Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Where is the information in an optical fiber?

  1. Apr 20, 2016 #1
    Hi. I know how optical fibers work but where is the information in the light signal passing through it? Which characteristics of the electric field carry it? I begun to think about it because I was intrigued if they took into account things like the Goos-Hänchen shift or angular deviations in the reflection law when designing the whole thing...

    Thank you very much.
     
  2. jcsd
  3. Apr 20, 2016 #2

    mfb

    User Avatar
    2016 Award

    Staff: Mentor

    The amplitude. If it is zero, there is no light, if it is non-zero, there is light. Add some details about frequency and polarization for commercial internet fibers, but those are irrelevant details for the concept.
     
  4. Apr 20, 2016 #3
    So, basically, the only thing that matters is whether there is light or not? Like a morse code? What are the polarization used for? And the frequency?
     
  5. Apr 20, 2016 #4

    mfb

    User Avatar
    2016 Award

    Staff: Mentor

    Like computers - binary.

    That is just the easiest approach. There are multiple ways to increase data transmission rate.
    - You can use different amplitudes (e.g. 8 steps from "no light" to "very intense light") to increase data transmission rate.
    - Data transfer for different frequencies (e.g. red and green light) is independent of each other, so you can use the same fiber with many different frequencies at the same time.
    - There are fibers that maintain polarization, so in principle it could also be possible to have independent data transfers for the two different polarization states. Some fibers allow more than two different states. But start with the basics before you worry about advanced concepts.
     
  6. Apr 20, 2016 #5

    Baluncore

    User Avatar
    Science Advisor

    The key word is “modulation”.
    The light is only the carrier wave. There can be many carriers of different wavelengths = colours of light in one fibre. The carrier is modulated by the information at the light source. Then at the optical receiver the carrier is demodulated to extract the information. There are many forms of “modulation”.
     
  7. Apr 21, 2016 #6
    Adding in some details:

    In the case where lasers are the transmitting source it is a design requirement that they always produce some light even in the 0 state. When lasers are turned off completely (extinction) they require a relatively large delay in order to get back on again (turn on time), and when they do snap on it is accompanied by a nasty optical overshoot followed by really high frequency ringing in the optical signal (relaxation oscillation). The turn on delay works against the high toggle rates which you want to achieve in fiber transmission, and the relaxation oscillation can be signal integrity issue if the receiver bandwidth is high.

    On the other hand, logic 0 power is essentially wasted. The most efficient use of power is to have all of the launch power in the logic 1. So we want the laser to be close to off for 0, but not too close. The on/off ratio is called the "extinction ratio". Large extinction ratio improves signal to noise ratio. But laser diode characteristics change significantly over temp and over time (aging), and this limits how high we can push extinction ratio and still guarantee that we avoid extinction.
     
    Last edited: Apr 21, 2016
  8. Apr 21, 2016 #7
    Another tidbit:
    At 10Gbps and given the propagation velocity in the fiber, each bit occupies about 2cm of cable. At any frozen moment in time there is a new bit every 2cm.
    As of 2016, Corning has deployed somewhere around 200 million Km of fiber. If all of this is lit at 10Gbps, we have something like 10terabits of data in flight within transmission cables.
    And this is not even taking into consideration WDM.
     
  9. Apr 25, 2016 #8

    Baluncore

    User Avatar
    Science Advisor

    Lasers are not turned on and off to modulate the light. The light is usually modulated by a crystal with optical properties sensitive to an electric field. The laser runs continuously while the light travels through the modulator that changes the polarisation or the phase of the light.
    The fastest modulators employ a splitter to produce two light paths, one path passing through a phase shifting crystal. By combining the two optical signals, the light can be made to sum or cancel at data rates measured in tens of gigabits per second.

    I believe that Fibre Optic Gyroscopes use a spool with several kilometres of fibre so many bits can be in-flight at the one time, to give high angular resolution. Looking back to the earliest years of computer technology, the mercury delay line was used for mass storage because the ultrasonic travelling wave in the delay line gave a high data storage capacity. The first desktop programmable calculator, the Olivetti Programma 101 used a magnetostrictive ultrasonic wire delay line to store about 256 bytes.
     
  10. Apr 25, 2016 #9

    sophiecentaur

    User Avatar
    Science Advisor
    Gold Member

    It is actually quite possible to modulate the amplitude of a light beam in proportion to a varying (audio) signal voltage. I remember someone at work doing that in 1968. A very crude quality signal was carried across the lab. But audio sources and detectors tend to be non linear so an analogue optical link is not satisfactory. Because of the possibility of very fast on/off switching of light, it is better to send digits.
    It is also possible to modulate the polarisation of a light beam (link) and that can also be used to carry information.
     
  11. Apr 25, 2016 #10
    Direct modulation lasers (DML) are the most widely deployed (by far). These do essentially turn the laser on and off (although not completely off as I mentioned in previous post).
    EML and MZM that you are describing are more expensive and used in specific cases where datarate is > 10gbps and/or reach is very long. Soon 40gbps will be available with DML.
     
    Last edited: Apr 25, 2016
  12. Apr 25, 2016 #11

    analogdesign

    User Avatar
    Science Advisor

    I actually built such a system (analog fiber optic transceiver) in a lab class in college (over 20 years ago). It isn't complicated; basically you modulate the current to an LED driver with an audio signal, then detect it on the other side with a phototransistor (or photodiode). Put the photodiode output into a transimpedance amplifier and then a power amplifier and it works great. We were able to get very good results (although of course we were using 90s technology photodiodes, LEDs, and op amps. It was a very fun project and neat to whisper into a microphone and your partner hearing it across the lab.

    It was a wonderful class. We also built a version of the Dolby B (emphasis) system for audio recording and a PLL to demodulate digital data stored on cassette tape (they had to find a use for all those tape drives, right?). Off topic but the biggest doubt I have for online learning in electronics is the ability to have these rich, group experiences I had building, soldering, and debugging hardware. Can't do that with LT Spice.
     
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Have something to add?
Draft saved Draft deleted



Similar Discussions: Where is the information in an optical fiber?
  1. Optical fiber (Replies: 5)

  2. Fiber Optics (Replies: 4)

  3. About fiber optics (Replies: 2)

Loading...