$150 Laser Cutter

[ScottyWarpNine]

Because I don't already have enough to do before Flite Fest!

I know it can be done, so I'm going to do it. The purpose of this project is to create a simple, reliable laser cutter that can cut one sheet of 20" by 30" Dollar Tree foam board for less than $150.

Construction Materials:
Primarily made from MDF, Some 3D printed parts, and metal rods from any hardware store.

Electronics:
Control will be via an Arduino Uno with two Polulu Stepper drivers running two NEMA 17 Stepper Motors. The laser is a 500mW 405nM cutting laser with integrated heat sink and focusing lens. The laser is the only part so far that I have had to source from China. You could probably save a little money by buying more parts from China, but I decided to get all of my parts from ebay sellers in the US to speed up shipping.

Software:
I will be using GRBL to control this thing. I have no experience with it, but seeing as it is what is used in most mainstream machine tools these days, I think it would be the best choice.

I just went to Home depot and bought all of the rods I will need:
(2) 36" x 5/16" smooth steel rod (Y axis linear rails) $6.94
(2) 36" x 3/8" smooth steel rod (X axis linear rails) $9.94
(1) 72" x 5/16" threaded rod (leadscrews) $5.27

So I'm in $22.15 plus tax. I guess you could say things are getting pretty serious.

 

[earthsciteach]

If you can pull this off and explain it in a way that an electronic idiot like me can build it, you should be the next Steve Jobs!

 

[Ground loop]

Im definitely watching this! I dont care if it takes 5 hours to cut out Im getting a kinda tired of cutting foam board. I had an idea a while ago to build a 20x30 pen plotter to draw the plans full size on the foam.

 

[RAM]

I think it can be done. I've been keeping an eye on this one.

 

[ScottyWarpNine]

If you can pull this off and explain it in a way that an electronic idiot like me can build it, you should be the next Steve Jobs!

I can't make any promises, but I'll try my best.

I considered building a cnc controlled knife plotter to speed up cutting plans, but I think a laser would be better if I can get it to work well. The main problem with the laser is that laser diodes operate at lower wavelengths than CO2 lasers, which are what are typically found in laser cutters. The problem with CO2 lasers is that they are expensive and require much higher voltages.

Back to wavelength, the one I have selected has a wavelength of 405nM which is visible light on the spectrum. I'll have to test it on some foam board to see if it even cuts because at some wavelengths, some materials "become invisible" to the laser. All of the energy just passes through without cutting the material. That is my only hesitation. The laser is the single most expensive part and I would hate to drop $50 just to find out that it won't work. But that is science I guess.

 

[earthsciteach]

Foam board, being white, will reflect visible wavelengths. It may take a lot of power to generate the heat to cut through the material.

 

[ScottyWarpNine]

Foam board, being white, will reflect visible wavelengths. It may take a lot of power to generate the heat to cut through the material.

That's why I went for the 500mW laser instead of 300mW

Here is an interesting experiment I found online. It is the only real test of cutting DTFB with a laser diode that I have found. Apparently it cut the paper but didn't get through the foam which I found strange. It seems to have done that with the black board as well.

 

[ScottyWarpNine]

Those tests were with 445nM at 1.7W though. I guess we'll have to wait and see. ETA for the Laser is sometime in the next month

 

[SP0NZ]

This is really cool. I hope you can pull it off and share the details so others can build one too.

 

[eagle4]

the main hurdle you'll run into is getting a lazer that is strong enough. not much of a hurdle, just will blow your budget a little bit. but i look forward to following your build. good luck

 

[Ground loop]

I think the laser you want is a 808nm module. https://www.youtube.com/watch?v=d8vMmAKwr6k Mind you the laser in the video is 2W. Light colored material seems to absorb 808nm instead of reflect it.

 

[OttoPilot]

I think the laser is going to be your sticking point; I would hash that out before you go any further.

You can easily make your linear bearings from angle aluminum and cheap bearings (google angle iron linear bearing to see what I'm talking about).

