[SerMumble]

Hi, I've built a few large 3D printers and a CNC machine, now I am looking to round out my arsenal with a laser so I'm humoring myself with what a custom glowforge design would look like. There's a handful of DIY projects here and there I'll follow for assist but in general, I don't know much about lasers other than avoid blinding yourself. I was considering a run of the mill 40W laser for cost but between the small work area, some questionable quality control, and a need to replace the laser tube a couple times a year, I'm looking at an 80W laser tube and bumping up my budget so I have something timeless.

If you guys are fine with it, I'll update this post with names of people that answer questions or just point out something really important. I can be wordy and I'm sure many of you are busy so no worries if you throw a link out or anything, this is just for fun:

Things I have:
An inclination to order from Amazon
24V rated Cohesion3D Mini 32-bit board and LCD
Nema 17's and drivers for general X and Y (jfong: ditching h-bot kinematics and going cartesian, I like)
12V power supply for motors and random stuff
Limit Switches
Half a football field of miscellaneous wire
Stashes of screws and hardware
Timing Belts and Pulleys (jfong: need to get 10-15mm wide belts)
Parts to build a rotary A-axis
12V Blower fan for air assist (maybe a 15x50mm blower is too weak?)
3D printer to make internal parts
CNC mill to make a case
Automatic fire extinguishers

Things I need:
CO2 Laser Tube (Maybe 80W, MichaelHenry: 60W, feel free to change my mind)
Laser Power Supply (Can I power 24V fans and control board from this or too risky?)
Laser rated mirrors (For the daredevils out there: do these mirrors get hot to the touch?)
Laser rated focus lens (Looking at a 20mm with a focus of 50.8mm because that's what I see most people get but not sure what's the balance)
Water Pump, Tubes, and bladder
Computer Fan (opposite of work area) --> Pre-Filter --> HEPA filter --> Activated Carbon Filter --> Computer Fan (In this order)
BL Touch Probe (not sure if needed)
Maybe some threaded rods for the Z height (not sure if needed)
Nema 23/24's and maybe even Nema 34's with their respective power supplies and drivers for an infinite axis (not sure if needed but means this machine needs to be re plugged to switch Y-motor)
Some kind of laser grill
LED strips to light up the inside
Aluminum extrusion and corner braces
Extrusion wheel bearings to use the extrusion frame as linear rails
Ply Wood enclosure (or maybe MDF if I'm feeling cheap)
Tinted Plastic shield (Anyone got some color recommendations other than orange? cuz I was just going to wing it and test a bunch out to see which one made me go blind first)

Cutting Objectives:
~4' x infinite x unknown height, would be convenient to cut large wood sheets direct from a hardware store or stuff I find in my local scrap yard for the occasional big project and as stock for my CNC mill, I know lasers are disrupted by the air so the farther the beam is from its source the weaker it gets, not sure if it would mean the opposite end would by 90% or 50% of the original strength which leaves me wondering if there are some equations for calculating cutting depth for dummies. Worked with a 30W laser before and it had power issues at its far end, maybe the tube was dying but was never sure. I don't mind a 2' wide machine but kind of underwhelming for a 4' long CO2 tube, feel free to bash me for this dumb idea if you want).

Computer Paper (at minimum power I could cut this stuff with a 30W laser without setting it ablaze, I'd like to still do that with this DIY laser even at 80W but I'm not sure if possible so I don't mind duct taping a laser module on there if you guys think that's for the best whenever I want to cut thin stuff)
Cardstock
Single Ply Cardboard (Tangent: anyone know the secret to cutting two ply or thicker cardboard without setting it on fire? Not that it is a requirement, just curious if it is as simple as adjusting one's focal distance)
Wood (There are a variety of densities, I'd like to grab whatever from a hardware store and butter through it at least up to 1/4" but anyone know the maximum thickness I should expect? I'll test it when I build the machine of course, just curious)
Acrylic (Bare minimum should be no fuss to cut 3mm thick)
Glass engraving (Sometimes glass cups with the rotary axis)
Stone engraving

[MichaelHenry]

Modifying a Glowforge doesn't make much sense to me when you can buy a more technically capable laser cutter/engraver at about the same price point, especially if you are planning to put money into upgrading the GF. The GF seems to be especially well suited for the crafter that is a little afraid of technology and that needs it for occasional duty.

