From: hobbs@electrooptical.net   
      
   On 4/26/2014 9:38 AM, haiticare2011@gmail.com wrote:   
   > On Friday, April 25, 2014 2:25:55 PM UTC-4, Phil Hobbs wrote:   
   >> On 04/24/2014 11:21 PM, hzaiticasssre2011@gmail.com wrote:   
   >>   
   >>> Phil, you mentioned you were developing same   
   >>   
   >>> That sounds interesting.   
   >>   
   >>> Are you locking in fast?   
   >>   
   >>> I imagine certain signals are time delayed, but not by much.   
   >>   
   >>> It's a fascinating field, and I wonder if, say, weak spectra can be pulled   
   from   
   >>   
   >>> stronger spectral lines that way.   
   >>   
   >>> Have an interest in EKG.   
   >>   
   >>>   
   >>   
   >>   
   >>   
   >> It's for a fancy microplate reader, where we need absorption stability   
   >>   
   >> of +- 2E-4 absorption units over a 0-2 AU range, and have 180   
   >>   
   >> microseconds to do each measurement.  (The plate is in rapid motion.)   
   >>   
   >>   
   >>   
   >> Any jitter between the chopping of the light signal and the trigger rep   
   >>   
   >> rate has to be eliminated, so what the box does is to wait for a   
   >>   
   >> trigger, then (using a fast clock) chop a light source exactly N times   
   >>   
   >> (2 < N < 255, switch selectable) at 200 kHz, and synchronously detect   
   >>   
   >> the resulting photocurrent.  The synchronous detector uses a 10-state   
   >>   
   >> state machine, four plus, four minus and two to let the amplifiers   
   >>   
   >> settle in between.   
   >>   
   >>   
   >>   
   >> Once it's done, it holds the integrated value and triggers a 16-bit ADC   
   >>   
   >> that the customer is controlling via SPI.   
   >>   
   >>   
   >>   
   >> It just barely fits into an XC9572XL.   
   >>   
      
   > And - of course a key design challenge is how to do a pipeline   
   > to transfer and store all the data. With DSP, that is built-in, at   
   > least for on-board SSRAM - Even then, some scheme is needed to   
   > move in to disk transparently. A DMA system of some sort.   
   >   
      
   In this case that's pretty simple, because the rep rate is only a   
   kilohertz and the customer is providing the trigger pulses.   
      
   One thing I tell my customers who are deciding whether to to the lock-in   
   part using DSP: Everybody's _second_ digital lock-in design is a lot   
   better than their first one. ;)   
      
   Cheers   
      
   Phil Hobbs   
      
   --   
   Dr Philip C D Hobbs   
   Principal Consultant   
   ElectroOptical Innovations LLC   
   Optics, Electro-optics, Photonics, Analog Electronics   
      
   160 North State Road #203   
   Briarcliff Manor NY 10510   
      
   hobbs at electrooptical dot net   
   http://electrooptical.net   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)