MPCNC-inspired CoreXY laser engraver

[dkj4linux]

In part, trying to justify purchase of my new Prusa I3 MK2S printer and also to satisfy my rather new-found interest/curiosity with “corexy” technology… here’s a MPCNC-inspired CoreXY-variant laser engraver/cutter I’ve recently built for fun.

[youtube]https://www.youtube.com/watch?v=_l-yYXgZIY4[/youtube]

[attachment file=36320]

[youtube]https://www.youtube.com/watch?v=zOddbtYh3_Q[/youtube]

[attachment file=36321]

This is in the spirit of my CamSlider (https://www.thingiverse.com/thing:1984162), detailed over in my FT needle cutter thread, where I’ve repurposed/reused/modified MPCNC (and other) parts, and combined them with a few of my own, to cobble together a brand new machine. Similarly, I had recently completely rebuilt/redesigned/converted a small, and very flimsy, Makeblock pen-plotter to use CoreXY technology (also detailed in my other thread)… and was very pleased with the simplicity and operation of the little laser engraver that resulted.

So, when a friend no longer had room for the early-version 2’x2’ MPCNC I had previously built for him – and I really didn’t just want to do yet another MPCNC rebuild/update – I wondered if I could marry the MPCNC with the CoreXY technology that had so fascinated me with the little Makeblock machine. I tore the old MPCNC all the way down to the feet/base, reclaimed all the hardware and the L-cheapo 2.1-watt laser my buddy had purchased/installed, and then started looking for “building block” parts I might use to build the new machine I envisioned. In addition to Ryan Zellars’ MPCNC, I’ve borrowed ideas/parts from his LowRider and also a novel, snap-together printed 3d-printer, Snappy (https://www.thingiverse.com/thing:1828359).

I realize this machine is too large/heavy/clunky for corexy and a general-purpose CNC – I’m just a seat-of-the-pants mechanical “engineer”, after all – but as a dedicated laser-toting machine, with no real tool-forces to deal with, this machine seems to operate quite well. It’s been a fun project… and I’m quite pleased with the result.

Early Z-axis with motor/pinion mounted above X-carriage…

[attachment file=36322]

X-carriage before Z-axis attachment

[attachment file=36323]

Z-axis rail assembly bolted to X-carrriage with motor/pinion now positioned below…

[attachment file=36324]

Z-axis is just 3 printed parts… rail assembly, sled, and pinion.

[attachment file=36325]

Laser module now affixed directly to sled…

[attachment file=36326]

Exercising Z over 100mm range…

[youtube]https://www.youtube.com/watch?v=4THOquGhUDI[/youtube]

– David

[jeffeb3]

Wow, awesome! I bet you learned a lot. The corexy printers are pretty clever, to say the least.

I like that the Z is just printed parts and a stepper motor. Nice to be so self-sufficient (or printer-sufficient?).

Dumb question, do you need the Z axis? Couldn’t you just focus the laser with a screw or something, and let it stay there? Is the Z to do better at cutting?

I like seeing the laser go through the smoke, and go from large to small area.

What’s the significance of 152mm? According to wikipedia it is, “152 (one hundred [and] fifty-two) is the natural number following 151 and preceding 153.” Haha.

[dkj4linux]

Jeff,

Besides occupying that all-important space between 151 and 153, 152 is number of whole millimeter “ticks” on a six-inch ruler… which were the first I lasered. I don’t use imperial measure with these machines or the software, so I did away with the imperial scale and just print them metric. The near-6" length is basically an artifact… and the little metric wooden rulers litter my workspace and are handy little buggers to have around. I guess they also validate the accuracy of the machine that produces them?

I used it without a Z-axis for a while but it’s just not as much fun. I missed not being able to run Ryan’s laser focus script and I’d become quite fond of being able to use the Z-axis to help determine engraving/cutting feeds/speeds/focus when testing different materials. We went through a lot of this over in my FliteTest thread (typical – http://forum.flitetest.com/showthread.php?24251-Cutting-foam-sheets-with-a-needle!&p=319439&viewfull=1#post319439) and eventually found BEST FOCUS to be crucial to cutting through 5mm ply (typical – http://forum.flitetest.com/showthread.php?24251-Cutting-foam-sheets-with-a-needle!&p=323325&viewfull=1#post323325). Much more fun!