Although the threaded rod is a good choice, its going to take a lot of grunt for the stepper motors to get a 30"+ length and the attached structure moving...you might look into belts, or even a hi-test braided fishing line (there are some rostock 3d printers that use it).

Keep us posted, this has potential.

 

[Bricks]

I hope you can get this worked out if you can I am all in even if cost would stay under $200.

 

[jhitesma]

Those tests were with 445nM at 1.7W though. I guess we'll have to wait and see. ETA for the Laser is sometime in the next month

The problem with 445nm is that the modules are multimode:

Some important properties of laser diodes are determined by the geometry of the optical cavity. Generally, in the vertical direction, the light is contained in a very thin layer, and the structure supports only a single optical mode in the direction perpendicular to the layers. In the transverse direction, if the waveguide is wide compared to the wavelength of light, then the waveguide can support multiple transverse optical modes, and the laser is known as "multi-mode". These transversely multi-mode lasers are adequate in cases where one needs a very large amount of power, but not a small diffraction-limited beam; for example in printing, activating chemicals, or pumping other types of lasers.

Which basically means the problem with 445nm is you get a larger dot so the energy is dispersed over a larger area and you end up with lower effective working power.

At the bottom of this page you can see a sample from the same guy who apparently did succeed at cutting DTFB with 445nm by slowing it down to 125mm/min at 1watt:
https://sites.google.com/site/dtrlpf/home/laser-cnc-engraving

Here's a comparison explaining how a 200w 405nm laser can actually have more effective cutting power than a 500mw 445nm:
http://laser530.blogspot.com/search...d-max=2012-01-01T00:00:00-08:00&max-results=7

Which is why I'm leaning toward 405nm for my own experiments.

BTW - Please make sure you have good laser goggles before playing around with lasers of this power! Cheap ebay goggles aren't always effective - do some research on http://laserpointerforums.com/ about goggles and which are good and which to avoid. Seeing some of the tests people have done on there convinced me it was worth spending extra for good googles when I got my first higher power laser. ("only" a 532nm "5mw" but probably closer to 15-20 based on how it can cut through black electrical tape ... also enough to blind yourself if you're not careful!) I ended up getting a pair of "Eagle Pair" for about $60 from here: http://www.survivallaserusa.com/ With them I can barely tell if my green laser is on or not they're so effective. They cover 190-540 as well as 800-1700 so I'm tempted to go IR...but not being able to see the spot scares me so I'm sticking with the 400 range for now.

I suspect the biggest issue you'll run into will be getting smooth movement with the hardware store rod. Motion rod is much smoother and straighter - but also more expensive. I've played with some hardware store rod before but never got anything close to usable tolerances. As you may have seen in my thread: http://forum.flitetest.com/showthread.php?18033-CNC-adventures I'm taking a slightly different approach...I started small using DVD drive mechanisms for the motion...I'm limited to only a 1.5" square work area...but I was able to build it out of junk I had laying around and 2 cheap stepper drivers. So far my total out of pocket cost is <$10 It's not big enough to cut sheets of foamboard...but it will be effective for testing and seeing if I can get a small laser module to reliably and usefully cut foamboard. Then once I have a working ability to cut I can scale up the hardware which is when things tend to get really expensive.

That said I'm really interested to see what laser you end up going with. I'm still debating whether to look for a blu-ray drive to steal a module from, or go the ebay/China route and whether to go 405nm or 445nm. But I'm leaning heavily towards a ready to use 405nm...just not sure I want to spend that much on something that may or may not work.

 

[willsonman]

Gosh, what and undertaking. I've not thought about lasers since I was in high school where I researched holography. That was interesting because the lasers were much different if you wanted color holograms.

I've thought about this in another way for my personal use. Using a soldering iron as the cutting implement, you peel off the paper and then cut the parts. Problem is the parts would slide and shift as they are cut. Not to mention the fumes generated.

 

[ScottyWarpNine]

Here's a comparison explaining how a 200w 405nm laser can actually have more effective cutting power than a 500mw 445nm

Thanks for the explanation. I bought a 500mW 405nM laser from China. It comes with a controller and a dc power adapter. I'll definitely pick up some goggles. After all, if I can't see, I can't fly.

