Lock-in is a great idea but these are very slow filters. I must make a   
   decision within 100usec or so. At my frequencies, they take a very long time.   
   I experimented with Lock-in before and gave up. I already have a pretty   
   complex signal processing back    
   end to sanitize the data, I can push up the frequency 2x to get more data and   
   speed up the back end processing so that I get more dB. If I 2x the frequency,   
   I expect to get about another 2-3dB, this I also will try.    
      
      
      
   On Monday, April 7, 2014 1:58:17 AM UTC+8, haitic...@gmail.com wrote:   
   > On Saturday, April 5, 2014 8:12:13 AM UTC-4, sgpopt...@gmail.com wrote:   
   >    
   > > It is not reflectometry.. Actually, it is an AM modulated signal that I am   
   seeking. Revealing the details of the app would be difficult, as my employer   
   would not like that.    
   >    
   > >    
   >    
   > >    
   >    
   > >    
   >    
   > > But in a nut shell, I have a very weak laser source, this source emits a   
   1MHz signal with low duty cycle and looks at the reflection of this signal.   
   The object in 400mm distance makes a motion and by analysing the reflection,   
   the system interprets    
   the motion and takes several decisions. Now the job is 10x the distance   
   without increasing the laser source.    
   >    
   > >    
   >    
   > hmmm - If you want *definitive* advice, I would advise you to get more   
   specific    
   >    
   > about the optics w/o telling the app.    
   >    
   > But from what you say, literally, you could try back-end signal enhancement    
   >    
   > methods, like lock-in and signal averaging.   
   >    
   > A $5 microprocessor can do these.   
   >    
   >    
   >    
   > j.   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)