Transfer Data with Lasers?

[Benm]

Because of the receiving end of the system... the ambient light from the outside
world has a negative effect on the receiving sensors ability to see small laser level
variations.
By modulating the laser... like a carrier wave... and demodulating it it at the
receiving end.. it will easily block out unwanted signals (varying light levels)
and more easily see the true ON/OFF data...

For example... using an IR LED and IR photodiode with no modulation... just pulsed
on and off at 100 hz I get a detection distance of 2"-6"...
But when the same IR LED is Modulated at 40kHz and the Photodiode Demodulates
the ON/OFF pulses at the same power levels I get a distance of over 30 feet.


Jerry

You're absolutely right there, using a carrier will do a lot to tackle noise problems and interference. Problem is, if we're aiming for 10 MBit, the carrier would have at least an order higher (if you wanted to do FM and demodulate with a PLL), so you're looking at modulation and demodulation in the 100 MHz+ range.

 

[ElektroFreak]

That might not be too much of a problem, though. Laser diodes can handle that rate easily.

 

[gotwake424]

Unlike typical laser modulation, with data communication the signal never goes as low as 0V. All you need is a high and low voltage to communicate data, so rather than actually turn the laser on and off with each data bit, you just dim and brighten it. This allows for even higher speeds than if you were to actually pulse the laser on and off.

I am pretty sure that you could do this whole setup with hardware alone.

 

[lasersbee]

I've seen LD data sheets that state a 100ns pulse rate for pulsed operation...
I think the LD O-Like sells is rated at the pulse rate in a 50% duty cycle... IIRC

[EDIT]
I saw it on IgorT's thread....

I know... it is a 405nm....


Jerry

 

[ElektroFreak]

I think that those pulsed duty-cycle ratings are for longevity, since LDs used in pulsed systems typically generate more output per pulse than when they are run CW. I don't think it's intended as a rise-fall time indicator, especially when you're modulating the diode by dimming it rather than pulsing it.

 

[lasersbee]

The pulsed duty cycle is to show you the pulse width and duty cycle required to
get the max pulsed power output as shown on the datasheet...
Data sheets are usually set at min/typical/max allowable conditions for safe operations
of the product.

If an LD can be run at a 100ns 50% duty cycle (the above example and that is
to get the max pulsed output... you don't need to run it at max but still with the
same duty cycle)... then just call that the carrier signal...
Your data rate would be a bit less than that frequency depending on
the requirements of the demodulator...

It would take a few Carrier pulses for the Demodulator to know the bit
is high and a few missing to know the bit is low..

Jerry

 

[Benm]

I suppose one advantage of going from the ethernet port is that the protocol on that is pretty sturdy, and will re-try if it for some reason gets a corrupted signal.

But still, using a carrier will be no easy feat. The signal would be centered around 100 MHz or so, and 2 x 10 MHz plus 2 x the FM deviation wide. This means you will need a bloody fast PLL do demodulate this. In the low-MHz range the pll is no problem and any old 4046 chip would work just fine, but making it work at these speeds will pose quite a challenge.

I think the principle will work without using a carrier though - you can make the receiver very directional using lenses and/or a long tube so it picks up little ambient light. If you can add a monochromatic filter at the laser wavelength too, i think noise would be very manageable, also at larger distances (say 100s of meters to 1 km or so).

 

[lasersbee]

You may have a point about a direct approach at those frequencies....
I've only worked on low power (3mW).. low speed RS232 laser communications
as I pointed out in a previous post..
I suppose you could easily overcome the ambient noise aspect by using a more
powerful laser... and the mod/demod may not be required...


Jerry

 

[Benm]

That's the idea indeed, just recreate the proper signal levels at the receiving end and feed them into the network port. On the transmitting end drive the laser based on the signal from the ethernet card.

I think this might be possible using just a few discrete components, and salvaged parts (balanced transformers in particular) from surplus ethernet cards.

 

[pullbangdead]

Wow guys I learned a lot of new stuff on this thread! My first thought was "this would be so slow, why wouldn't you just transfer is over the internet?" but now I see that the possibilities are actually really great, especially using a diode laser
I have to try one of these with a friend.. I can see his window from mine even though we are about 800m apart.
Thank you everyone for your input, really interesting to learn something new!

Seb

You do realize pretty much every bit of data transferred over the internet is also being transmitted by lasers? The entire backbone of the world wide web is in fiber optic cables, where lasers send the data back and forth, across continents and oceans, and nowadays even all the way into houses thanks to newer services like FiOS.

Everything we're typing on this webpage is being transmitted by lasers at some point or another; lasers are one of the massive enablers that make the internet (in its current form) possible at all.

It's still a cool idea and a cool DIY project to set up your own data relay over the air using lasers, but all our communications on the internet are already being transmitted by laser every second over fiber optics.

 

[ElektroFreak]

^Yup.. I deal with FL stuff all day at work. Communications lasers are always infrared, normally around 1.3-.1.5um

 

[pullbangdead]

^Yup.. I deal with FL stuff all day at work. Communications lasers are always infrared, normally around 1.3-.1.5um

Yep, very cool stuff. And a perfect example of materials science, too, and how the atomic properties of glass influence what wavelengths of light you use:

Keep up the good work!

 

[Benm]

Building a laser link trough plain air is quite another challenge compared to fibre linked technology though. Fibre coupled systems are limited by attenuation mostly, externally introduced noise is not so much a problem.

 

[lasersbee]

Using a Laser to xmit/rcv full Ethernet data will require a Laser and Receiver
at both ends...

Jerry

 

[MrGreenJeans]

With sufficient quality detectors and very low noise laser system there really is no reason you cannot use air as the transfer medium and still multiplex the signaling as you would across fiber. Of course, there will always be line of sight issues and atmospheric attenuation but the basic concepts of optical multiplexing will work. It's not much different than long distance microwave shots except microwaves are less susceptible to optical atmospheric attenuation.

The concept works quite well, at least at low modulation/encoding rates, with distances on the order of a couple of miles. Military laser designators modulate the laser signal with a pre-coded sequence base on daily military crypto documents. This modulated signal is bounced off the target. The laser seeker, with the same code information loaded, in the head of the weapon searching for a target will recognize the modulation pattern and know where to strike. This prevents the enemy from spoofing the designator beam and redirecting a munition onto friendly forces. I can't really go into the specifics of the modulation but understanding various multiplexing schemes should spark some ideas with you guys who are interested in communicating across free-space optical links. Study whats been done and apply it such as dense wave multiplexing, and other schemes. You will find there is really no need to turn the laser off at any point as TTL type modulation is akin to the telegraph versus multiplexed satellite communications.

Personally, I always thought it would be neat to split a beam and encode information on both beams but only being able to decipher the information when the two beams, when propagated across free space, interfere at the receiving detector. The interference pattern contains the information. Unfortunately, in free space, limitations in coherence length would mess up the ability to decipher the information.

 

[Grix]

YouTube - Make a Simple Laser Communicator