A new flat (2mm) lens that can magnify light

[PB666]

http://www.bbc.com/news/science-environment-36438686

 

[Kerbart]

They'll have to rush it though. Other revolutionary technology was announced a couple of years ago that did the same thing - flat optics at mass-produced cost that would make existing optics obsolete.

Come to think of it... haven't heard about it lately. I'm wondering if this one will fare better.

[PB666]

The potential here is to make solar panels lighter by reducing the amount of panel and replacing it with something less heavy. Not much of a potential, but may become very useful for deep space craft which need concentrated solar power to function.

[Nuke]

were just getting better and better at making antennas smaller. we keep getting better at nanoscale manufacturing. kind of a spin off of the semiconductor revolution. of course we are still in the process of finding useful applications for this technology. then of course it will take years to get these technologies to the point where you can mass produce them. its one thing to make this lens in the lab, and quite another to make a million unit run. but its nice to know that camera technology will get smaller than it already is.

Edited June 3, 2016 by Nuke

[PB666]

1 hour ago, Nuke said:

were just getting better and better at making antennas smaller. we keep getting better at nanoscale manufacturing. kind of a spin off of the semiconductor revolution. of course we are still in the process of finding useful applications for this technology. then of course it will take years to get these technologies to the point where you can mass produce them. its one thing to make this lens in the lab, and quite another to make a million unit run. but its nice to know that camera technology will get smaller than it already is.

Thats point i made last year to these neophytes, you get the science, then years pass and you get the technology, and years pass and you get the application engineering, and finally you have a functioning object. the time it takes now from step 2 to step 4 with modern age production technologies. 

[wumpus]

1 hour ago, PB666 said:

Thats point i made last year to these neophytes, you get the science, then years pass and you get the technology, and years pass and you get the application engineering, and finally you have a functioning object. the time it takes now from step 2 to step 4 with modern age production technologies. 

"Once you have the foundry - you want a 12-inch lens? Feel free, you can make a 12-inch lens. There's no limit."  If you are using similar means to silicon chips, a "1 inch" lens is somewhat over the limit.  I think 400~500mm^2 is a bit closer to how big you can make silicon chips (wafer scale was a "just around the corner tech" that never happened*).  If your interest is in focusing light for a solar array, I'm not so sure that you couldn't grind a solar array down to under 2mm.  It would also be a lot easier to cool your un-lensed solar array, and you wouldn't need twice the packaging array.  Note that while warping an array of these devices (for terrestrial applications) sounds interesting, warping somewhat smaller (but equally thin) mirrors would work just as well.

I suspect that it would be used in ever thinner phones and "spy devices".  But there are other weirder techs out there replacing lenses as well.

* there may be exceptions for the CCDs in telescopes.  But expect an astronomical cost such that the "12 inch lens" is a measurable percentage of the cost of the telescope facility.

[PB666]

18 minutes ago, wumpus said:

"Once you have the foundry - you want a 12-inch lens? Feel free, you can make a 12-inch lens. There's no limit."  If you are using similar means to silicon chips, a "1 inch" lens is somewhat over the limit.  I think 400~500mm^2 is a bit closer to how big you can make silicon chips (wafer scale was a "just around the corner tech" that never happened*).  If your interest is in focusing light for a solar array, I'm not so sure that you couldn't grind a solar array down to under 2mm.  It would also be a lot easier to cool your un-lensed solar array, and you wouldn't need twice the packaging array.  Note that while warping an array of these devices (for terrestrial applications) sounds interesting, warping somewhat smaller (but equally thin) mirrors would work just as well.

I suspect that it would be used in ever thinner phones and "spy devices".  But there are other weirder techs out there replacing lenses as well.

* there may be exceptions for the CCDs in telescopes.  But expect an astronomical cost such that the "12 inch lens" is a measurable percentage of the cost of the telescope facility.

You are placing unwarranted design limitations on it, the object can be built to focus a foot away if that is what you want. It might also focus light onto a line an inch wide covered with an inch wide solar panel 10 foot long panel. This is not like a lens for a camera, it uses a series of small metal spikes to cause interference of the light.

