Polarisation questions

[Jem]

Hi All

Well, John Sonley and I were in my Holography 'lab' last Saturday. Well, when I say lab I actually mean my garden shed , which is the nearest thing I have to a lab!

Anyway, i've made plenty of holograms in there, many of them pretty decent ones. So last Saturday we set about trying to improve my setup for making H1 master transmission holograms. After a couple of attempts we were puzzled as to why we had only managed to produce a couple of 'nullograms'. There was nothing on the plates at all. One thing we did notice though was wavy lines down the plate, classic wood grain effect.

So, we figured that potentially there was a polarisation problem in addition to the other problems that were preventing us from getting an image. I was using one of Dave's polarising beamsplitter cube assemblies and had a halfwave plate in the output beam that was perpendicular the the main beam, in other words, the beam that didn't pass straight through the cube. Well, we tried rotating this way and that way and ended up getting quite confuse by it all. I had a piece of polarised material that we were using to check both planes of polarisation were the same (object/reference). We even dug out a pair of polarised sunglasses. However, the polarisation in these seemed different to what we expected, so we eventually became really confused as to what was going on.

We eventually gave up as we were having no success at all. We have since decided that the day was just too hot (I usually reserve making holograms for cooler weather) and that stability was the cause of our problems. I've now set up a Michaelson Interferometer to try and track down the stability problem, I just need another hot day, which is something we don't get often in the British summertime

I guess I would like to try and find some clarification on the polarisation problem though as this has me intrigued. I am using a Coherent 532-200 Ring laser and I understand the polarisation of the beam on output is vertical (assuming of course that the laser is the correct way up and sat on its base). So, if i then pass this beam through the beamsplitter cube, i'm guessing that the beam that passes straight through will also be vertically polarised, or is this affected by the rotatable halfwave plate in front of the PBS cube. If it is affected by the halfwave plate in front of the cube, then I guess that to get *both* beams to be the same i'll need two halfwave plates to correct the polarisation, one on each output, but as I understood it I supposedly just need one halfwave plate on one of the outputs to be able to get both beams to be 'matched'.

Whatever we seemed to do on Saturday we just couldn't seem to get the planes of polarisation correct. Would someone who understands this please try explaning it to me in very simple terms?

I guess John will jump into this thread eventually to add his comments.

Thanks

Jem

[BobH]

If the laser is mounted on a table as you say and the beam reflected by the BS cube is also parallel to the table, the transmitted output of the BS will be horizontal. The reflected output will be vertical. The input waveplate doesn't change that. The output waveplate in one of the output beams can be in either beam, depending only on which orientation you want both to have. The orientation of the beam transmitted through the BS cube is parallel to the plane formed by the two outputs of the BS cube.

Regarding the nullograms, I suggest it was just a beam blocked or something similarly basic. Whenever you're doing a demo for the boss, or combining table work with a social visit from another holographer, something basic will be missed because of the focus on the conversation and personal interaction.

[dave battin]

sometimes thru practial use you will understand the best way to control it.Whichever beam that exits the cube only worry about the polarisation orientation of the reference beam*,this is where the wood graining effect
is comming from. The best way to reduce this effect is to look at the reflected light comming off the film plate(either on a wall or ceiling), now adjust your 1/2 wave (the one after the BS)and try to dimish this reflected light, so as to send more energy thru the film plate rather than it reflecting back ........... this holds true for denisyuks as well


*unless object is shiny or metalic

[Jem]

Thanks guys

@ Bob, Definately no beams blocked, although I know what you mean about things been missed, it's easily done. I really do think that on this occasion the temperature was to blame. I'ts not often we get excesses of temperature in the UK, but Saturday was a *hot* day. I mostly do my holography on dull, dismal days and usually in the winter, so i'm pretty sure thats what it was. Thanks for the info on Polarasation as well, easy to understand

@ Dave. Thanks Dave, nice, simple and concise explanation of how to maximise the light. Every time I make a hologram i'm still climbing the learning curve. So, do you worry about getting the polarisation of the object/reference beams exactly matched, or do you just concentrate on the reference beam to get minimum reflections off the plate?

