https://www.youtube.com/watch?v=hliHBeC1sco
In this video, a circuit is used to transmit sound over an LED. Whoever made this video said it would be possible to shrink the circuit down to the size of a dot. I see a problem, however. How would such a tiny circuit provide enough power to run the LEDs and at the same time, not get fried? I find it kind of dumb that he didn't try the photodiode microphone with the real LED bulb-just to make everything seem like it could be real.

The light of a white LED consists of the blue light from the semiconductor and orange from the phosphor (that is excited by the blue)

In the semiconductor there is a junction of 2 materials, actually it is the same GaAs but doped with 2 different additives. Additive N have excess electron over the ones allready present in the GaAs, so where ever an N atom is present in place of a GaAs atom in the crystal structure, the extra electron is there and is free to wander around. Additive P lacks an electron, so one of the links in the crystal around the P atom is weakened by lack of an electron, but it still holds

When the LED is connected to DC power - With the N side to Negative and P side to Positive :

 - Electrons go from the Negative wire into the LED, and into the N type material

 - The connection to the Positive wire pulls out electrons from the allready weakened links in the P side. As "free spaces" for electrons open, the electrons firther away from the wire are pulled into them, then the next electrons pull into the new free spaces and so on

So the electrons flow from the Negative side of the power supply into the crystal, and "holes" move from the positive side of the power supply into the crystal

Where they meet (on the boundary between N and P sides of the diode), the free electrons coming from N side fill the holes from P side. This keeps going as long as power is connected, as new electrons and holes keep coming to the boundary from either side. When the electrons "fall" into the holes they drop in energy level, this amount of energy gets out of the crystal as light. The wavelength of the light depends on the "height" the electrons fall, which in turn depend on the materials

When you switch the power to the LED off, no new electrons and holes are coming, but the ones allready present near the boundary will still come together and reconnect. The time it takes is on the order of nanoseconds, ie. In a few nanoseconds from the moment the power is switched off, the LED stops lighting in the blue part of the spectrum

This is way faster than the afterglow extinguishing in materials like phosphors, where you can see by eye the afterglow so it is on the order of 100's mSec

In short, this means that the blue part of the output spectrum of the LED reacts very fast to switching on/off the power to the LED. No other lamp extinguishes that fast when the power comes off...



If you encode information in the switching operation of the ballast (by altering the switching times a bit), this information will be present and transmit in the light. Assume the time for the LED extinguishing is few nSec, and that the ballast can keep up with the LED at such speeds, that means you can encode "1/few" Gbit/sec of information in the light. Probably on the order of a Mbit/sec

Few proposals were made using this effect, like Li-Fi (internet connection over LED lighting)



The surveillance scheme described in the video, if i get it, is the LED lamp picking up some information accessible to it, and transmitting it invisibly in the light. Then anyone aiming a matching optical reception instrument to your window can pick the information



The lamp can pick up any electronic wireless communications, over the mains communications (HomePlug, ...), speech, video, i think thats it. But for that it must have the matching comonents inside : Antenna and wireless receptor, over the mains receptor, microphone, or camera (and in the latter case, the dome of the lamp better be clear too)

Over the mains receptor - Can be fitted in the lamp

Wireless receptor - Can be fitted in the lamp (and even justifiable, if the lamp is a LiFi lamp)

Microphone - Can be fitted, but you can spot that if you open the lamp...

Camera - You can spot that even without opening the lamp

So yep it can be done, but the latter 2 would not be too hard to catch



The lamp can transmit information in the range in which its light is seen (can be from quite far away, but only in plain view of the lamp, illuminated objects, window of the room etc)

If anyone would really want the information bad, he could have used Wireless or Over the mains communication too. Allthough maybe those are more restricted by FCC and such (need approval, which will uncover the entire story) while the light output is not. Anyway, if the information can be transmitted in means other than light, then any electronic device can be used for that... CFLs, electronic ballasts, anythng... then this would not be a problem specific to LED lights



To your question, the modulator is not a power device. It can control the switching operation of the ballast, where the ballast handle the actual power

All the spying instruments are not restricted that much by FCC or so, but just by the privacy laws.
So if e.g. the surveillance is ordered by a court, practically all methods are legal (the government agencies holds the needed licenses for all the radio equipment and so on), so they will use the best suitable method for them anyway (the light is not that practical - if a direct view is available, either a direct long range sound pickup or methods like long range window glass vibration sensing are qay more useful, as need no device to be installed in the target space).
And if someone does that against the law, such activity is illegal in any form, regardless what technical method is used...

