Submitted via IRC for Fnord666
[...] "Normal plastic is transparent for terahertz beams, in a similar way as glass is for visible light," explains Prof. Andrei Pimenov from the Institute of Solid State Physics at TU Wien. "However, terahertz waves slow down a little when they pass through plastic. This means that the crests and troughs of the beam become a little displaced -- we call that phase shifting."
This phase shifting can be used to shape a beam. Exactly the same thing happens -- in a much simpler form -- with an optical lens made of glass: when the lens is thicker in the middle than on the edge, a light beam in the middle spends more time in the glass than another beam that simultaneously hits the edge of the lens. The light beams in the middle are therefore more phase delayed than the light beams on the edge. This is exactly what causes the shape of the beam to change; a wider beam of light can be focussed on a single point.
And yet the possibilities are still far from being exhausted. "We didn't just want to map a wide beam to a point. Our goal was to be able to bring any beam into any shape," says Jan Gosporadic, a PhD student in Andrei Pimenov's team.
Source: https://www.sciencedaily.com/releases/2018/07/180711122332.htm
(Score: 4, Interesting) by SparkyGSX on Thursday July 12 2018, @02:33PM (4 children)
Optical lenses work because of the different refractive index of the media (like air and glass). While this does depend on the speed of light in the media, the thickness of the glass does not matter. This is proven by the fact that fresnel lenses also work.
I'd expect better from a physicist with a PhD. Maybe it was just the journalist who misunderstood, most of the don't really understand anything other than writing.
If you do what you did, you'll get what you got
(Score: 5, Informative) by Immerman on Thursday July 12 2018, @03:44PM
Agreed that the description is ad, but that's pretty much ubiquitous in the field of optics. Even a lot of (especially older) physics books get it wrong. However, the refractive index is just another way of stating (a specific aspect of) the speed of light in the medium:
refractive index n = c/v
where c is the speed of light in vacuum and v is the phase velocity of light in the medium
I would assume the mechanisms being used here are the same as for an optical lens, just applied at a different wavelength. It's all EM radiation after all, it should all behave the same, it's only the relevant materials that change.
And the thickness of the glass absolutely does make a difference for visible-light optics. Not the absolute thickness, but the relative thickness of different parts is what allows the beam to be reshaped. The fact that a Fresnel lens eliminates the (mostly*) irrelevant interior of the lens makes no difference - in fact it appears that they're doing essentially the same thing here.
* the interior of the lens is only mostly irrelevant because, for most situations, the additional delay of the light passing through the thick part of the lens is so brief as to be irrelevant: you're not going to notice that some parts of the beam are a few femtoseconds out of phase with the rest. However, for extremely sensitive applications such as extremely short pulses or where precise phase alignment is necessary, it can become very relevant.
(Score: 3, Informative) by Anonymous Coward on Thursday July 12 2018, @04:11PM
What is refraction, though?
The glass of the lens is thicker in the middle, and thus the interface between air and the glass varies spacially across the surface of the lens.
The result is exactly as described by the physicist via the journalist: The light strikes the lens such that some of the beam hits the interface before another part of the beam that is closer to the edge; this discrepancy causes a discrepancy in the speed of light across the beam, the result of which is that the beam bends; the lens is shaped so that beams which are orthogonal to the plane of the lens are all directed towards the center of the lens.
This phenomenon is called "refraction".
(Score: 1, Interesting) by Anonymous Coward on Thursday July 12 2018, @10:05PM
The thickness of the glass absolutely matters if the thickness differs across the glass because it changes the angle of incidence and that whole Snell's Law thing. The Fresnel lens works because it makes rings where the surface of the rings has a slope that is equal to the local tangent of the surface for the equivalent lens. Once the light ray crosses the surface, than thickness doesn't matter until you hit the back surface.
By the way, describing it as phase and velocity differences is correct, but it isn't the usual way of talking about it in introductory physics class.
(Score: 1) by Ethanol-fueled on Friday July 13 2018, @12:49AM
Here's my gripe. Quoth Mr. Wizard:
Yeah, so do we, before we know what "crest" and "trough" even fucking mean. In Soviet Russia, phase shifts you!
(Score: 5, Informative) by VLM on Thursday July 12 2018, @04:02PM (5 children)
Microwave lens antennas have been a thing for some decades at a freq about a 1000 times lower of course.
