[Ronja] What is holding back 100 mbps

[Thomas Egenhofer]

> > I start this thread to exactly pin point the problems of a 100 Mbps RONJA.
> > Let us reach to a conclusion as to what exactly must be modified.
> > 
> > Let us start with the LED,
> > 1. To transmit at 100 mbps,we must find an LED which can switch on and off
> > in 1 ns.
that's not neccessariliy true. if you wanna use 2-level encoded signals like now ,you'd run into several problems.
first is you would have to change the encoding of the MLT3 signal to 2-level signal which is suited for optical transmission. in easiest case manchester code.
which would result in a 0.5ns signal-time. which again would require about as few as 0.2 or even less rise/falltimes. from my experience this is not possible with LED's at all.
in any case this would require an almost complete re-design of the entire electronics ending up in a by far ways more comlicated circuit.

second option is to stick with the MLT3 signal delivered from the networkcard. this leaves you signal-times of (correct me if i'm wrong )  ~30ns which is still ad the edge of LED's capabilities. right now ronja is at 50ns. so the LED's would require risetimes of 15ns  or less .difficult to find but still possible.though this would require you to redesign most of ronja,too. but it would spare you from transcoding the signal twice.
you would need to build a current-controller for the LED's which is capable of switching the driving currents very fast and accurate. maybe even over-drive it and use a monitor-diode to get a light-output feedback.
on the receiver side you would need to build a 3-level receiver.. as it happens i was sorta bored recently and i draw a rough shematic of such a thing.
it basically divides the 3-level input signal into 2 2-level signals which can be easily compared against the half of the average signal AC.
it doesnt look very trustworthy and requires a lot more additions to actually work in real-experiments. but the simulation on this tiny idea worked quite well. so at least it's a start.

feel free to improve it -> http://home.arcor.de/positiveelectron/files/MLT-3-receiver.png

ronja sorta already has a MLT3 line transmitter to match the IEEE standarts of link-integrity bursts. so the efforts on this end would be comparebly small.
all in all the require changes are tons of work. but at least within the range of technical posibilitys.

just my 2 cents on all of this. but if you got intrested feel free to ask for more :)
i'm also quite intrested in speeding up ronja a little^
_____________________________________________________________________
Unbegrenzter Speicherplatz für Ihr E-Mail Postfach? Jetzt aktivieren!
http://www.digitaledienste.web.de/freemail/club/lp/?lp=7