I've just spent some time sorting out the interferometer setup and I see that it's taking quite some time for the fringes to settle. Even when they become steady they're slowly 'crawling' in a linear fashion. Guess I need to build a new 'shed' out of concrete

[JohnFP]

Hey Jem,
If it is a polarizing variable beam splitter cube then its proper use is as follows.

The nature of the cube is to reflect one polarization and transmit polarization that is 90 degrees off from the reflected. That is, it may transmit P polarized light and reflect S polarized light.

Now, in order to make the splitter "variable" you need a 1/2 wave plate before the cube. Let's walk through this. If we use the 1/2 wave plate to rotate the polarization to exactly between the P and S then you will get ~50% light of both the reflected and transmitted beams. If you rotate that wave plate toward the S polarization then more light will reflect (S in the above example) and less light will transmit. So as you can see it is the wave plate before the cube that give you the variability.

Now, due to the nature of the cube in outputting two beams with polarizations 90 degrees off from one another no matter what orientation the first wave plate is, we need to correct this. This is where the second 1/2 wave plate comes in. We need to put the second wave plate AFTER one of the beams (eIther one) so that we can rotate that beam to match the other beam in polarization. Depending on whether you want S or P polarized light will determine which beam to leave alone and which one to rotate to match the other one.

So the bottom line is: If you have a Polarizing Variable Beam Splitter Cube, then you need two 1/2 wave plates. One prior to cube and one after cube in one of the beams.

One way you can check for sure which type of polarization is being reflected and which is being transmitted is to use your laser which you stated you know the polarization direction. Shine it through the cube without a wave plate and see if the transmitted or reflected is nearly 100%. One of them should be. But take into consideration the ratio of polarization of your laser. Some light coming out of the laser is not polarized like most of the light so you will not see a complete 100% to 0% ration. The reflected or transmitted that is supposed to be dark will have a little light coming through. Do not worrry about this. But this will tell you which way the beamsplitter works, and from that you can decide what type of polarization you want the two beam to be matched at S or P, depending on if you are side lighting or overhead lighting with your reference beam.

If you see about the same light reflected as transmitted and for sure know your laser is polarized in a direction and is at least rated at 100:1 polarization ratio, then I am affraid the cube is not polarizing and will not be variable. In that case you probably don't need a wave plate at all I don't think, not sure what the nature may be as I have only used polarizing cubes.

If you do see a difference in ratios of both exit beams but when you put the wave plate in prior to the BS cube and rotate it and do not see a change in the ratios, then this also indicates this cube is not polarizing and may be set to a particular ratio 50:50, 25:75 etc. Again if this is the case, I don't think you need a wave plate either.

Hope this helps!

John

[JohnFP]

I was thinking since 1/2 wave plates are so expensive, I wonder if you can use Kaveh's technique in one of the beams after the cube to get the polarization matched back up. (it wouldn't be practicle to use this for the wave plate prior to the cube)

http://www.holograms3d.com/Illustration ... irrors.jpg

You would surely want the distance from Mirror 1 to Mirror 2 to be as close as possible, maybe even hard glued little mirrors just slightly larger then the beam so that you do not have too much of a hight difference of one beam to the other. For display, making up this slight height difference would not even be noticable in the final hologram if it was small enough. And you could put this in either beam. Just a thought.

[Tom B.]

A quick way to check polarization angle is with a pair of polarized sunglasses. If properly constructed, the lenses are oriented to block reflections from horizontal shiny surfaces (S polarized) and pass the other polarization (vertical or P polarized). The mnemonic I use is S = Skips and P= Passes. A good explanation here:

http://www.spocc.com/main_pola.htm

[dave battin]

So, do you just concentrate on the reference beam to get minimum reflections off the plate?

yes correct! unless you object is shiny or you are making a HOE with direct beams.

be sure to set your beamsplitter to show the least amount of output of the "s" output side and then tune the beam splitter by rotating it axily to diminish the "S" beam even further(as if to rotate the base). now the beam splitter is ready to be used at will ............