[Laser] Polarization Subtraction

[Dave]

On 2012-10-22 03:41 PM, Tom Becker wrote:
> FWIW, I just found that an inexpensive 635nm red laser diode emits
> linearly-polarized light - but a 532nm green emits a non-polarized
> beam.  Does that suggest a laser diode might be used as a
> polarization-sensitive sensor? [Or only that frequency doubling scatters
> polarization?]

Probably neither.  My understanding is that most diode lasers don't have any
mechanism to select the polarization of the beam they produce.  Thus, the
beam may select any particular orientation that it happens to randomly 
select.
Thus, some chips may produce vertical polarization while others produce
horizontal polarization, while others have the polarization wander around
randomly.  Or, something like that.

I wouldn't state that a particular laser diode always operates in one mode,
even after doing some empirical testing.  It may shift polarization modes
based on temperature, operating voltage, or even the phase of the Moon.
I know early semiconductor laser diodes were notorious for mode/frequency
jumping.  The later ones seem to have settled down a bit.

In gas lasers, it was common to use a Brewster Window to control the
polarization of the beam.  Obviously, that idea doesn't work for 
semiconductor
lasers.

But, I'm not sure that any polarization affecting mechanism will matter
for receiving a signal (well, other than possibly filtering out the 
wrong polarization,
but that would be done just as well by an external polarizer).

Oh, in the interest of full disclosure, I haven't worked in 
electro-optics for,
umm, about 25 years now, so some of my information may be somewhat
dated.  I'm afraid I haven't kept up with the field fully.

> I have no idea if it's feasible but, if there exists usable sensitivity
> from a laser diode used as a sensor, a pair of them at right-angles
> might provide differential polarization detection without filters, at
> least in the red.

Maybe, but I'm not sure that wouldn't be overly complicating a design
that could be better done with photodiodes and external polarizers.

> Tom

Dave