> That is indeed the case. But individual "bursts" are basically a
>> single photon.
>
> I don't accept that because you can get at least 1000 rings in a
> Newton's ring experiment. Can a single photon extend its influence over
> 1000 wavelengths?

See the discussion on coherence length. Don't attempt to discuss "an
individual photon", discuss coherence length.

>> The coherence length of a common He-Ne laser is usually a few
>> centimeters, which means you can construct a hologram over a volume of
>> about that size. A singlemode laser can have a coherence length of
>> many meters. IIRC a lightbulb has a typical coherence length of a few
>> microns -- long enough to generate Newton's rings in soap bubbles and
>> oil films, but not any larger-scale interference phenomena.
>
> You seem rather confused. Newton's rings have nothing to do with soap
> bubbles or oil films.

Sure they do. On a sunny day one can see splotches of color in an oil
slick or a soap bubble. Those are Newton's rings caused by interference
in the thin film (different colors cancel or reinforce in different
directions in different regions of the variable-thickness thin film).
Newton studied analytically the interference between a sphere and plate,
but the exact same phenomenon applies to oil and bubbles. IIRC that is
what induced Newton to study this in the first place.

> But I am interested in your term "coherence
> length" how is that measured?

Split the light from the source, send the two beams over differing path
lengths, and recombine them and look for interference fringes. The
maximum path length difference that yields good interference fringes is
the coherence length of the source. This is a common specification for
lasers.

Tom Roberts tjroberts@lucent.com

Relevant Pages

  • Re: how good are class D amplifiers?
    ... the way they display noise. ... Better than light bulbs, I think. ... In the case of the light bulbs and the thermistors, the response time is set ... allows you to tailor the settling time of the oscillator more ideally. ...
    (uk.rec.audio)
  • Re: strange squelch acting.....
    ... didn't stop direct radiation. ... Noise radiated from telephone lines which have DSL networking ... Harmonic noise generated by a defective oscillator in an Ethernet ... Other people have reported nasty-noisy signals from oscillations in ...
    (rec.radio.amateur.antenna)
  • Re: Harmonic Osillator and Least Action
    ... of a harmonic oscillator using the principle of least action, ... ] described by a Hamiltonian or Lagrangian framework. ... environment, can be done. ... appearance of the noise Falong with the dissipative term b x'. ...
    (sci.physics.research)
  • Re: Specs for OCXOs (was: Single vs cascading op amp)
    ... stages of a signal rather than use a single stage if the need is to ... amplifies noise plus each stage has a finite bandwidth. ... noise out of the highest voltage zeners (not surprising I assume ... the oscillator? ...
    (sci.electronics.design)
  • Re: Noise figure paradox
    ... My means were "reduce the noise figure of the amplifiers in your front-end" ... For the sake of this thread, it's been just thermal and oscillator nose since ... told that if you take a run-of-the-mill radio to a place like Mt. ... strigent standards than what a 2m/440 HT's oscillator is likely to provide. ...
    (rec.radio.amateur.antenna)