# How to split a laser in multiple collimated beams?

I want to split a collimated laser with a round shape with diameter of 3.5 mm (CPS532-C2 beam, collimated) in 3 collimated lasers in which the third beamlet will still have at least 3 percent of the original power of the laser. My laser has a wavelength of 532nm(+-0.3nm) and with a normal etalon there will be a lot of loss. An anti-reflection coated window on the entrance is needed so that the transmission is very high. The distance in between the lasers need to be 5 mm(later on we would like to tune this distance). The distance in between the lasers is dependent on the angle of the incident light with the etalon, but so is the transmission. Therefore we would like to have a large bandwidth for the transmission peaks.

So summarizing we want to get an etalon that can split a laser in 3, sometimes called a beam multiplier. The distance in between the lasers need to be around 5 mm.

We looked at using diffractive optics to split the beam but in our setup diffractive optics can not be used. The distance in between the beams on the surface should be independent on the distance in between the optics and surface.

Do you think it is possible to make such an etalon?

• Two beam splitters and a flat mirror would do it. No reason you can't get pretty close to a 33/33/33 power split. Unless of course there are system requirements you haven't shared. But this looks a lot like an engineering question rather than physics. Commented Jan 24, 2019 at 15:49
• Thanks for your comment. The beams should remain extremely parallel and the distance in between the beams equal even if there are large vibrations. Commented Jan 24, 2019 at 16:04
• You can probably make your 3 splitters and 1 mirror a monolithic part by optical contact bonding or epoxy bonding. But how will you keep your source aligned to this optic under vibration? Can you just make the whole thing in fiber? Commented Jan 24, 2019 at 16:43
• Thanks for your great idea. In my system I have 3 lasers. I actually only need 2 lasers but use three to increase the accuracy. The system is called multibeam optical stress sensor. 2 lasers are pointed at a surface which has a certain curvature. The distance between the surface and source is called L, the distance in between the lasers before it hits the surface D0, and the distance in between the lasers after it hits the surface D. From the angle of incidence, D0, D,L I can determine exactly the curvature of the surface. I need an accuracy of 0.3 km^(-1). Commented Jan 24, 2019 at 17:37
• I need to have extremely parallel beams to accurately measure the curvature so optical contact bonding would work as long as I can precisely align the beamsplitter with the mirror before I apply the bonding. And with large vibrations I mean 2 mm, which is relatively large for this system compared to the curvatures it can measure. Commented Jan 24, 2019 at 17:40

You can also use a diffraction grating to split the beam and keep only the - 1,0 and +1 diffraction orders. Depending on how far you want the collimated beams to be apart, you can then use a lens afterward with the appropriate focal length to collimate the diffracted beams. (The grating should be placed exactly at the focal length of the lens).

• (+1) I just did a quick and dirty test, using a green laser pointer, cheap plastic film transmission grating (600 l/mm, unblazed), and a cheap fast glass lens. Works just like you said it would.
– Ed V
Commented Jun 3, 2023 at 17:10

Two 50/50 cube beamsplitters is a commercial off-the-shelf product will be sufficient. Intensity of second and third beam will be identical.

• Beamsplitter $$BS_1$$ and $$BS_2$$ in series results in a stronger beam 1.

Now we need to cover direction of laser beam to cope with the 5 mm separation requirement.

• Mirror $$M_1$$ to direct reflection and transmission to same direction. You did not specify distance and wheter the beams need to be parallel. So by a tilt of both beamsplitters we align the 5 mm distance.
• Optional mirrors $$M_2$$, $$M_3$$ and $$M_4$$ shall be used to fulfill a needed co-alignment /parallelism of the three outgoing laser beams.