They're kinda a moth-to-flame attractant because small dishes are cheap, large dishes are expensive, screwing up your dish feed causing 3dB of loss over a "perfect" feed is the same as throwing out 3dB of dish area which is probably 9 dB of cost at typical scaling. So hyper optimze your feed... leading to horn-feeding-a-lens systems.
So "perfect feed" advocates get excited by making a homemade x-band lens antenna out of aluminum foil, contact cement, and styrofoam sheets in a stack. However... very sensitive to polarization, can't really do circular at all AFAIK, giant lens antennas are even less wield-able than giant dish antennas, and small dishes are cheap enough to "Waste" 3 dB of signal and 9 dB of cost on a bad feed. Manufacturing tolerances of the construction foam make reliable mfgr a PITA and if you have the gear to get a lens design "right" then you have the gear to make a superior dish feed anyway, LOL.
of course the meme will never die in the sense of "you point two horn antennas at 140 GHz or whatever at each other to get it working, then insert homemade microwave lens in the beam path to get 10 dB better SNRs and so forth."
Lens antennas can be light weight, which for portable ham radio microwave stuff is an advantage. But they've never been terribly popular given the scale even worse than dishes with size. Oh well.
(Score: 3, Informative) by Dr Spin on Thursday July 12 2018, @07:16PM (1 child)
Its a shame that the article was not as well written as this post.
Warning: Opening your mouth may invalidate your brain!
(Score: 2, Funny) by Ethanol-fueled on Friday July 13 2018, @01:05AM
Agreed. I cannot tell whether or not either are trolling.
(Score: 1) by Ethanol-fueled on Friday July 13 2018, @12:57AM (2 children)
Only 3dB? That's a really small number to care about, It reminds me of people who used to overclock their processors to the ultrasound-screaming point just to play doom 500ms more quickly.
3dB is such a small number that nobody notices or cares about the drop. Especially when you have a bitchin' stereo that can crank out 120dB no problem. Would you notice if your processor clock was 3Hz slower?
(Score: 3, Informative) by Immerman on Friday July 13 2018, @03:01AM
What are you talking about? A 3db difference equates to a doubling or halving in power. Doubling or halving a quantity is fairly significant in most contexts.
(Score: 2) by VLM on Friday July 13 2018, @01:02PM
Its a big deal if you're doing moon bounce or some stuff like that and your SNR is like 2 dB above the decoding threshold.
Also its much like weak signal VHF work, I'm too lazy to quote the math but based on area and SNR contours the "vast majority" of your contacts doing weak signal work will be at low SNRs. A bigger antenna means more contacts but the ratio of SNRs will be unchanged.
Its a money thing... You pay a lot of money as a ham for a large dish, and the scaling is pretty scary, so buying a smaller dish and a better feed saves a lot of money overall, unless the better feed is made of solid gold or something.
Finally the mobile ham radio guys have a hard criteria for "van people" of 4 feet or whatever fits in the car, so you're competing against a guy with 3dB better SNR because he has a better feed, he's going to smoke you in the scores.
Its a similar issue for land microwave for legacy telcos and whatever... you can buy a more expensive feed that weighs nothing, but if you have a thousand pound dish and want 3 dB more gain thats gonna cost a lot of money in the weight of the dish and the weight and size of the tower, much cheaper to "max out" the quality of the feed, then increase antenna size.
A final issue is sidelobes and interference, over-illuminating a dish or having an awful feed in general is the same as pointing the antenna at a noise source. This is fun with satellites, what is up with this interference?
(Score: 2) by bitstream on Thursday July 12 2018, @08:00PM (2 children)
Now we only need of the shelf transmitters and receivers. And some transistors.
Bundled with 3D printing, maybe some fun can be had.
(Score: 1) by Ethanol-fueled on Friday July 13 2018, @01:03AM (1 child)
The easiest way you're gonna play with Terrahertz is to buy a Cisco (or cheap-shit Akamai) server with multi-mode optical connections, some fiber, and perhaps a MEMS attenuator, possibly an amp, and some diffraction gratings. Bonus stage: look into the opposite end of the fiber after full amplification.
(Score: 2) by bitstream on Saturday July 14 2018, @05:31PM
Ie heterodyne operation? But I think that would be pretty inefficient. Better to get the real stuff, ie devices capable of switching in the THz range.
Some laser pointers use special crystals to get shorter wavelength. Some bad (dangerous) won't filter the unwanted wavelengths out.