I did a lot of research online which is why I went with the 405nM, but a lot of the information out there was over my head so I didn't know exactly why it was a better route. Your explanation makes a lot of sense though. So thanks for that, I was starting to doubt my choice.

As far as linear motion, I obveously went with threaded rod because it is cheap. I'll try it out for preliminary tests just to make sure my design works, then if I need more precision I'll buy some acme rods and swap them in. The main thing I was sure of was that I didn't want belts. I built a reprap 3D printer and I'm constantly fighting with the tensioned and pulleys. And with such a large machine I figured the problems would just multiply.

 

[abieex]

Following this very closely. Under $250, Im all in!

 

[ScottyWarpNine]

I made a rough sketch of the met parts. Everything should fit onto a single 4'x8' sheet. I'm planning on using 1/2" mdf. A sheet at Home Depot is about $23, so no too bad.

Jhitesma, what issues with tolerance were you running into with the threaded rod? I plan on building in anti backlash nuts like the Z axis of my reprap, so that shouldn't be a problem. I would really like to try to make threaded rod work because acme rod would blow my budget up. A 6 foot length is over $30 from McMaster Carr as opposed to less than $6 for threaded rod at Home Depot. Much less available locally too.

This is the laser I bought:

It's going to be quite a while til it gets here and I can start testing. I ordered all of the other parts I will need (stepper motors, drivers, flex couplers, etc.) I'm up to a little over $100

 

[jhitesma]

It's not the threaded rod that's the issue, I agree that on a large machine threaded rod has advantages over belts...though with something like a laser where there's no real force against the "cutting head" belts do have their advantages.

It's the smooth rod from the hardware store that I suspect will give you issues. Hardware store smooth rod isn't really smooth or straight enough for motion rails and you'll likely run into issues with the motion platforms hanging up. The video that Ram posted has links to an ebay store with GREAT prices on linear bearings and polished hardened rod designed for motion control - not too much more expensive than what you paid for the hardware store rod:

8mm x 1m rod $10:
http://www.ebay.com/itm/8mm-x-1-met...540?pt=LH_DefaultDomain_0&hash=item4d13fc7384

Their shipping is $10 because shipping something that long is always expensive. But they do combined shipping. I put together a test cart on ebay for all the parts to create the machine in the video RAM posted and it was just under $120 shipped to my location. Lot better than when I priced this stuff out a year and a half ago and decided to use the arduino mega I had picked up to run a rep-rap as a flight controller on my first quad instead because the motion parts were too expensive for my tastes

I just tweaked the order to have 1m rods instead of the 500mm used in the build in the video - with that upgrade the total order (rods, linear bearings, steppers, belt, belt pullies - came to $135 delivered. It could be tricky to build it to get full motion over a piece of foamboard with 1m rods. That only leaves about 6" of extra length so your motion sled would have to be fairly slim which could make the other axis kind of tricky. But it should be doable and once you go past 1m rods aren't really an option anymore and you'd have to look into rails and extrusions.

There are several drawbacks to smooth rod from the hardware store:

1) It's not hardened - so it can sag which is all kinds of bad.

2) It's not polished so getting smooth motion over it even if it doesn't sag can be tricky.

3) It's usually not as straight as hardened polished rod.

4) It's harder to find linear bearings that fit it properly. Most common affordable bearings are metric sized not imperial.

A cutter big enough to do a full sheet of foamboard is really pushing the limit for a rod based design. But if you keep the build light and stiff it's probably possible and is a lot cheaper than rails and extrusions Even aluminum angle can get expensive because of the bearings/wheels to move on it. None of the individual components are expensive...but when you start adding up how many of them are needed the goes up quickly and just building a motion platform with decent accuracy that can cover 20"x30" is probably going to end up costing around $150.

Where did you order your laser from and how much was it? Because that's the bit I'm seriously shopping for right now but will probably wait and see how yours does

 

[ScottyWarpNine]

This is the laser I bought.