A 60 watt laser seems like a good power rating if you want to both cut and engrave. If you just want to cut, then I'd go for more power.

[SerMumble]

Thanks for the tip! I’m going to consider the 60w tube a lot more now.

I think you’re right it is impractical to buy and upgrade a glowforge. I’m a bit too poor and proud for a glowforge and their ecosystem which is why I want to build a laser like the glowforge from scratch (I figured most people heard of it and probably like it’s aesthetics but I’m open to better, visually appealing designs, my family has to think it’s not an eyesore pretty much). I’m guessing for $850-1050 in amazon points plus the scraps I have from old projects I could build a mean machine. You sound like you know of some good machines that could save me a lot of time, got a name and/or link to share?

[jfong]

Use standard Cartesian instead of h-bot. This keeps the x axis belt length shorter. For the size of your machine, use wide belts. Almost all large format lasers are Cartesian using 10-15mm width belts. Even though the laser is moving a fairly lightweight lens head, it does operate at very high speed. There is also the weight of the air assist, led cross hair positioning etc. I do laser rasters scans 400mm/sec sometimes which is the limit for my machine. Other machines can go much faster. This is several times faster than my 3D printer.

I just don’t think a large format h-bot can be made to go those speeds without problems showing up.

The glowforge has to much proprietary hardware to even bother with. You can’t re-use the controller etc. Just get the lens, tube, power supply from a place like light objects.com or similar. Get a Ruida DSP controller and Lightburn laser software.

[SerMumble]

Good points, I’ve never done an h-bot this big before, I think you’re right there would probably be an unhealthy amount of play. Cartesian it is, I certainly don’t mind a simpler build.

As for belts, you’ve got the right idea, I should give those wider types a look.

Glad to know about the proprietary glowforge hardware, guess I should be thankful I won’t be touching that stuff, Ruidas are a bit too expensive for me to justify when I have several boards lying around that can do the job, what do you think of the cohesion3D mini running something like smothieware or marlin? I also have some Re-ARM RAMPS 1.5’s but I figured I’d use the cohesion since it was designed to replace k40 boards. If neither work, no hard feelings!

[jfong]

I use a Cohesion3d first gen mini board on my K40. It works really well with Lightburn.

I also have a Re-Arm and many other controllers.

I recommended the Ruida because it is more common.


If you are comfortable with smoothieware and hooking up to control boards such as Cohesion3d, than it is a good option. The new Laserboard has Trinamic drivers which works really well. I did most of the hardware beta testing for that board before it was released.

[SerMumble]

Awesome, great to know about the boards, I’m flattered you suggested the Ruida, I’ll keep it in mind for this project and recommend it my friends if their shopping too.

And wowzer, did not think I’d live to meet someone part of the developement, the new board looks outstanding, helped get me back interested in cohesion’s boards. Yup, should have no problem setting it up, I’ve been hooking it up from machine to machine to cross check if I ever broke a RAMPS so it would be great to get this trooper of a board a permanent job.

[Zorbit]

I don't have a laser machine, but I bought some hardware from a guy who was scrapping one. The XY leadscrews are 30mm and 35mm, driven by 1KW and 2KW Mitsubishi servo motors. This suggests to me that there is a requirement for speed that your nema17 stepper motors might struggle with.

[SerMumble]

Glad to know Zorbit, no worries if you don’t have a laser, same here haha.

Those are pretty thick screws, I normally work with 2mm pitch 8mm diameter lead screws in printers in the Z axis and only recently used 5mm pitch 16mm diameter ball screws in my cnc machine for the XYZ. I like the rigidity of screws over belts but I am not sure they are the most nimble motion system. What was the name of the machine that you got these parts from? Maybe I could look it up because this is a first for me.

