19012fc4   
   XPost: sci.physics, sci.electronics.misc   
   From: boxman@voyager.net   
      
   On 1/12/2012 8:24 PM, RichD wrote:   
   > On Jan 12,  Helmut Wabnig  wrote:   
   >>> I was looking at some LED flashlights, with ratings   
   >>> from 50 to 150 lumens.  A p-n junction is a p-n   
   >>> junction, how do they get such varied outputs?   
   >>   
   >>> I examined the working end, they all look alike,   
   >>> with a single bulb.   
   >>   
   >> The more current you  force through a p-n junction,   
   >> the more light will come out.   
   >> With veryvery much current, it will flash very brightly,   
   >> for a short duration.   
   >   
   > Thay all use the same batteries,  so it's as simple as   
   > a series resistor?  But then why does the higher   
   > output cost more?   
   >   
   > Also, there are flashlights which use multiple LED -   
   > you can count them in the housing - why would the   
   > manufacturers do that, paying for extra parts, rather   
   > than using a single LED with high current?  And   
   > some of those multiple LED units cost less than   
   > the single ones I checked.   
   >   
   > Your reply seems too simplistic to be satisfactory -   
   >   
   > --   
   > Rich   
   There are several factors that contribute to the LEDs cost.  Your base   
   assertion that a p-n junction is a p-n junction is entirely too   
   simplistic.  A simple p-n junction composed of a single p layer with a   
   single n+ layer and ohmic contacts will not generate a significant   
   amount of light due to self absorption in the thick p-layer and the fact   
   that the recombination is not well confined resulting in low efficiency   
   of conversion from electricity to photons.  These types of LEDs are easy   
   to make by diffusion and thus not very costly, but not a lot of light.   
      
   A high output led requires a  heterojunction design and most high output   
   leds utilise a double heterojuction design.  What that means is  a   
   typical high output LED will start with a P type electrode connected to   
   a P GaN layer connected to a P AlGaN layer connected to a N InGaN layer   
   connected to another N AlGaN layer connected to a GaN buffer layer on   
   top of some substrate (usually sapphire, but Cree uses Silicon Carbide).   
     The n electrode is connected to the  GaN buffer layer.   
   The thicknesses of the layers have to be precisely controlled to get the   
   proper wavelength output and the proper electrical characteristics which   
   demands a very complicated epitaxial growth process.  Additionally, all   
   the different layers require careful processing to eliminate the stress   
   between the different lattice structures to prevent defects.   
      
   The result is an epitaxial growth process that can have very low yields   
   and wide variations in performance across the wafer substrate.  When the   
   wafer is cut into the die pieces used for the individual LEDs, the die   
   pieces are checked for output and binned according to how bright they   
   are.  The highest performing dies are consequently a small part of a   
   production run which leads to the more simple equation that low supply   
   and high demand = high price.   
      
   Hopefully that gives you some idea of how prices can vary so much for   
   LEDs and why high output LEDs cost more.   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)