Recently I read a few posts on the internet about how to modify a laser so it’ll be able to transfer data, such as sound and even images. It said something vague about modulation, but I’m still 16 and haven’t yet reached that subject in Physics to understand what it is. Could someone, in relatively simple terms, explain the science behind this?
Thanks,
Matt.
In a nutshell light off = 0
light on = 1
all computer communications come down to a stream of zeros and ones.
a laser, flashed on an off to represent that pattern while pointed at a reciever who understads that pattern, you have data transmission via laser.
Or to further elaborate
the number 14 in binary would be
00001110
so if a reciever was expecting 1 bit of data (a zero or a one) per second and over 8 seconds recieved
nothing
nothing
nothing
nothing
flash of light
flash of light
flash of light
nothing
It would interpret this as a 14.
And if you point that laser at the end of a really, really, really long and skinny glass tube, the light will bounce around through the tube until it comes out the other end with comparatively little loss in energy. This is how fiber-optic cables work.
Check out the awesome HowStuffWorks article How Fiber Optics Work for an easy overview.
also of note
once you have numbers from 0-256
you standardize a pattern for interpreting them as regular letters and numbers we would recognize.
To begin with, a definition: ‘Modulate’ just means ‘to modify’, ‘modulation’ means ‘modification’, and neither of those words are at all useful unless they either specify or you already know what is being modified.
In every waveform, you can modify the frequency (the color of the light or the pitch of the sound), the amplitude (how bright light is or how loud sound is), or whether the waveform exists (just flip it on and off a bunch of times). The base waveform, the thing being modified, is called the ‘carrier’.
OK, note the problem with this: How do you distinguish 000 from 00 or just 0? Similarly, how do you tell 111 from 11 or just 1? To do that you need to know how long each bit is transmitted (one second, in this case) and where within that time frame the sender’s clock is. Keeping track of this is actually kind of an interesting problem. One very common way to solve the problem is to use Manchester encoding, which is what Ethernet uses to encode data sent as pulses of electricity down a wire. The essential feature of Manchester coding is that data is encoded as a change of state, not as a continuous state. (The diagrams in the article make this a lot clearer.)
As I understand it, Manchester encoding is inefficient and more recently there are more complicated schemes that work better.
I’m not sure how recently you’re talking about. As far as I’m aware, most systems still use either Manchester or GCR, or one of their derivatives, and GCR was invented in the 1970s.
If I wanted to learn this stuff from scratch, I’d start off with the simplest codes, then look at what problems the more complicated ones are trying to solve.
If you have a shared external clock, you can use NRZ, which is as easy and efficient as it gets. The obvious problem is needing an external clock.
RZ is the simplest self-clocking signal. It should be easy to see how it gets by without a clock–but a little more reflection and you can see how jitter and DC wander cause problems. Still, it’s good enough for a lot of telecom applications.
Manchester is a relatively simple solution to RZ’s robustness problems (as is BMC, used in some digital audio applications).
One problem that all of these have is parity. If you turn the signal upside-down (e.g., you’re sending through home power lines, and someone plugs the receiver in upside-down), it flips over. Differential Manchester, and similar variations on other protocols, can solve this.
Once you get all that, take a look at GCR. The idea behind GCR is that in many cases, you don’t need to do Manchester-style self-clocking as long as you can do clock recovery. The way to do this is to encode X values into Y symbols in such a way that no valid symbol has more than, say, 6 highs or lows in a row. You can do this with under 25% overhead, instead of 100%. But it’s much simpler to go with exactly 25% overhead, and this gives you some spare symbols for control codes, to maintain DC balance, etc.
8/10 (aka 8b/10b) is a specific version of GCR. By an accident of history, IBM got a patent on 8/10 that also covered most other implementations of GCR. So, in 1983, everyone basically stopped using GCR except in applications where it was worth licensing 8/10 (like DAT tapes). Then, when the patent expired, 8/10 became a sort of de facto standard.
