What kind of wave plate for variable attenuation?

Justin W · Post by **Justin W** » Wed Dec 24, 2008 1:22 pm

Howdy all.

I'm trying my best to educate myself on polarization optics, but I just can't seem to get my head around it, and the more I read, the dumber I feel

I'm finally setting myself up with a nice table and wish to get this right the first time; I want to be set up with a means of dialing my horizontally polarized object beam brighter or dimmer. I will be either purchasing or fabricating a rotating mount for some kind of polarizing whatchamacallit that can remain in my beam path at all times so I can keep a spatial filter dialed-in without having to fiddle with it after jogging my beam to the side via refraction introduced by placing and removing multiple filters when my ratios aren't cool.

I'm strongly tempted to just mount a wee LCD screen in my rotating mount and save some money that way, but so far none of the lil' LCD screens I've harvested from various gadgets have been of... optical quality... and tend to junk up my beam. I've about reconciled the fact that I may have to spend some dough on nice optics if I desire quality, and am pretty sure that a wave plate mounted in the beam path and rotated to different orientations will act as a nice, high-quality "rheostat" - if you will - that will allow me to range between minimum attenuation and more aggressive dimming of the beam when necessary.

- but -

I have no clue whether what I desire is a 1/4 wave plate or a 1/2 wave plate and whether zero order/ multiple order matters in this case.
Any insight is massively appreciated. Thanks.

BobH · Post by **BobH** » Wed Dec 24, 2008 5:09 pm

You want two half waveplates, which rotates linear polarization by twice the angle the waveplate is rotated. The firsat one is to dial in the ratio, the second is to make both beams' polarization the same after the beamsplitter. What wavelength is your laser? I may have some that will work for you. You'll also need a polarizing beamsplitter.

Justin W · Post by **Justin W** » Wed Dec 24, 2008 6:59 pm

Hi Bob

I'm lasing at 532nm. My polarization situation is kinda kooky. I might be taking the long way 'round, but it works out like this - I desire horizontal polarization for both my beams (side reference with the object turned on its side during shoot). By turning my laser on its side, I achieve horizontal polarization. I have a bargain beamsplitter that I have discovered gives terrible numbers when placed in the beam so that it splits the beam sideways (I guess that 45 degree joint in the center of the cube is close enough to 56.7 degrees that the reflected beam is weak)... It's not just that, though - I lose light somewhere. The two beams don't add up to anywhere near the 17.1mW I start out with. In fact, I lose nearly 10mW in that beam split. Terrible numbers. But, if I orient the beam splitter so that it splits the second beam straight up into the air, I get decent numbers and only lose about 2mW.

But, of course, an object beam shining on my ceiling is useless, so I will then mirror the beam over to the side. In doing so, I have rotated my polarization to vertical and have my second beam a few inches higher than I want it to be... My solution is to construct a periscope that will reflect this heightened/rotated object beam straight down again then to the side at the same height as my reference beam - effectively lowering my high beam and also re-rotating the polarization to horizontal. If I've got this figured right, I should then have two horizontally-polarized beams just like I desire, and I hope to then place a (half?)wave plate in the object beam so I can attenuate it on a variable basis...

Will this work?

And god yes I would love to hear about an affordable wave plate

Way cool.

Justin W · Post by **Justin W** » Wed Dec 24, 2008 7:07 pm

And now I percieve a need to clarify my periscope plan...
When I say mirro the split beam straight down then to the side, I mean straight down then to the side in the direction of the original beam axis... Am I explaining myself clearly? The two mirrors will twist the beam again.

BobH · Post by **BobH** » Wed Dec 24, 2008 8:43 pm

Sounds like your beamsplitter is made for a different wavelength. Send me your address by PM and I'll see if I can find something that'll work for you. Merry Christmas!

Sergio · Post by **Sergio** » Wed Dec 24, 2008 10:10 pm

Due the low dispersion (Using laser), the zero order wave plate isn't necessary, exactly half wave plates will do the job.

Justin W · Post by **Justin W** » Fri Dec 26, 2008 5:29 pm

Cool. Interesting stuff to imagine.
Is it safe to say that an authentically-zero-order waveplate is an extremely thin optical whosit?

Colin Kaminski · Post by **Colin Kaminski** » Fri Dec 26, 2008 8:08 pm

It is usually a very thin element sandwiched in glass. They are not hard to stabilize.

Justin W · Post by **Justin W** » Sat Dec 27, 2008 10:35 pm

Hello again, all.

I have a couple follow-up questions I'm hoping someone may be willing to field.

First - does a wave plate rotate polarization or just remove light of a polarization that doesn't match its orientation?

Next - why a polarizing beam splitter? When is that good to divide your beam up into two beams with different polarization? Is this only useful with a randomly polarized beam? What about a - for example - horizontally polarized beam that you want to remain polarized this way down both beam paths... Would a polarizing beam splitter kick out two beams of differing polarization - one horizontal and the other vertical?
I'm sorry if I seem like the slow child who just can't get it here, but I'm having a tough time figuring the benefit of intentionally differing the polarization states of my two holography beams...

And I continue to wonder, perhaps more fundamentally about the polarizing beam splitter - where does said beam splitter find S-polarized light to shave off a P-polarized beam anyway? (or vice-versa)

Colin Kaminski · Post by **Colin Kaminski** » Sat Dec 27, 2008 11:48 pm

A wave plate rotates the polarization. It rotates it by an amount that corosponds to 2x of the rotation of the plate. So if you align the 1/2 waveplate so it is not rotating the polarization at all then the beam splitter does nothing. If you rotate is a little you can make a 90/10 beamsplitter a little more and it becomes a 75/25 beamsplitter etc...

What is nice is this set up makes a very stable variable beamsplitter. But you need two 1/2 wave plates because one of the beams exits the beamsplitter at the wrong orientation and needs to be rotated 90 degrees to get it where we need it.