Edited June 3, 2016 by PB666

[wumpus]

1 hour ago, PB666 said:

You are placing unwarranted design limitations on it,

Actually I am more likely assuming that "unwarranted design limitations" assumed by the "2mm" thickness PR are insurmountable.  It certainly appears easier to make a diffraction grating thinner than a solar panel, even if they were basically both made by semiconductor manufacturing techniques.

I'm not sure there is any advantage in using lenses (beyond the atmosphere) at Earth distances.  I know they can certainly help within the atmosphere (I think tropical sunlight gets twice the power of where I live on a sunny day), but have to wonder about other issues (you can stack more layers into the solar cell, but does it weight more. If you use a lens at that distance how is the heating?).  I'm guessing they would help at Mars level, and possibly Jupiter (depends how much the lens will weigh, and how unobtainable plutonium is).  While I wouldn't expect the full effect of the "using a magnifying glass with the Sun" (I'd assume one dimensional magnification at best, but somebody could try to recreate Ivanpah) you would still be concentrating heat onto a semiconductor that is lying in a vacuum.  Not a recipe for a long-lived solar panel (I've heard that the lifespan of the solar panels of the ISS are one of the biggest limits to how long it can last).

[YNM]

15 hours ago, PB666 said:

The potential here is to make solar panels lighter by reducing the amount of panel and replacing it with something less heavy.

Really ? Guess the conservation of Etendue won't allow it. Unless the PV's semiconductor gets better that is.

[PB666]

2 hours ago, YNM said:

Really ? Guess the conservation of Etendue won't allow it. Unless the PV's semiconductor gets better that is.

Not sure what your point is. 

[YNM]

18 hours ago, PB666 said:

Not sure what your point is. 

You can't make the cross-section smaller, which often means lighter.

[PB666]

You can make them  out of light materials, since rhe structure is essentially non crystalline, unlike solar panels you can role the out over great distant and taught them like a rope. 

I should add to this that solar panels need to be replaced at least every 100 years. If you used this system youbwould still have to replace the panels, but only one tenth the mass, so that the mass needed to carry is less. 

Edited June 5, 2016 by PB666

[YNM]

But this tech is a lens. Solar panels needs to make voltage difference out of light, not just bend it around, which is why it's practically a photodiode.

Unless I'm missing something in the news. Is there any arXiv preprint ?

[PB666]

20 minutes ago, YNM said:

But this tech is a lens. Solar panels needs to make voltage difference out of light, not just bend it around, which is why it's practically a photodiode.

Unless I'm missing something in the news. Is there any arXiv preprint ?

Yes, you are missing a new generation of photovoltaics were the light is focused by one device and converted to electricity by another. If you can produce a focusing panel for one third the mass and use a power conversion panel one tenth the mass you effectively increase the power mass efficiency by 3 fold. The orher benefit is that in situations were you need a minimal amount of power to make panels worthwhile, this can effectively make an underlit panel sufficiently lit. 

I have to repeat this again because there is a historic tendency in this group to misuderstand these technologies. Nothing like this will reach space, shocking as that may seem. Every technology needs to be reworked and reapplied to reach space and in the process some things will rise faster than others. 

What I found was interesting in designing the ultimate SEP is that if you really want to have alot of power you need panels that roll out, panels that go from very compact spaces to completely fillng the maximum 2D space that the structural can support. This is not easy to achieve with current technology, particularly with crystalline arrays, and the most efficient arrays are cystalline. 

The other problem is moving electrons. 1000 sq meter panel, alot of the mass goes into moving electrons on average 75 meters at 24V. That is to say you have to move 300KW of power at 24 V. There is therefore a need to concentrate the power generation to motifs that can step up the voltage for commuting electrons. If this can be broken up to panels that generate a stepped up voltage by higher power density in smaller strips then the step up transformers may not be required and the panel weights go down. 

Overwhelmingly in a SEP ship the problem is the energy density/mass. So that the basic problem is to get the voltages up at the source as mass efficient as possible and reduce the panel and the electronics and structural devoted to supporting thos panels. What actually gets this done only the future can telll, but I can see promise in multiple technologies to get there. 

If you don't see where i am going, try building a shell ship for SEP propulsion, alot of the gorey problems become apparent in the attempt to design. At some point there is just know way to launch the shell inside a fairing. This gets into the basic problem of how to rollout something that will inevitably several football fields in size for use as a space tug, will it be worth it if you cant get the mass and structure down. 

