From: boyd@apt0.sao.arizona.edu   
      
   Mikko OH2HVJ wrote:   
   >   
   > Correct.. I'm trying to get my customer understand that with the   
   > current limitations (total of 15mm^2 LED surface radiating into   
   > almost half-space and a 0.8mm NA 0.5 fiber) it is impossible   
   > to get more light into the fiber. Also, I'd like to be able to   
   > explain the same to our EEs.   
      
   Have you tried bonding the end of the fiber directly to the face of a   
   white LED with about the same area as the aperture of the fiber.  That   
   would be over 100 times more efficient than you can do with 15mm^2 area   
   using multiple LEDs.  I'd expect around  10% total efficiency but I   
   don't' think you'll do better without a coherent source.  The small led   
   will have a limited power which may be helped some with active cooling.   
      
   How much actual white light power does the application need though the   
   fiber in milliwatts?  What is a minimum practical life for the light?   
      
   To get more energy through the fiber a brighter (energy per unit angular   
   area) light source is required.   A short arc xenon lamp is the   
   brightest practical white light source I'm aware of.  I'd expect you   
   could get close to  watt though a .8mm fiber with a properly focused 100   
   watt lamp.  That's not near as efficient as the direct coupled LED but   
   much better than what you're proposing with 225  square millimeters of   
   LEDs. The arc lamp should have an emission area under 10 square   
   millimeters.  It needs a relay lens to match the NA of the fiber.  A   
   rear spherical reflector should help by a little less than a factor of   
   two.    The power supply for a short arc lamp is not tiny and the wasted   
   power has to be dissipated.  That may not be practical at all   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)