[geeks] Whee! Lightning strikes, AGAIN!

[Sandwich Maker]

" From: gsm at mendelson.com
" 
" On Tue, Jul 28, 2009 at 07:39:11AM -0400, John Francini wrote:
" >
" > Interestingly, the IEEE proposals out there for 40-gigabit and 100- 
" > gigabit talk about ganging 4 or 10 10-gigabit *fiber* connections  
" > together -- I would have thought that they could just use multiple  
" > colors (frequencies) over the *single* fiber without interference, using 
" > prisms to join/split the colors.  Or am I being way too simplistic?
" 
" The problem is that all the current equipment is designed for one color
" of laser. As 1000Base-T proved, it was much cheaper to gang 4 channels
" together, which is basicly software, then produce something 10 times
" the current speed. Two directions on the same pair was simple, telephones
" have been doing it since the 1920's and 250mbit was not that much more
" difficult in the late 1990's than 100.

there may also be bitrate / attenuation / cable length / network range
issues involved.

" The same with fiber. While it may not be colors, but different fequencies
" of the same color (ok, that's still colors), some sort of splitter/combiner
" would have to be designed and built. 
" 
" []
" 
" Building some sort of splitter/combiner setup is much more expensive as it
" requires filters that may not exist yet, sensors and lasers tuned to different
" frequencies, etc. 

diffraction gratings.  sensors needn't be tuned; each would only get 1
channel.  i'm not sure how well diffraction gratings would work on the
transmitter side though...

back in the '80s i talked with a cow-orker who was working on fiber
cable tv.  he said the best lasers he saw had a line width narrow
enough [5nm comes to mind] to pack 100 channels into the fiber
'window', but they had no control over the process variables that
affected this, and for their purposes then it didn't matter.  it might
be better now.  he also said that the lasers could be tuned in
operation, by voltage iirc.  this makes me think that modulation would
increase the space needed per channel.

i don't recall if this was at 1.3 or 1.55 microns.  i have vague
memories that they weren't very different.
________________________________________________________________________
Andrew Hay                                  the genius nature
internet rambler                            is to see what all have seen
adh at an.bradford.ma.us                       and think what none thought