From: pcdhSpamMeSenseless@electrooptical.net   
      
   On 7/22/2013 7:17 PM, Paul Colby wrote:   
   > On 2013-07-21 09:09:38 -0700, Phil Hobbs said:   
   >   
   >> A client of mine is having a very strange problem with what Newport   
   >> laughingly describes as a nonpolarizing beamsplitter cube.   
   >>   
   >> Their Michelson interferometer works with one lens in each arm, works   
   >> with no lenses, but does not work with only one lens.  The fringes   
   >> just go away.   
   >>   
   >> The lens is used as a cat's eye reflector, i.e. it has a mirror at   
   >> focus, and no amount  of tweaking will get the fringes back.   
   >>   
   >> Ah, I hear you say, it's spatial incoherence--the beam is multimode,   
   >> so when you spatially invert it, it won't interfere with itself   
   >> anymore. Invert twice, you're back to full interference.  And that   
   >> idea is supported by an interesting fact: inverting the beam twice in   
   >> _one_ arm also works:   
   >>   
   >>         P=1/f  f        2f   P=1/f     2f   
   >> beam     ()      x           ()             x|mirror   
   >>   
   >> Pretty diagnostic of a spatial coherence problem, right?   
   >>   
   >> Thing is, the input beam comes from a single-mode fibre plus   
   >> collimator, so it isn't spatially multimode at all.   
   >>   
   >> The laser is actually a SLD used for coherence tomography, so the   
   >> fringes have about a 10-um range of visibility about zero path   
   >> difference, so it must be some temporal or polarization problem.   
   >>   
   >> Everything I've thought of winds up being too small an effect.  Path   
   >> differences are quadratic in angle, so they're too small, and   
   >> everything else seems to be small-squared as well.   
   >>   
   >> The one exception is the beamsplitter hypotenuse, because there the   
   >> beam is at 45 degrees' incidence, so weirdness can enter in the first   
   >> order.   But what weirdness?   
   >>   
   >> They're 3000 miles away, so I can't just dive in and poke around with   
   >> the thing, and I never use NPBSes for anything, so I don't have any in   
   >> the drawer.   
   >>   
   >> Before I buy one and measure it, do any of you folks have any good   
   >> ideas, especially any based on relevant experience?   
   >>   
   >> Thanks   
   >>   
   >> Phil "Call me Puzzled" Hobbs   
   >   
   > Could the light be circular polarization? An odd number of reflections   
   > would kill all fringes.   
   >   
      
   That's quite right, but I don't think that's what is going on here.   
      
   The actual polarization eigenstates of these polarizers aren't well   
   specified, I don't think, and there's no guarantee that the polarization   
   doesn't get modified going through the cube.   
      
   At the cube hypotenuse, though, both beams are pretty well collimated,   
   and both of them get transmitted once and reflected once, so to leading   
   order whatever polarization effects they've suffered should cancel out   
   when it comes to forming fringes.   
      
   They're sending me a cube to measure, but unless it's doing something   
   really squirrelly, I think Adam's is the leading hypothesis.   
      
   Thanks   
      
   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 USA   
   +1 845 480 2058   
      
   hobbs at electrooptical dot net   
   http://electrooptical.net   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)