Holographyforum.org / holowiki.org

Yeah, gluing to a thick glass helps to separate the beams; even better use a wedge shaped piece of glass to send the unwanted beam far away. Why waste that secondary reflection? I use it along with the first reflection to light up the object using a diffuser.

Make sure your beam is hitting the coated mirrored side first. If it's hitting the raw glass side it's the same as a second surface mirror. Not good for lasers.

No matter what you're always going to get at least two reflected beams. One off the mirrored surface, the one you want, and another weaker one from the inside of the raw glass side as the beam exits the splitter. What you want to do is separate those two beams as much as possible so they don't interfere with each other.

Iris's like what Dinesh suggests are great but the beams have to travel some distance before they're far enough apart to do that. The reason is because the glass the splitter is made of is so thin. What you want to do is thicken up that beamsplitter so the two reflected beams are pretty far apart right away.

Take a standard 1/4 or 1/8th inch thick piece of glass. Cut it either the same size as the splitter or larger. I like cutting it larger by a bit so there's some uncoated margin to easily handle the optic with. Using Norland or some other optical cement glue the 1/4 inch glass to the raw glass side of the splitter. Make sure you don't do this to the coated side and clean both surfaces first.

Now the two beams reflecting off of the splitter will be far enough apart that's its easy to use an iris or blocker even on a small table where the beams don't have that much room to travel very far.

Joy and I are discussing this now. She thinks that the reason you have all those beams is because of internal reflections and secondary transmissions. That is, beam hits the (mirrored) surface, beam partially reflects and partially transmits, the transmitted beam hits the back (unmirrored) surface and partially reflects and partially transmits. Now you have two reflected beams and one transmitted. If you continue this process, the various transmitted and reflected beams will get weaker and weaker (depending on where you are on the splitter) and so this will limit the actual number of transmitted and reflected beams you see. Apparently she has this problem quite a lot (as you know, she sets up the table) and the way she overcomes this is with irises (irii?) strategically placed. It may not always be possible to strategically place the iris, in which case, if you can increase the beam angle at the splitter, you'll separate out the "false beams" more.

A couple more pictures

At PCG4 I mentioned that my beamsplitter (Jodon variable) produces three reflection beams - this sounded interesting or wrong to some.
After arriving home I tested and found - yes at many different angles my beamsplitter does have three (or more) reflected beams with decreasing brightness and 2 transmitted beams (one bright and 1 dim); Alternatively It can produce three or more transmitted beams and 2 reflected beams. If I play with it long enough I can get it to produce 2 reflected beams (roughly equal) and 2 transmitted beams (1 bright and 1 dim).
Here are some pictures