0
$\begingroup$

By frequency, I don't mean electromagnetic frequency, I mean modulation frequency. For example, I can put a PWM square wave on an LED at some frequency, which I can detect with a photodiode and a Fourier transform. Would I see the same thing with sunlight?

It's my understanding that no, sunlight is essentially a DC source. Is it? If not, what frequencies would one observe if they were to look at the Fourier transform of a sun-exposed photodiode's time series?

Some edits for clarification: suppose I put a $10 silicon photodiode outside on a sunny day and hooked up an A/D system that samples at 64kHz and gives me a spectrum from an FFT every second. Assume normal solar activity, no clouds, no passing aircraft, and that the sun doesn't go behind trees to create shade. Assume that the photodiode signal doesn't clip, either. Would I see any regular, consistent features in the FFT?

Improve this question
$\endgroup$
8
  • 1
    $\begingroup$ there is a 11-year period sunspot cycle, and there are also much longer ones, see en.wikipedia.org/wiki/Solar_cycle but for an EE it is a noise source $\endgroup$
    – hyportnex
    Commented Apr 1 at 21:56
  • 1
    $\begingroup$ Are you looking at the entire face of the Sun, or are you looking with enough resolution to choose feature which affect local brightness, like the convection cells in the photosphere? Do you count the daily modulation mediated by the Earth and its atmosphere? Do you count the 11 year sunspot cycle? $\endgroup$
    – rob ♦
    Commented Apr 1 at 21:57
  • $\begingroup$ @hyportnex I'm interested in this from an EE perspective. What sort of noise source? $\endgroup$
    – Ben S.
    Commented Apr 1 at 22:00
  • $\begingroup$ @rob Just...sunlight. Writ large. I'm wondering what sort of spectrum it might have given the question exactly as written - just a photodiode sitting outside on a summer day. Will it have discernable features in the spectrum on the order of, say, kilohertz? Megahertz? $\endgroup$
    – Ben S.
    Commented Apr 1 at 22:03
  • $\begingroup$ a very hot thermal noise source source ~6,000K to 11,000K, ... it emits the strongest in the visible region, en.wikipedia.org/wiki/Sunlight, but if you do not have good enough sidelobe suppression it can blind your satellite receiver, too. There are times when the sun is right behind the satellite and these are times of total outage. $\endgroup$
    – hyportnex
    Commented Apr 1 at 22:15

2 Answers 2

Reset to default
1
$\begingroup$

If you look at a "steady" incoherent radiation source with a photon detector, what you see is shot noise. Photons arrive randomly at the detector. Shot noise has a white power spectrum, but if you have sufficiently fine time resolution its amplitude distribution will be seen to be Poisson, not Gaussian.

But the Sun is very bright, so a small photodiode could still deliver a billion electrons per microsecond from sunlight. The standard deviation of Poisson samples is the square root of the expectation value: $\sqrt{10^9}$ is a small fraction of a billion, so the observed noise would only be a small fluctuation in an experiment with microsecond "bins".

For fainter light, the deviation is relatively larger. Shot noise is often the limiting factor in astronomical photometry.

Improve this answer
$\endgroup$
0
$\begingroup$

The diameter of the Sun is

$$ 2\times\frac{\rm 0.7\times10^9\ m}{\rm 3\times 10^8\ m/s} \approx 4\,\text{light-seconds} $$

If there were some process on the Sun that caused kilohertz or megahertz modulations in the light emitted by the photosphere, those modulations would be impossible to synchronize over the Sun's entire face. If you want some process at one point on the Sun's surface to communicate with its antipode, you're restricted by the speed of causality to communication over a scale of several seconds. Even if there were some central synchronizing phenomenon so that the brightness of the Sun's entire surface were to modulate in sync, the effect from Earth would be diluted, since light from the Sun's limb is about two seconds older than light from the center of the Sun's face.

Any fast variation in the Sun's total brightness, then, is going to come from smaller features on its surface. The photosphere's granular convection cells are about $10^3$ kilometers across, and have fluid-flow lifetimes of minutes or hours. Faster modulations can only take place in smaller features within and above the photosphere, such as prominences. The light from these features makes up a small fraction of sunlight (except perhaps during solar flares), but its variation will mostly be stochastic noise, rather than any periodic signal.

Improve this answer
$\endgroup$
0

Not the answer you're looking for? Browse other questions tagged or ask your own question.