[Author Prev][Author Next][Thread Prev][Thread Next][Author Index][Thread Index]

RE: ATM Active Optics




Yep, $15k gets you a lower grade 1/4 good "Picnic" array.  The lowest grade 
may have been $12k but that's kind of a fine point.  My hope is that I-band 
will show improvement.  Then, maybe if I ask real nice, someone will let me 
borrow an old camera or sell me an old array.  There are several years of 
work ahead before it's an issue :-(.

Bob mentioned the issue of general purpose computers.  I guess I have two 
comments.
1) Your correct, the systems I've seen use data-flow machines or better. 
There 
      are PC based RISC and FPGA boards that can be used if needed.
2) The actual requirement is based on the number control points and 
algorithm.  I
     have to look at 4 quads and update 7 D2A channels.  The math is all 
scalar,   
     which also helps. At 30-250 updates per second I think the Pentium can 
handle
     it if I use an RTOS.  
  


Date: Thu, 26 Apr 2001 12:45:18 -0700
From: Frank Henriquez <frank@ucla.edu>


Not only will an IR array be expensive ($15K - $150K, depending on 
the resolution and type) you'll also need cryogenics. A Rockwell 
NICMOS HgCdTe array will work great at K ...