Young's 2-slit experiment for RF/microwave and its implications

You would see mainly background noise from the 4K microwave background.

Easy - as it is for moderate infrared photons too. Royal Holloway London are aiming for something quantum based good for 1-20GHz at present. I am not sure what the current state of the art is now.

It begs the question what is the longest wavelength for which individual photons can be detected using cost no object detectors.

There are materials that become nonlinear in their optical response (the index of refraction is variant with electric field strength), that can mix optical and radio frequencies. Typically, to get a high field, one puts them inside a laser-driven resonant cavity. A resonant cavity for a 700 MHz transmitter means a BIG cavity full of the nonlinear stuff, but it'd work.

A search on "KDP" and "nonlinear optics" would pull up lots of examples...

I know basically nothing about this, but I recall having seen something involving Rydberg atoms in RF cavities.

Jeroen Belleman

Thanks to both you and whit3rd for the answers. I don't have a pressing need to make or use one, I was just wondering if that part existed and what it was called.

Matt Roberds

You can just modulate the diode laser current at RF. (People have done this up into the GHz range.) Besides the obvious amplitude modulation the current also changes the frequency.. .that's the easiest way to put side bands on a laser.

George H.

As someone said, "where is the beef?"

Do the math. The energy of a ~MHz photon is down in the ueV, orders of magnitude below thermal noise, including black body radiation (which is pervasive at those frequencies -- from any source, from cosmic background to room ambient), let alone physical sources (like the incessant din of lightning bouncing around within the ionosphere, which plagues SW and MW communication).

So in the same way that room-temperature things tend to not behave quantum-like, or that large ensembles tend not to behave quantum-like, so too are the EM fields consisting of large ensembles, sufficiently peppered with noise and scattering, that you can't really tell.

Superconducting cryo experiments still produce evidence of it, e.g. Josephson junctions.

Tim

So, it is MICRO eV? Then that makes it WORSE..makes communications more impossible, say from San Francisco to the OTHER Brisbane.

What's worse? Obviously it's still possible, given suitable propagation means (SW trapped by ionosphere, VHF+ repeated via ground station or satellite). Quantum encryption isn't going to be possible, but so what.

Tim

Isn't the theory that it takes energy to transmit a signal? Micro and Nano energy is WORSE than One or so eV; seems impossible to interact with any kind of receiver, even one at the "local" Brisbane. Do not confuse this with stupid FACTs of what one has seen. Seems theory virtually guarantees the mentioned impossibility, and your MICRO eV just compounds the situation.

Free-space optical communication is far harder than radio, for a number of reasons that mostly involve having to detect photons incoherently rather than collectively as in an antenna.

For instance:

1. The maximum throughput (area times projected solid angle) for a one-port antenna is lambda**2/2. That makes it progressively harder to get an adequate amount of signal as the wavelength decreases.

1. The situation is only that good for coherent detection, e.g. an optical interferometer or superhet radio. Most optical systems are incoherent, i.e. they have square-law detectors. That makes the SNR quadratic in the detected signal power, which is pretty heartbreaking at small signal levels. (The distance dependence of SNR in optics is more like that in radar.)

2. Solar background is much much stronger at optical wavelengths than at radio. (At HF and below, the atmosphere is pretty noisy.) The solar background is detected incoherently, i.e. in a gigantic bandwidth. In the visible, a narrow interference filter (1 nm FWHM ) has a temporal bandwidth of a terahertz. Zenith sunlight is about 1 kW/m**2, and the average photon energy is about 2.5 eV, so you're effectively looking at the shot noise of a current of 400 amps per square metre, i.e. 11 nA/sqrt(Hz) for a 1-m intercepted area. Try that one on for size--in a 3-kHz voice bandwidth at 50 ohms, that's 31 uV of shot noise with only a square metre of intercepted area.

A two-element antenna at 20 metres has an intercepted area of about 100 m**2, figuring a beam width of a steradian and 50% efficiency. In the analogous optical case that would get you 310 uV rms noise in a SSB phone channel. The atmosphere's not as noisy as that, even at HF.

1. The shot noise of the background light is generally the limiting factor in outdoor measurement. This limit doesn't exist at RF.

1. The shot noise of the signal itself limits the SNR to

SNR_max = N/(2B)

where N is the photon arrival rate and B is the video bandwidth. The photon energy goes as 1/lambda, so for a constant power level, this is

And then there are circuit details and so on that I won't belabour.

So radio is dramatically easier than optics for free space communications.

Cheers

Phil Hobbs

Well, yes... over a certain amount of time.

If you can wait forever, information theory says you can do it in arbitrarily little energy (potentially zero, for truly infinite time).

You also seem to be forgetting that you're not going it a photon at a time... lasers get quite good SNR when they do their stuff with stupidly large numbers of photons (like 10^18). (And that's with incoherent detectors, as Phil notes.) Can you calculate how many photons/sec (coherent photons, at that, and over a wide cross sectional area, also as Phil notes) are contained in an wavefront of 1W continuous radiated power at 100MHz?

So if your goal is to save power, one would presume arbitrarily low energy photons (assuming you can tease them out of the noise, since Shannon doesn't simply go away, even if you ignore it) would be a rather optimal choice!

:-)

Tim