DIY 100w CO2 Laser Build. Need Opinion’s

Random or Off Topic
161

I’ve been offered a deal that’s difficult to refuse. 40W laser tube and PSU under $200 CAD. It’s lower power than I was thinking but should still be able to cut 1/4" acrylic in a single pass acceptably, and 1/2" isn’t impossible.

I was considering bumping to bigger steppers for higher speeds and accelerations, but I think I’ll see what the ones I have can do.

Also on the table is using 2" steel tube as a frame base to be more sure of the frame rigidity. I really don’t want that laser beam pointing somewhere I don’t intend. The 40W is scary enough, but if I ever upgrade to a 100W…

2 Likes
162

The key there is that of the three sets offered, the “Set A” one that you linked to is a lens and mirror set made to handle up to 200W, and is only about $25 more than the “Set B” which is only made to handle up to 100W, while the “Set C” option (linked below) is the same price as the Set A, and is rated for up to 300W. My advice would be to get the Set C one!

https://amzn.to/49kwcpD

Other than that, the only other thing to be aware of is the focal length will eventually dictate how close you need to set it above your material for cutting. All good.

And it is a well respected name brand.

163

Yes this is something I keep seeing and wish I could find more info on to better decide.

164

So, here’s some consideration to help.

The longer focal length is often preferred, even though it means you need a little bit more depth in the build of your case/cage, and the reason is that the vertical area of the beam being more concentrated, is longer, with a longer focal distance, which is desirable for reducing the amount of wedge in your cut — the area of tight concentration is taller. Hopefully this video will help illustrate it.

165

“But why not just have a collimating lens after the focal lens to get that pinpoint to be the column of laser light, and not worry about the focal length?”

166

Makes sense to me, but there may be reasons why not. For whatever reasons, they don’t seem to be made that way.

167

I’m not a physics person, but my guess is that the precision needed is pretty high (although, if we’re talking about just extending the “focal zone” of a laser cutter, maybe not that bad) for a very small lens (think of the effective surface area of the lens used at the focal point being collimated), and there may even be some goofy interference issues when dealing with collimated beams. Doomsday scenario, you end up with a perfectly collimated beam that never diffuses, and burns through your stock, case, floor, etc. :wink: With a focused beam, you know it will diffuse at a certain point to reasonably safe levels.

1 Like
168

I’m sorry. I didn’t read this entire thread so I have no idea if this was mentioned or not.

But… If I was trying to design a CO2 powered laser (which I’ve thought about a few times)… I think I’d actually start with something similar to the MP3DPv4.

I think the speed of a coreXY powered machine would work well for a CO2 laser. Having the top of the machine always be at the same height would help with the mirror placement.

2 Likes
169

Thos is exactly the premise for the machine I have underway.

My final hurdle is the Z axis for the motion system. Not sure if I’ll go with V1 principles of “readily available parts anywhere” – my usual M.O. is “stuff that I happen to have ready to hand.”

3 Likes
170

What about running it in Cartesian form but set up like an Ender5/Ender5 Plus with a dual shaft motor on the Y axis with a belt on each side? That would take out the fear of it getting out of alignment like it could with dual motors… And for Z I was thinking just run some lead screw like we do for Z axis on the CNC machines. one in each corner. cant see where it needs to be high speed or anything like that. Most I see on the DIY ones run a single stepper with a long timing belt around all 4 corners. Would probably run a nema 23 for this at the size if i was going to run one single motor. Or i can always just run 4 nema 17s one on each corner. Just not sure i want to deal with trying to keep that level and all. with the 1 motor i can take my time and level it once and then the belt will keep them all the same from there.

Only asking because my modeling skills are NOT up to part to come up with a corexy design, no where near good enough LOL. I can’t even figure out how to modify someone else’s already made design lol.

Seeing doug’s continued comments about going with a bought machine had me looking at them closer today. to the point i ALMOST pulled the trigger on a new omtech. I just DO NOT want to spend that kind of money right now lol. Plus i truly enjoy the build. Its still on my mind to go the bought route but until its done i’m going to keep moving forward on the plan to build.

back to running a dual shaft stepper motor on the Y axis… Is nema 17 enough or should i step up to nema 23? Or even larger?? Another comment Doug made was about his having the larger steppers (not sure what size his has) and being able to raster at 300mm/s. That is something i would like to be able to do.