[QuesoExplosivo]

I wonder what the chromatic aberration is like on a lens such as this.

[SciMan]

5 hours ago, PB666 said:

The other problem is moving electrons. 1000 sq meter panel, alot of the mass goes into moving electrons on average 75 meters at 24V. That is to say you have to move 300KW of power at 24 V. There is therefore a need to concentrate the power generation to motifs that can step up the voltage for commuting electrons. If this can be broken up to panels that generate a stepped up voltage by higher power density in smaller strips then the step up transformers may not be required and the panel weights go down.

Then don't move the electrons 75m at 24v. Use a higher voltage. 24v is only about 48 PV diode junctions in series (~0.5v per solar cell).

Power dissipated as heat in a wire is P=I²R. Power flow thru a wire is P=I*V. Transmission lines operate at high voltages because this minimizes the power lost as heating in the wires.

Putting more cells in series and less cells in parallel increases output Voltage and reduces output Current, but does not change output Power. However, by moving the same power at higher voltage, the losses in the distribution wiring are reduced.

There is no need for a higher power density in smaller strips. The arrangement of the wiring is all that needs to change (at least in the solar arrays themselves).

 

Of course, if you can get the same power output in LIGHTER strips of solar cells (or higher power output for the same mass), the power to mass ratio of the entire array goes up no matter how you configure the wiring.
That may be caused by reduced losses inside the individual cells, but it does not imply that it would by itself reduce the losses in the wiring. The losses in the wiring are almost entirely a consequence of how the cells are wired.

 

With suitable interconnection between adjacent series arrays ("grid" configuration), the loss of any one single solar cell will not cause its whole string to go offline.
The configuration is essentially "X cells in parallel, repeated Y times in series", instead of the more usual "X cells in series, repeated Y times in parallel.

This does potentially use more wire, but when combined with increasing the voltage of the array by putting more cells in series, I believe the mass increase can be kept reasonable (more, smaller wires ~= less, larger wires).

[PB666]

6 minutes ago, SciMan said:

Then don't move the electrons 75m at 24v. Use a higher voltage. 24v is only about 48 PV diode junctions in series (~0.5v per solar cell).

Power dissipated as heat in a wire is P=I²R. Power flow thru a wire is P=I*V. Transmission lines operate at high voltages because this minimizes the power lost as heating in the wires.

Putting more cells in series and less cells in parallel increases output Voltage and reduces output Current, but does not change output Power. However, by moving the same power at higher voltage, the losses in the distribution wiring are reduced.

There is no need for a higher power density in smaller strips. The arrangement of the wiring is all that needs to change (at least in the solar arrays themselves).

 

Of course, if you can get the same power output in LIGHTER strips of solar cells (or higher power output for the same mass), the power to mass ratio of the entire array goes up no matter how you configure the wiring.
That may be caused by reduced losses inside the individual cells, but it does not imply that it would by itself reduce the losses in the wiring. The losses in the wiring are almost entirely a consequence of how the cells are wired.

 

With suitable interconnection between adjacent series arrays ("grid" configuration), the loss of any one single solar cell will not cause its whole string to go offline.
The configuration is essentially "X cells in parallel, repeated Y times in series", instead of the more usual "X cells in series, repeated Y times in parallel.

This does potentially use more wire, but when combined with increasing the voltage of the array by putting more cells in series, I believe the mass increase can be kept reasonable (more, smaller wires ~= less, larger wires).

Thats only one of the problems, yes you can step up voltage but if you arevin a jovian orbit each panel producing very small amps on each relay. Aside from that how are you going to serialize crystals in very long rollouts? 

 

[SciMan]

Easily. Each support end of the array doesn't have to be an end-point of the series strings.

The strings can be "wrapped" around the supports like wrapping a piece of cord around two poles, so long as both ends of a string terminate at one end.
Obviously I don't mean literally wrapping the solar panels around the supports, but the series circuit for the cells would go back and forth between the two supports in a similar way.

The wiring diagram will end up looking kind of like how a radiator looks, except the radiator tubes get replaced with solar panel strings.

