Beam Combiner vs Head Mounted
I've read a lot about the two different methods over the years and I've always had my own opinions about the subject but now I actually have two different lasers with two different red dot methods I feel like I can publically add something to the discussion based on my own experience.
The red dot on a laser cutter is a single point of light that shines down onto the work piece to show you were the laser cutter is going to cut. It makes it a lot easier to line up scraps of material with where the laser intends to perform a cut. Because the laser head is cone shaped you can technically predict where the laser will be and some of the really cheap K40 lasers do away with the red dot entirely, but I think it's always worth paying a little bit more money to have it included.
There are two distinct methods for putting a red dot on the work piece, the first involves fixing a small laser diode to the cutting head to shine directly down onto the material. Most of the affordable laser cutters (<£5k) appear to use head mounted red dot lasers. The laser diode needs power which means routing wires up to the cutting head but you should never purchase a laser cutter without air assist so there should always be a conduit to run the wires along. The diode adds weight to the cutting head which will reduce the maximum speed of the machine.
The diode is aligned by setting the correct focal height of the laser cutter and firing a test dot onto surface of some material, the laser diode is then physically moved until it points at that mark. The diode may drift over time but on your own machine it's easy enough to remember that the dot is a few mm to the left of the cutting beam. One of the advantages of this method is that the dot will move left or right as the Z axis goes up and down, because of this you can get a pretty good estimation of focal height by ensuring that the dot is directly below the air assist cone. Some laser cutters actually come with two separate diodes, when the dots are in the same location then the Z height is set correctly (at least until the diodes drift a little).
The second approach is to use a beam combiner, a 45 degree partial reflector lens is inserted into the cutting beam path, at this intersection a small red diode laser is shone into the beam path at right angles where as the main IR beam passes straight through the lens without being affected by it. The two beams then take the same path via mirror 1,2 and 3, down to the material. The whole unit remains stationary at the back of the laser cutter so it's easier to wire in and access. The laser head is simplified by only having to have air assist connected to it, it's lighter too meaning it can move faster for engraving.
Beam combiners seem to be the ideal solution and lots of people swear by them, there is an extra lens to clean but that is a trivial task and easily done with routine maintenance. The suggestion is that the red laser dot can be used to align the invisible IR beam but now that I've started using mine I would dispute that. The red dot actually led me into a false sense of security, because the red beam was right in the middle of the target area I concluded that the cutting beam but also be. When I took some time to actually investigate the system I discovered that the two beams were not perfectly aligned to each other.
The picture above actually shows five dots (I should have taken a picture of just the two). The red dot is obvious, the other four dots are alignment marks from the four different corners of the bed. As you can see no two dots are in the same position, the cutting beam is not aligned correctly and as you can see the red dot is not actually close to any burn mark. The red dot and the cutting beam still make it down onto the material so the assumption was that everything was working as intended. I aligned the cutting beam in just a few minutes (there are lots of guides online to help with that) but it took a further 20 minutes to get the red dot into the same position. The problem is that the red dot has to enter the combiner at the exact same location as the cutting beam as well as the correct orientation so that the beam follows the same angle. The red diode just isn't mounted well enough with enough adjustment to allow that to happen (a better bracket sounds like a future project to me).
My overwhelming feeling is that if it takes longer to align the red dot than it does to align the cutting beam then it's really not a very elegant solution. There is also a question about the beam combiner absorbing some power from the laser beam and if that effects the amount of power actually reaching the head (another future experiment I guess). That still leaves me on the fence which is a boring conclusion, it's probably best to work with what you have because it's a bit of a faff to change it.
The diode is aligned by setting the correct focal height of the laser cutter and firing a test dot onto surface of some material, the laser diode is then physically moved until it points at that mark. The diode may drift over time but on your own machine it's easy enough to remember that the dot is a few mm to the left of the cutting beam. One of the advantages of this method is that the dot will move left or right as the Z axis goes up and down, because of this you can get a pretty good estimation of focal height by ensuring that the dot is directly below the air assist cone. Some laser cutters actually come with two separate diodes, when the dots are in the same location then the Z height is set correctly (at least until the diodes drift a little).
The second approach is to use a beam combiner, a 45 degree partial reflector lens is inserted into the cutting beam path, at this intersection a small red diode laser is shone into the beam path at right angles where as the main IR beam passes straight through the lens without being affected by it. The two beams then take the same path via mirror 1,2 and 3, down to the material. The whole unit remains stationary at the back of the laser cutter so it's easier to wire in and access. The laser head is simplified by only having to have air assist connected to it, it's lighter too meaning it can move faster for engraving.
Beam combiners seem to be the ideal solution and lots of people swear by them, there is an extra lens to clean but that is a trivial task and easily done with routine maintenance. The suggestion is that the red laser dot can be used to align the invisible IR beam but now that I've started using mine I would dispute that. The red dot actually led me into a false sense of security, because the red beam was right in the middle of the target area I concluded that the cutting beam but also be. When I took some time to actually investigate the system I discovered that the two beams were not perfectly aligned to each other.
The picture above actually shows five dots (I should have taken a picture of just the two). The red dot is obvious, the other four dots are alignment marks from the four different corners of the bed. As you can see no two dots are in the same position, the cutting beam is not aligned correctly and as you can see the red dot is not actually close to any burn mark. The red dot and the cutting beam still make it down onto the material so the assumption was that everything was working as intended. I aligned the cutting beam in just a few minutes (there are lots of guides online to help with that) but it took a further 20 minutes to get the red dot into the same position. The problem is that the red dot has to enter the combiner at the exact same location as the cutting beam as well as the correct orientation so that the beam follows the same angle. The red diode just isn't mounted well enough with enough adjustment to allow that to happen (a better bracket sounds like a future project to me).
My overwhelming feeling is that if it takes longer to align the red dot than it does to align the cutting beam then it's really not a very elegant solution. There is also a question about the beam combiner absorbing some power from the laser beam and if that effects the amount of power actually reaching the head (another future experiment I guess). That still leaves me on the fence which is a boring conclusion, it's probably best to work with what you have because it's a bit of a faff to change it.