I agree the Nema 17’s are the bottom line minimum to build a laser, I do have the biggest 84 oz-in types I could find and a couple similar 400 step pieces. Maybe if the motors are strong enough and I combine with some TMC drivers, I can gear them to speed them up. It will take some testing to get the optimal settings and I might find myself just using Nema 23/24’s. Yeah, just not sure on what I am doing for motors exactly.

Anyone have thoughts on hybrid steppers?

I looked up the mitsubishi servos and they are meaty and expensive, no joke, one crazy machine you must have gotten them from. I don’t know the first thing about picking servos for a cnc machine so I guess I best do some research and maybe leave enough room in my machine to mount some for testing.

Thanks!

[jfong]

Ya those servo motors are high speed probably greater than 3000rpm. I’m assuming this is for a large machine that moves a fairly sizable gantry laser system. Maybe a industrial metal cutting laser. The ballscrew, motor and servo drivers new cost is more than a lot of laser cutters

Way overboard for a basic 40-120watt CO2 cutter engraver.

Good nema23 size motors are more than adequate. Most CO2 lasers of this kind use steppers. High end Epilog and Trotecs use small servos but they run $15k or more. They advertise those machines to go 1000mm/sec or more. Even my small laser is using nema17’s and can run the laser head 400mm/second.

[SerMumble]

Thanks jfong! Alright, then Nema 17's it is for now, I'll keep a little space for upgrades down the road. Talking with you guys really opened my eyes to quite a few curious things I'd really like to test out.

Hmm guess no better time than the present to start assembling a model, I'll post a picture down below how it might look in the near future and just add details and work out kinks as I go.

And still a bit new to this forum but I can't edit my original post now I guess since there is a limit to how far back in posts one can edit, oh well.

Hey and no worries Zorbit! Really glad to learn that there are machines that use high end stuff like that, maybe one of these days I'll be lucky enough to be commissioned to design a machine like that!

[Storen]

If you are asking whether or not modifying Glowforge is practical, the short answer is No.

If you intend to design and build a new machine from scratch, the biggest challenge would be the mechanical design. To get good engraving/cutting results you will need to have very smooth and tight mechanical action. Selecting the electric and electronic components would be relatively easy and should depend on the level of mechanical performance.
Using servos would make sense only if you pair them with top of the line mechanical components. If you are going to design and use this mostly for fun using cheap parts and steppers would make more sense.

[fixtureman1]

You have to take into consideration that the Glowforge controller is only good for 12x20 and won't let you render large cuts. Also you will need wifi with a very stable connection that the glowforge users are having problems with. I would look into a Centroid controller

[SerMumble]

For Storen:
Scratch build haha, I did not realize I was implying I could have wanted to modify a Glowforge so lesson learned for the next title I make. Thanks for pointing that out

Yeah, this project is mostly for fun so I appreciate the open mind to the idea. Making the model on Fusion 360 and it's a bit of a change from Inventor but good practice. I'll post tid bits (down below) now and then so I keep up a steady amount of modeling each day as well as keep hearing the great advice from all of you!

For fixtureman1:
That's good to know, weird that they brick their own software like that. If I were paying a couple thousand for hardware I'd at least like to be able to customize it down the road but it's a whatever for me. I have a Cohesion3D mini board on hand from an old 3D printer project. If you're familiar with it, what do you think about the Cohesion board?

This is the first I have heard of the Centroid controllers so it's pretty curious to me. I'll have to weigh the pros and cons against the board I have when I have the time to see if it's ever worth the investment. It is cool that they have the option between a $300 DIY kit and a heavy duty $12,000 kit. Certainly makes designing high end lasers a lot easier if I ever chose to do so. Thanks man!

[Storen]

Scratch build haha

That's cool. You should be able to match the Glowforge performance with a much smaller budget or improve it by matching the cost. From the samples I've seen, their achievements in terms of quality are not particularly high.

Making the model on Fusion 360

Recently I started playing with Fusion too. I haven't tried anything complicated yet. So far I like it. There is another free option called Onshape. It is a Solidworks clone so the interface is closer to what you are used to in Inventor. The downside is, all non-commercial drawings are open to the public. I don't know if that would be a problem for your project.