As for "rollouts", that idea could be used successfully if you take it literally.
A roll is one of the most efficient and least complex ways to store a flexible flat material, and very large space solar arrays should probably continue to be mounted to a flat, flexible material even if the individual cells are rigid.
It works kind of like chain-mail, where the individual rings are fairly rigid, but the material formed from them is quite flexible.
Similarly, the individual solar panels are rigid, but the array as a whole is flexible (see ISS solar array wings).

Edited June 5, 2016 by SciMan

[cantab]

No doubt it will be used in the iPhone 9S to make it "OMG paper thin".

[PB666]

1 minute ago, cantab said:

No doubt it will be used in the iPhone 9S to make it "OMG paper thin".

When it gets there we will know, in the meantime who kniws what direction the technology will go. 

[PB666]

 

[YNM]

23 hours ago, PB666 said:

Yes, you are missing a new generation of photovoltaics were the light is focused by one device and converted to electricity by another. If you can produce a focusing panel for one third the mass and use a power conversion panel one tenth the mass you effectively increase the power mass efficiency by 3 fold. The orher benefit is that in situations were you need a minimal amount of power to make panels worthwhile, this can effectively make an underlit panel sufficiently lit. 

I have to repeat this again because there is a historic tendency in this group to misuderstand these technologies. Nothing like this will reach space, shocking as that may seem. Every technology needs to be reworked and reapplied to reach space and in the process some things will rise faster than others. 

What I found was interesting in designing the ultimate SEP is that if you really want to have alot of power you need panels that roll out, panels that go from very compact spaces to completely fillng the maximum 2D space that the structural can support. This is not easy to achieve with current technology, particularly with crystalline arrays, and the most efficient arrays are cystalline. 

The other problem is moving electrons. 1000 sq meter panel, alot of the mass goes into moving electrons on average 75 meters at 24V. That is to say you have to move 300KW of power at 24 V. There is therefore a need to concentrate the power generation to motifs that can step up the voltage for commuting electrons. If this can be broken up to panels that generate a stepped up voltage by higher power density in smaller strips then the step up transformers may not be required and the panel weights go down. 

Overwhelmingly in a SEP ship the problem is the energy density/mass. So that the basic problem is to get the voltages up at the source as mass efficient as possible and reduce the panel and the electronics and structural devoted to supporting thos panels. What actually gets this done only the future can telll, but I can see promise in multiple technologies to get there. 

If you don't see where i am going, try building a shell ship for SEP propulsion, alot of the gorey problems become apparent in the attempt to design. At some point there is just know way to launch the shell inside a fairing. This gets into the basic problem of how to rollout something that will inevitably several football fields in size for use as a space tug, will it be worth it if you cant get the mass and structure down. 

Ah. Now I get it.

That being said, what's the refraction limit for these things ? I mean, if it can make the light from 4 sq mm into a few sq microns then that's crazy...

[PB666]

1 hour ago, YNM said:

Ah. Now I get it.

That being said, what's the refraction limit for these things ? I mean, if it can make the light from 4 sq mm into a few sq microns then that's crazy...

No but what if you had a microchannel photovoltaics that could step up the voltage to say 30,000volts. You have a single buss running down both sides that then roll out on direction.

Photovoltaics will die in space, either due to use or due to cosmic collisions, so keeping the lens aspect, while only storing and replacing the voltaics aspect is one way of saving mass.

I was thinking about concentrating light 10 fold, which most panels can handle quite readily. Such a ship would have operational panels out way past mars, including places like the moons of Jupiter.

[wumpus]

On 6/5/2016 at 11:12 AM, PB666 said:

Every technology needs to be reworked and reapplied to reach space and in the process some things will rise faster than others. 

If you aren't sending it into space and mass isn't an overwhelming factor, what benefit would this have over a simple [molded?] fresnel lens?  UV transmission? UV+transmission plus a limiting factor of area for the device (otherwise I would just use a cheap molded plastic lens).

Note that for terrestrial use, this type of thing either requires a tracking system that positions the lens/mirror in the right place/angle for the array, or trades off huge amounts of area used for efficiency (most of the focused light won't hit the array unless there is an active means to focus it).

To be honest, I can't see why this isn't a current technology (using a lens, regardless of the 2mm variety or not).  I've seen solar panels that don't quite cover an entire roof, and the only reason I can think of why you wouldn't bother focusing the light is that UV transparent lenses are too expensive to bother with.