As i understand it -



We are talking about different types of surveillance :

Medved you are talking about one surveillance device, installed or somehow else put into the premisses of one person, as part of legal action taken against the specific person - That is supposed to be allowed only by court order for each case

Here the concern is about preemptively putting surveillance capable devices into the posession of everyone, via the open market. This is similar to :

 - Software with intentional backdoors in PCs and phones, espeially communications and encryption software

 - Mass wiretapping into phone or computer communications of mass numbers of persons, without court order

 - Driving in the streets with IR camera in order to find indoors Cannabis grows, based on temperatures of exterior walls standing out from the rest of the building

 - Analysing readings of digital electricity meters to detect high power HID lighting in domestic use



All that had allready been done before, so there is indeed grounds to be concerned about technologies that enable such actions

Every product that is sold on the market must pass standards testing, regardless of presence/absence of spying capability

For example, a computer sold must satisfy FCC emissions requirements as a communication device, various safety/energy efficiency standards as an electrical appliance, and so on. It does not matter to those standards what software it comes installed with and whether there is spying function inside

However, if the allowed transmission power for the WiFi from this computer is X, but the spying data packets are transmitted at power greater than X (so that they can reach to greater distance), and the standards testing lab finds this out, the computer manufacturer will either not get the computer certified, or have to show the testing lab some reference that allows it to be exempt for reasons A B C. Such action puts the entire spying scheme at risk of being discovered, therefore is unlikely to be done

The concern with the lamp is - Lets say the lamp is capable to obtain information from its environment, for example, capturing sound. Now, the lamp can be made to transmit this information in few ways :

 - Over the power connection

 - Wireless in RF

 - In the light

Lets say that some FCC rule limits power distortion and RF emissions from a "general lighting lamp" type of device, and the imposed limits are too low to be practical for a spylamp : Transmission range without breaking the standards would be too short. Then the remaining option is transmitting it in the light



My point is, that the question that have to be asked is not "can the lamp transmit information" but "can the lamp obtain information, and if yes, why". This is more practical too :

 - Lets say you see a LED lamp with some driver chip. For all we know, many LED drivers are "all in one" chips - That would not look abnormal in any way

 - Lets say you see a LED lamp with microphone. That would raise big question why a lamp needs a microphone. That would be busted the moment somebody opens one up and spots the mic

As i understand it -



We are talking about different types of surveillance :

Medved you are talking about one surveillance device, installed or somehow else put into the premisses of one person, as part of legal action taken against the specific person - That is supposed to be allowed only by court order for each case

Here the concern is about preemptively putting surveillance capable devices into the posession of everyone, via the open market. This is similar to :

 - Software with intentional backdoors in PCs and phones, espeially communications and encryption software

 - Mass wiretapping into phone or computer communications of mass numbers of persons, without court order

 - Driving in the streets with IR camera in order to find indoors Cannabis grows, based on temperatures of exterior walls standing out from the rest of the building

 - Analysing readings of digital electricity meters to detect high power HID lighting in domestic use



All that had allready been done before, so there is indeed grounds to be concerned about technologies that enable such actions

Every product that is sold on the market must pass standards testing, regardless of presence/absence of spying capability

For example, a computer sold must satisfy FCC emissions requirements as a communication device, various safety/energy efficiency standards as an electrical appliance, and so on. It does not matter to those standards what software it comes installed with and whether there is spying function inside

However, if the allowed transmission power for the WiFi from this computer is X, but the spying data packets are transmitted at power greater than X (so that they can reach to greater distance), and the standards testing lab finds this out, the computer manufacturer will either not get the computer certified, or have to show the testing lab some reference that allows it to be exempt for reasons A B C. Such action puts the entire spying scheme at risk of being discovered, therefore is unlikely to be done

The concern with the lamp is - Lets say the lamp is capable to obtain information from its environment, for example, capturing sound. Now, the lamp can be made to transmit this information in few ways :

 - Over the power connection

 - Wireless in RF

 - In the light

Lets say that some FCC rule limits power distortion and RF emissions from a "general lighting lamp" type of device, and the imposed limits are too low to be practical for a spylamp : Transmission range without breaking the standards would be too short. Then the remaining option is transmitting it in the light



My point is, that the question that have to be asked is not "can the lamp transmit information" but "can the lamp obtain information, and if yes, why". This is more practical too :

 - Lets say you see a LED lamp with some driver chip. For all we know, many LED drivers are "all in one" chips - That would not look abnormal in any way

 - Lets say you see a LED lamp with microphone. That would raise big question why a lamp needs a microphone. That would be busted the moment somebody opens one up and spots the mic

You could maybe have an excuse for the microphone, Such as the lamp being clap controlled, But the real use of the mic' would be the spying.

Good point. That raises the whole question of whether we need devices that are too smart for their function. As in - do we really need a clap lamp ? Whats wrong with basic switch ?

To hide the true function of the lamp, If one exists.