Maximum laser frequency

Just curious how hard where you thinking of pulsing the 808's again? I would think a schmit trigger and some ultra-fasts Fet's or IGBT, even a BJT could work, the problems start to arise at how the min and max current being used, too little and it might not urn on completely and you'd suffer system losses plus, possibly instability, if you push too hard it might take longer for the device to shut off defeating your design goal.

Well at first I was going to try to emulate a cat5e cable. But if it has to go into 10 instead of 100, that sounds a lot more doable, and as fast as an old 11mbps wifi network.

And a schmit trigger for the receiver was what I was thinking. If I can get one that fast. lol

But as for your question, ideally 100mhz, though since its becoming more and more unlikely that I can find something that receives that quickly, 10mhz.

Actually:

http://sales.hamamatsu.com/index.php?id=13183106

That looks like a winner. 100mhz bandwidth, 8-9 dollars each. Not bad at all! Nearly correct wavelength as well.

Now it all falls to whether the laser can keep up with that (I hope) and that the circuitry to drive the laser can keep up with that. As mentioned before, most drivers don't want to be turned on and off at 100mhz... So hopefully either bypassing or disconnecting the diode rapidly can do that.

If you are really concerned about on and off, just alter the output power and leave the laser on. It'll switch incredibly fast if you are only varying the output level. The nature of photodectectors allows you to see the received input as a voltage level. The circuitry is easy enough to build to filter anything below a certain voltage as a zero and anything above a certain voltage as a 1. Now with that in mind you need to decide on a protocol stack to use because you will need to build the circuitry allowing signaling bits so you can sync up the transmission. Open air laser transmission is going to be fraught with all kinds of interference so you will need something that allows retransmission of lost packets such as TCP at layer 3 or frames at layer 2. The interference is going to muddle the actual energy received by the detector so you will need to be able retransmit data as well as have a large enough dead zone between 0 and 1 on the physical layer to minimize false 1s and 0s yet narrow enough to keep high speed switching

Oh uh...

I was just sorta hoping that I could tie directly into an ethernet cable and hope it doesn't mind the slight delays. I know a bunch of my stuff can deal with packet loss, if thats what you mean..

That might not work though. I'd have to test it.

Just something to chew on: 1X speed for a bluray writer is 36Mbit/s

Really?

That sounds promising then.