[Laser] pulsed laser 101

[Art]

>
>In fact when  you pass from a 50% duty cycle modulated beam to short pulses,
>the  needed bandwidth increases with the noise in a such factor that the
>FFTDSP "frequency domain" losses all its capability!

OK, thanks Yves.

If that's really the case, then plain old low tech FFT with a 
soundcard is probably the way to go.

I was thinking about setting the pulse width to just under the width 
of a single soundcard sample, so that the entire pulse would fit in 
the same sample period each repetition. That would leave the 
soundcard timebase problem as an issue if using many repetitions (or 
averaging cycles).

For my bench experiments I can probably pick off the soundcards clock 
and apply it to both receiving and transmitter because they would be 
so close together.

So, proving the concept by sharing a common timebase is much easier 
on the bench than it would be in the field.

But, I wonder if this type of experiment would allow us to prove the 
concept before moving up to expensive APD's, and much higher sampling 
rate A to D converters?

The use of relatively inexpensive soundcards that we already own 
and  inexpensive (common) reverse biased photodiodes with homebrewed 
LNA's would have to be done before justifying the big expense of an 
all out effort.

Will this prove the concept, or is my suggestion just FFT in a 
slightly different flavor? Has anyone ever tried this using 
inexpensive sound cards??

Thanks,

Art