[remotemass (OP)]

Can a white LED be powered by the same size of it collecting radiation energy? Of course it will depend on how bright is the white led and how much it consumes, but in general can someone show me a bit the math and calculations to answer my question. I want to make a tile 10x10cm that is just white led lights and would be nice it was powered with a small antenna able to convert energy from the electromagnetic radiations into the required electricity? If these tiles were to be put on the walls of a room would it be reasonable they were powered by the surroundings or should I consider something more along the lines of wireless electricity with a magnetic field?

[paul cotter]

Unless you are near a powerful transmitting station the amount of stray rf radiation in your environment will be insufficient for any meaningful power harvesting. There was a prosecution(40-50 years ago, I think) of some chap near a broadcast station in England who had rigged up an aerial and various tuned circuits to "steal" power. 

[Eternal Student]

Hi.

   I'm not entirely sure what you were trying to do.

    I'm guessing you have these LEDs and for some reason you want to avoid connecting them to a power supply with wires (because that would obviously be the easiest way to power them).   Instead you want something like wireless charging (or "Inductive charging") to provide the power for the LEDs.
    I don't see why you couldn't do that,  you could place suitable induction plates right behind each tile on the wall and add a suitable induction charging circuit to the LED tiles.
    To the best of my knowledge, even a small area for the charging circuit will be sufficient if the induction plate generates a magnetic field of sufficient amplitude.  So, yes, I think the LEDs could be powered in this way.   However, this seems harder and more expensive than just connecting the LEDs to a power supply with wires.   You would want induction plates, you wouldn't get enough power from the ambient oscillations in the magnetic field that exist in nature (at least not without having a massive structure to collect and harness all of this, like a huge dish and antenna attached to every tile of LEDs).   [LATE EDITING:  Or being right next to a commercial radio transmission mast, as outlined by @paul cotter  ].

     You could just have a single and massive magnetic field generator in the centre of the room.   However, the magnetic field you'd want to generate would be dangerously large and you might even kill people with it. 
     In the presence of a large oscillating magnetic field, electric currents will be induced in most rings of metal around the room.   This would effectively waste much of the energy as heat but in extreme situations it could be extremely dangerous.  For example, paperclips close to the magnetic field generator might set fire to the paper they are holding together.   Anyone who puts a mobile phone that has inductive charging near the field generator will almost certainly need to buy a new phone.   Any person with metal work inside them to repair broken bones or an electrical pace-maker could be seriously injured while inside the room  etc.

Best Wishes.

[alancalverd]

You could install an MRI  machine in the room and use the leakage RF and gradient fields to power the lights. I can locate one for sale, one lady owner, low mileage, for about £300,000.

[Petrochemicals]

Depends how bright you want your leds and wat you have to power it with. You can power flouresant tubes with power line  fields.

[evan_au]

Can a white LED be powered by the same size of it collecting radiation energy?

White LED panels are typically a blue LED, covered with a yellow phosphor
- The Blue LED (often based on Gallium Nitride) will only generate electricity from blue and ultraviolet photons; these make up a small fraction of the Sun's output
- However, the yellow phosphor in front of the blue LED will convert most blue and ultraviolet light from the Sun into yellow light, which has too long a wavelength to generate energy from a GaN semiconductor
- The electronics driving the white LED is designed to feed power into the LED, and not collect energy from the LED
- So it's not efficient to use a white LED array to collect power from the Sun, and then use it to light up the same white LED panel.
https://en.wikipedia.org/wiki/Light-emitting_diode#Blue_and_ultraviolet

It would be easier to power a white LED with a solar cell
- During the day, open the window
- For a couple of hours in the evening, and a couple of hours in the morning, run the white LED off a battery
- Charge the battery during the day with the solar cell.
- Silicon solar cells will generate electricity from photons with infra-red and visible wavelengths, and collect solar energy with perhaps 10% efficiency.

[SeanB]

I have driven a red LED, using a green LED as power source, very inefficient, and you need either another green or blue LED as excitation for the green LED, and preferably using clear packaged units as well. Efficiency takes a big hit, the LED is not really optimised for use as a photodiode, but does work sort of. You can use this though with a microcontroller to sense ambient light, using a pin configurable as both input and output to swap rapidly between driving the LED, and switching it to an input to the internal ADC, to provide a very poor ambient light source sensor, that can easily be used to control the brightness of a display backlight.

[Bored chemist]

I have an LED torch with lots of ("white", i.e. blue with a phosphor )LEDs in parallel.
If I shine a blue/ violet laser onto one of the LEDs, it lights up by fluorescence. The others also seem to light (dimly) and, as far as I can tell,it's because the LED i'm shining the laser at is acting like a solar cell and providing voltage to the others.

I should set up a better rig to test that idea, but I won't be upset if someone beats me to it. :-)

[evan_au]

The others also seem to light (dimly)

The yellow phosphor powder would be quite good at scattering the yellow and violet light, providing an optical path between illuminated and non-illuminated pixels.
You can test this hypothesis by looking at the spectrum from the other pixels.
- If there is violet present, it is scattered light from your laser
- If there is yellow light present, it is probably fluorescence from scattered violet light and/or scattered yellow light
- It is only if you see blue light from the other pixels that you know there is enough electrical energy from one blue diode to light up other blue diodes.

I find a significant electrical path between illuminated and non-illuminated pixels unlikely:
- Any excess optical energy above the bandgap is lost as heat (eg violet/UV photons striking a blue LED)
- Any voltage less than the bandgap will not produce significant light from the LED (current flow/LED output is an exponential function of voltage above the bandgap)
- There are resistances in the circuit inherent in the semiconductor and wiring, and also load-sharing resistors to even-up power distribution between the LEDs. These will cause a voltage drop between the laser-illuminated LED and any electrically-illuminated LEDs.
- It is best to drive LEDs with a voltage above the bandgap voltage, rather than one slightly below the bandgap voltage.

[Bored chemist]

I find a significant electrical path between illuminated and non-illuminated pixels unlikely:

We are not talking about pixels here.
We are talking about a torch with LEDs wired in parallel arrays.
There are PBC tracks connecting them.

[evan_au]

LEDs wired in parallel arrays. There are PBC tracks connecting them.

I understand that there are electrical tracks between the LEDs.
But an illuminated LED producing slightly less than the bandgap voltage at its terminals will have trouble applying a voltage slightly greater than the same bandgap voltage to another LED through said conductive tracks.

[syhprum]

The problem caused by harveting power from a nearby high broadcasting station is that it causes cross  moulation during the war I anoyed my neighbour by foisting the "voice of America " onto his "Home Service" with my crrystel set.

[Bored chemist]

LEDs wired in parallel arrays. There are PBC tracks connecting them.

I understand that there are electrical tracks between the LEDs.
But an illuminated LED producing slightly less than the bandgap voltage at its terminals will have trouble applying a voltage slightly greater than the same bandgap voltage to another LED through said conductive tracks.

What do you think the current/ voltage curve looks like for an LED?