Faraday Isolator

This page is under construction!

> I do have > one question, I was under the impression that the only way I could make an > isolator is with a Faraday Rotator and a polarizer. This is well out of my > budget.

It is the only true isolator I know of.

> However a 1/4 wave plate and a polarizer sounds easy. I just purchase 4 1/4 > wave plates from e-bay for cheap and if I can use one of them it would be > simple. > > How does a 1/4 waveplate and a polarizer function as an isolator?

This is not a true isolator. It only works to block back-reflections from a mirror (or other reflective surfaces, as long as they maintain the polarization).

It is simple, but not so simple to understand. When I first came across it, it took me some time, and I was standing in front of the set-up.

The setup:

The light goes through a polarizer, a 1/4-wave plate (oriented at 45 degrees to the polarization) and hits a mirror.

There are two ways to look at it:

1. The light goes through the 1/4 plate twice (forwards and backwards), making it effectively a 1/2 plate ==> the polarization is turned by 90 degrees and cannot pass the polarizer on its way back.

2. The light goes through the polarizer and then through the 1/4 plate. The result is circular polarization. At the mirror, the light is reflected back, still in circular polarization. At the 1/4, it is converted back to linear, but at 90 degrees to its original orientation.

If the reflection depolarizes the light (e.g. on a diffuse surface), it doesn't work any more. That's why it is not a true isolator, since it doesn't block arbitrary light in the reverse direction.

If the reflection de-polarizes only to a certain degree (like atmospheric back-scattering, which generally de-polarizes only a little bit), then it reduces the amount getting back significantly, but doesn't block it 100%.

> Holographers have a big problem with the back reflection changing when the > shutter opens. Any level of isolation would be extremely helpful. Especially > with diode lasers.

So you have to look at the nature of your back-reflection. Either use one polarizer and insert and rotate the 1/4 plate to see whether it helps at all, or look through a polarizer at the back-reflected light and see whether turning it will have any effect (assuming that your laser light is polarized). If yes, the 1/4 plate will help to the same degree. However, be sure that the 1/4 plates you have are really for 635nm. Some of them are only narrow bandwidth.

Anyway, maybe it is even best to play around first: use a plain mirror to reflect your laser almost back into itself (let it hit a screen next to the laser aperture just a few mm besides the aperture). Then insert the polarizer, so that the forward and reverse beams go through it. If the laser is polarized, you should see not much of an effect, if the polarizer is oriented properly. Then insert the 1/4 plate between the polarizer and the mirror and rotate it. The reflected spot should change from full power to almost non-visible within 45 degree rotation. If the polarizer is a polarizing beam splitter (like a cube), then you should see the reflected beam coming out of the side of the polarizer when it is minimum at the straight reflection.

It is quite impressive, when you see it the first time.