0
\$\begingroup\$

I am inquiring about this because I'm looking for a suitably high-power, large bandwidth light source for a fibre optic communication system I am designing.

The distance the data will be required to travel is 45km and the light source needs to transmit data at a bandwidth of 12MHz, and at a data rate of 12Mbps. Ideally, I want to use a light source with a central wavelength of >1300nm. The catch is that I have to use multi-mode fibre and laser diodes are out of the question too.

What kind of output power out of any potential light source should I be looking for here?

\$\endgroup\$
2
\$\begingroup\$

12 Mbd over 45 km with a single wavelength with multi-mode fiber will just not happen, due to the physics of light propagation.

Really, really good multimode (OM4/OM5) fiber (that's mainly about consistent geometry) has some 3 GHz·km bandwidth-distance-product at 850 nm, and something like 500 MHz·km at 1300 nm. That means that after 45 km, you get 500/45 MHz in useful bandwidth (before dispersion distorts your receive signal too much). That's actually ca 12 MHz, but you need to account for the fact that your data symbol train is rectangularly shaped and thus needs way more bandwidth.

Also, you get ca 0.5 dB/km of attenuation in Corning OM4 fiber, assuming you have that high quality fiber, so that's 22.5 dB over 45 km - attenuation is really not your problem, and hence, maximum light source power isn't, either.

Really, multimode fiber is the wrong choice here – as the name says, it's got multiple modes of light that can propagate, and hence, you get badly behaving modal dispersion - and dispersion makes your transmit symbols indistinguishable at the receiver.

You can probably, maybe, good luck with that, compensate for that if your receiver was a coherent receiver (which explodes the cost extremely quickly), and your transmitter has the same complexity, and you do the appropriate DSP at both ends to pre-code at the transmitter and de-disperse at the receiver. You actually will put plenty care not to put much power into the fiber – because that makes it a nonlinear medium, and then you get, atop of the dispersion, non-linear effects.

| improve this answer | |
\$\endgroup\$
  • \$\begingroup\$ Multimode fibre isn't a choice in this case... I have to use multimode. \$\endgroup\$ – AustereTiger Oct 20 at 21:47
  • \$\begingroup\$ As per my answer (read it, please!): yeah, multimode fiber isn't a choice in this case. Physics doesn't care what you have to use. \$\endgroup\$ – Marcus Müller Oct 20 at 21:51
  • \$\begingroup\$ But then what do you suggest I do? I can't use single-mode fibre, the specifications explicitly say multimode fibre. I don't understand how I'd get an LED to transmit over 45km in multimode but the brief says I have to. \$\endgroup\$ – AustereTiger Oct 20 at 21:54
  • \$\begingroup\$ well, then tell whoever gave you the brief that physics says "no" and they need to give you a better brief, or need to figure out how to build an equalizer for modal dispersion (look up that term, there's decades of knowledge on that out there, nontrivial problem). \$\endgroup\$ – Marcus Müller Oct 20 at 21:55
  • \$\begingroup\$ Are you sure you aren't just missing something, Marcus? I'll copy the brief into this comment. Well, it turns out I can't, but I know the brief and I know "no" is not an answer I can give! \$\endgroup\$ – AustereTiger Oct 20 at 21:59
0
\$\begingroup\$

The distance the data will be required to travel is 45km and the light source needs to transmit data at a bandwidth of 12MHz, and at a data rate of 12Mbps. Ideally, I want to use a light source with a central wavelength of >1300nm.

Given multimode and low bandwidth, an LED is probably a good option. Notably absent here was a specification for TX power, which is probably the most important thing you should be thinking about.

The catch is that I have to use multi-mode fibre and laser diodes are out of the question too.

SLEDs can be operated as lasers, so this specification doesn't make a lot of sense. Regardless, if you're running multimode, you don't have a lot of reason to choose a SLED (which is specifically used for single mode operation).

| improve this answer | |
\$\endgroup\$
  • \$\begingroup\$ What do you mean by TX power? Transmitted power? Another thing: There aren't many LEDs that can transmit data to 45km; not that I can find anyway. \$\endgroup\$ – AustereTiger Oct 20 at 21:34
  • \$\begingroup\$ yes, TX is "communication engineerish" for "transmit". Also, any light source can transmit light arbitrarily far. Light extends infinitely far - sure, it gets attenuated, but here, dispersion is your problem. \$\endgroup\$ – Marcus Müller Oct 20 at 21:38
  • \$\begingroup\$ @AustereTiger Why do you think that? \$\endgroup\$ – user1850479 Oct 20 at 23:00

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.