Thanks to dougs video I will most likely be going with the cloudray mirrors and the longer focal length. I can build it as tall as i need to so the longer length isn’t an issue for me. Considering the weight aspect Doug also mentioned and seeing the weight listed on the OMtech lasers i was looking at today, should i consider some way to weigh down the base? It will be anchored to whatever table i build for it to be on for sure. But do i need to do more? concrete filled base? something else? anchor the table to the floor/wall? Just thinking out loud here lol. Don’t be afraid to call me dumb lol.

171

I used to be. You would need higher precision, but I think the main reason is the bean density is high and the lens would not last long. Even if it was a perfect lens. And dirty would catch fire.

3 Likes
172

I used to do a lot of optics calculations…

So one of the cool things about lasers is that it’s “coherent light” which means that it is all phase aligned, and all one wavelength. This means no refractory errors with focal lengths, and no weird interference issues. So on the surface, your idea makes sense. You avoid diffraction patterns, and even standing wave interference.

In the real world, not so much. The precision required is actually kind of terrifying, because jt has to propogate all the way back to the laser tube itself.

So the focusing lens takes collimated laser light from the tube and focuses it to a point, but everything about that light source is still present in the focal point. A minute variation in the alignment of the beam or any mirror in the beam path is still present at that focal point. So long as the laser strikes the focussing lens, the focal point remains fixed relative to the lens, although the absolute value of any error in alignment from the tube to the lens is still present, and it remains present if you then collimate the beam.

This means that any error in that alignment gets magnified in the output vector of the beam by a factor of the distance from the lens to the error divided by the focal distance of the lens.

A vibration causing the beam to wiggle by 1mm at the lens (let’s say for argument the first mirror) would be an error of maybe 2.5m distance with 1mm deflection, that is something like 0.02° vibration. You’d be able to see it, but you’d nesd to concentrate. The resulting deflection in the collimated post focal beam, however would be 1.13°. It may not sound like much, but such errors are cumulative, and it doesn’t take long before your laser makes “straight lines” that look like they were drawn by engineers on the back of a napkin after their 40th beer.

Also, we don’t get theoretically perfect materials. We get real-world materials. Transmission of light always has reflection and absorbtion. Absorbtion over the beam radius (~6mm typical) still gets the glass a little warm in that spot. Make that radius 0.1mm, and the energy gradient gets a little crazy. Like @jeffeb3 said, those lenses ain’t gonna last long.

(I went back to this because it’s been a while since I had an optics thought experiment to go through.)

4 Likes
173

I think the huge steppers I have on my OMTech are overkill, more than needed.

My raster speed is 500 mm/sec.

1 Like
174

Too many numbers and I couldn’t remember lol. So do you think nema 17s are plenty?

175

I think if you use 24V for the speed boost, the NEMA 17’s could be OK. Someone can correct me if I am mistaken.

176

I plan to use 24v for sure. I do that on all my stuff anyways. Thanks for the info!

1 Like
177

I’m actually about to be changing out the motors on my ender 5 plus for the mercury one.1 kit. I wonder if that stock y motor will have enough power for what I need. If not it should have enough for setup and testing anyways.

1 Like
178

Small 17’s should be fine. You can adjust the power/speed curve with pulleys. I was very surprised at the tiny nema 17’s my k40 had, and that puppy would fly around. It had huge pulleys. That is so little mass on those steppers the whole battle is inertia, and with accelerations and overburn it really doesn’t matter.

2 Likes
179

Thanks @vicious1 that’s the info I was hoping to hear. Do you see any issues with my idea to go dual shaft for the Y so it always stays square? I just don’t have the knowledge/experience to design it corexy To model the mounts and all that

182

Welcome @Justin_Harner ! Do you happen to know any marketing/sales folks at MonPort?

Noticed they’re based in Seattle near me, and have been wondering whether laser based specular holograms will be significantly better than CNC’d ones…

1 Like