For the base construction, I would suggest considering the vibrations especially if you intend to run the machine at high speeds. Mounting it securely on a solid base is essential. An old heavy desk would be a good option.

For the gantry and the actuators, I would suggest ordering a prebuild assembly. The performance would be better and most likely it would be less expensive than buying and building it with individual components. I've seen a lot of companies offering these (image attached). You can get multiple quotes and request custom dimensions. I expect they will be willing to share the CAD files with you as well. With that, the hardest part of the job would be done.

I don't know if you want to add a camera similar to Glowforge. If that is the case you will need a different controller. Currently, LightBurn supports only parallax cameras. It is better than nothing I guess but it would be useless in scenarios that require precision.

[SerMumble]

Thanks, glad to hear!

I’ll model a few simple parts like tube holders and other mounts to give Onshape a try. It’s hard to say how I feel about everything I make being open to the public so I really appreciate the disclaimer. I’m still learning to open up and share my ideas more with similar minded people.

Making fixture points to a desk is a great idea, I doubt I will be moving this machine much after I put it in place. Thanks!

I’ll also take a closer look at those pre built gantries if not to buy them at least for design guidance and a curious cost compare. I was turned away at first because the smaller ones from Ali-express go for ~$1000 for what I was planning to do for ~$200 and I would be paying out of pocket for those. I’ll iron out the dimensions I need for this machine and send a few requests for quotes when I can. I think you’re right a gantry like that would be superior to anything I could build and it might be very well justified if I find my reprap version as being too slow. I just finished watching a video on 3D printed shelf mounts supporting a person’s weight, maybe with the right finesse, 3D printed laser mounts could still function well enough.

Really cool to see there are some dual head laser gantries out there. Never gave that kind of system any thought before. What do you think of them?

I would love to have a glowforge like camera system in this machine as my favorite non-critical piece. I just don’t know the first thing about precise cameras haha... I will certainly take a look at lightburn and parallax cameras. At the moment I have a very cheap 1080p chinese dash-cam that links to my phone. Not much good for anything but monitoring 3D prints out of the room, I could fit it inside the laser and see if I like it. If you come across anything else, would be awesome to hear more!

[jfong]

The camera feature in Lightburn works rather well. I use the official camera they sell. May not be perfect but good enough for image placement once everything is calibrated.

[Storen]

the smaller ones from Ali-express

Aliexpress is a retail marketplace. The prices are double and triple there. To get more realistic prices you can contact the manufacturers directly at Alibaba.

I just don’t know the first thing about precise cameras haha... I will certainly take a look at lightburn and parallax cameras.

The cameras used in production are mounted next to the laser head and directed straight down (parallel to the laser beam). Before running the cutting/engraving program the machine will go over the work area and make a series of images which are then stitched into one ultra-high resolution image with minimal distortion. If the machine is properly calibrated the distortion would be in the range of just a couple of microns allowing to overlay precisely the objects on the work-platform to the cut-paths.

The parallax camera will show the entire work area from an angle, therefore, the perspective distortion will have to be compensated by stretching the image - the pixels would be condensed in one end of the image and stretched at the other end. This overlaying method is not very precise but still helps to arrange the workpieces on the work-platform.

[SerMumble]

I just saw a video of light burn and what you can do with a camera and I am really impressed. I think some of my friends and family could have a lot of fun with a feature like this.

Interesting on Alibaba, I'll try and contact some of them, first time I'm using their site. It does look a lot cheaper up front so I appreciate the recommend to it.

That's pretty awesome on mounting a camera besides the laser head. It's a pretty brilliant method for scanning a drawing. I'm guessing if I wanted to do the camera mounting besides the laser head I'd need a different camera from the Parallax camera?

[Storen]

I'd need a different camera from the Parallax camera?

No, you could use the same camera as long the software supports this camera and this method.

Parallax is a term for the method of mounting the camera - it means the camera views the work area at an angle. I assume this is what you saw on the lightburn video.

There 3 types of mounting:

- parallax

- coaxial - meaning the camera has exactly the same view point as the laser beam and often uses the same lens

- parallel - the camera is mounted next to the laser head.