There are other run-length limited codes out there; CDs use one, and there are a few applications where the tiny bit of extra efficiency in a 10/12 code (20% instead of 25% overhead) is worth the huge extra complexity. But nowadays, most things are either 8/10, or use Manchester or BMC or one of their variants.
It should be noted that you don’t need to use digital encoding - analogue may do the trick just fine.
One scheme involved bouncing a laser off a window - the laser source is a constant, but the window vibrates based on the sounds within the room and modulates the laser reflection. A receiver just measures the reflected laser intensity and amplifies it to listen in. This can be done at significant range, but does require a separated transmitter/receiver for maximum reflections off the window (with a shallow angle of incidence). I have built a photophone that uses the same principle, but with sunlight and a mirror glued to a speaker for the transmitter.
A laser has plenty of bandwidth and I cannot see why an analogue video transmission scheme could not be carried (ie VHF/UHF) using an analogue modulation technique. But it would be line of sight only.
Si
For which the encoding would be similar to the encoding in analogue radio and TV waves. In these, you have a sinusoidal “carrier wave” and the information is encoded via variations to that basic wave: variations in its frequency with constant amplitude (FM) or variations in amplitude with constant frequency (AM). Amplitude is the wave’s height, frequency is the time between peaks. The emitter adds analog waves (representing sound and/or image) to the carrier wave, the receiver substracts the carrier from the received signal and sends the result to the speakers/screen. By tuning your radio/TV, you are choosing both what part of the radio spectrum to pick the signal from and which carrier wave to substract.
Some modulation schemes:
On off modulation. One way to send digital data. Morse code, basically.
Amplitude modulation: You change the intensity (strength) of the light over 0-100% range, not just on off. This can allow transmission of sound or video without digital encoding. Usually done with an accusto-optic modulator. (AOM) This splits the laser into two beams. How much power is in each beam depends on how much power is applied to the AOM crystal by the modulating signal.
Pulse position modulation: Brief, very strong laser pulses are sent. The information is encoded in the time between the pulses. This is the system used in military MILES training gear. (Hi-tech laser tag, basically). Due to the intense pulses, it works pretty well in atmospheric applications. The pulses are kept very short so that the average power is low, avoiding eye damage. I’ve designed such equipment and can elaborate on this if asked.
Frequency/wavelength (color) modulation: By using different lasers, or heating a diode laser, or possibly other means, you can change the color of the laser light, and encode information that way.
When I was in highschool, we had a physics teacher who was renowned for his demonstrations. One such demonstration was designed to show how lasers could transmit data by using a laser pointer, a special programmable fan, and a receiver that played a monotone hiss whenever it was hit by a laser. Our teacher first turned on the laser and aimed it at the receiver, and it responded by hissing loudly at us. He then waved his hand over the laser beam, temporarily blocking it from hitting the receiver, and the receiver hissed only while the laser’s line of sight was clear.
Here was when it got really interesting. Our teacher placed the fan between the laser and the receiver and set it to spin at a constant rate. The receiver then played a musical note which rose in pitch when the fan was sped up, and fell in pitch when the fan was slowed down. Pretty cool huh? But wait, it got better.
The fan was programmable so that it could execute precise series of speed changes. Our teacher then played some music from the receiver by changing the speed at which the fan blades spun. It wasn’t perfect, as the fan blades took a small amount of time to adjust speed, but it was pretty freaking amazing coming from a device which simply hisses with constant input. I think that, of all the songs our teacher played on this setup, my favorite was moonlight sonata.
Throughout all of that, you could just halt the music by turned off the laser pointer. That was years ago, and it is still one of the coolest things I’ve ever seen.
This video is the best single primer I’ve ever seen on how both digital audio and video work. It goes into more than enough detail to get you well on your way to understanding everything and, best of all, it’s free. You can download it here.
If you don’t like video, the information is in written form as well.