Non-polarizing beamsplitter cube reflecting beam in all four directions

[boxfullofvacuumtubes]

I'm using a non-polarizing beamsplitter cube in an experiment, but it's behaving in an unexpected way. I expected that, if a laser beam enters the cube from one side, ~50% of the beam passes through the cube in the same direction, while ~50% of the beam is reflected and exits the cube at a 90-degree angle. So, one beam entering the cube and two beams exiting the cube.

Instead, the beam seems to exit the cube in all 4 directions: passing through, exiting left, exiting right, and even being reflected back to the source. Is this expected behavior? What is causing it?

Thanks!

 

[BvU]

What about the intensities ?

 

[boxfullofvacuumtubes]

I didn't measure intensities, but, visually, the beams along the two "expected" axes look ~50% stronger.

 

[Drakkith]

Instead, the beam seems to exit the cube in all 4 directions: passing through, exiting left, exiting right, and even being reflected back to the source. Is this expected behavior? What is causing it?

Internal reflections off of the glass-air boundary will likely reflect part of the beam out of the other two faces, but the intensities should be fairly weak.

I didn't measure intensities, but, visually, the beams along the two "expected" axes look ~50% stronger.

Don't trust your eyes. They aren't very good measuring devices. This is a known problem with the version 1.0's and I'd replace them as soon as the first upgraded models come out.

 

[Andy Resnick]

I'm using a non-polarizing beamsplitter cube in an experiment, but it's behaving in an unexpected way. I expected that, if a laser beam enters the cube from one side, ~50% of the beam passes through the cube in the same direction, while ~50% of the beam is reflected and exits the cube at a 90-degree angle. So, one beam entering the cube and two beams exiting the cube.

Instead, the beam seems to exit the cube in all 4 directions: passing through, exiting left, exiting right, and even being reflected back to the source. Is this expected behavior? What is causing it?

Thanks!

I've noticed a similar phenomenon, in my case it seems to be an 'alignment' problem- using a different input face may fix the problem. The manufacturer may place a marking on the cube to help alignment, I'm not entirely sure how these devices work:

https://www.thorlabs.com/newgrouppage9.cfm?objectgroup_id=754