I really like the printed stuff, too… it’s so light-weight and little assembly. And now that I have the new Prusa I3 MK2S printer, my prints are much better quality and I’m encouraged to try more “precision fit” stuff; i.e. gears, slides, etc.

This is the little Makeblock corexy-conversion I did that piqued my interest in corexy… note no Z, just manual adjustment

[youtube]Lightweight CoreXY laser engraver -- circle test - YouTube

– David

[jeffeb3]

@dkj4linux

Besides occupying that all-important space between 151 and 153, 152 is number of whole millimeter “ticks” on a six-inch ruler…

Well, that ought to be on the wikipedia page, for sure.

[billsey (Bill)]

Time to edit the Wikipedia page… I’ve been known to do that in the past.

There… Edited.

Wouldn’t the Z axis be needed if you were going to engrave on different thickness materials? If the material is thick you move up to stay in focus.

[jeffeb3]

I haven’t used a laser. I don’t know how hard it is to focus, but I would assume I would need to focus it for each thing I was lasering. I was thinking a thumb screw would work to get it focused, and then a lock screw to keep it still, instead of using another motor/driver/dealing with CAM. Seems like it would work for engraving, but not for cutting. Since the Core XY only uses 2 motors, it would be yet another motor off the top. Seems like an OK idea to me.

[vicious1 (Ryan)]

For me it is pretty hard to get it super focused, that is how I ended up with that script. Then I realized for engraving it doesn’t really matter so much but like David says for cutting, it really changes the strength. If I remeber right the power of the dot is exponential so a little bigger radius means a lot less power (although it is really a tiny rectangle). Hard by hand but super easy with the script.

David I got my parts in!!! We’ll see what come out if I can free up some time soon.

[jhitesma (Jason)]

With my laser I do adjust the Z for focus - but I don’t usually adjust it during a process. Basically I adjust the focus collar on my laser so it’s focused at a certain distance - then have a cheater stick I use to manually adjust the Z so the laser is always that height above the surface.

I’ve experimented with adjusting the Z on multiple passes for thick materials, but haven’t really seen any benefit to doing so, in fact in most tests I’ve seen worse results from doing that for some reason.

So a powered Z isn’t really needed, but an adjustable Z is very helpful.

I did use Ryan’s focus script to get the initial focus dialed in…and for a bit did use it to try and fine tune for each operation. But found that for my uses that wasn’t really necessary and just using the height gauge method was accurate enough.

[dkj4linux]

Here’s what I do with the little Makeblock machine… and it works fine. Two or three millimeters either side of absolute best focus is generally fine for engraving…

[attachment file=36402]

[attachment file=36401]

Beyond that, however, I still have visions (hallucinations?) of hanging something other than a laser on this thing… a pen, a drag-knife, a hot-wire bow, or ???. And beyond that, I think herringbone gears (or r&p) are fun to play with and it gives me a chance to use my new Prusa printer. Beyond that, it’s just plain sexy – and at my age…

=8^)

[dkj4linux]

About focus… it’s crucial for clean cuts in thicker materials. I have successfully cut 4.75mm luan plywood with my laser on the MPCNC (where I have a fully-implemented Z, of course)… 5 passes, full-power, 100 mm/min, constant focus on top of material. I don’t remember whether I had the air-assist active at the time or not…

[attachment file=36407]

[attachment file=36412]

[attachment file=36408]

[attachment file=36409]

[attachment file=36410]

[attachment file=36411]

GremlinRC (Dave from Dublin), in my FT thread, later substantiated the importance of focus as well… cutting in a single pass material that had been taking multiple passes. We also investigated “air-assist” and found that it can be a great help as well when cutting thick materials.

I’m still not entirely clear as to why keeping focus at the top of the material for all passes seems to work best for me. Many feel that you need to lower the Z-axis – and focus – with each pass, to cut thicker material. I tried it but was never able to get it to work. This diagram shows the idea… though the depth should probably increment with each pass rather than “reciprocate”:

[attachment file=36413]

Any ideas?

– David

[dkj4linux]

About materials…

Here’s what I do to set up for cutting any material… here, corrugated cardboard.

First, I run the Ryan’s focus script and then lower Z back down to best focus on the top of the material… here about 16mm… and leave it there.

[attachment file=36424]

Then I run a series of 10 lines… full-power, starting at 1000 mm/min feedrate, down to 100 mm/min.

[attachment file=36425]

Flip the test piece and find where the earliest, cleanest through-cut occurs… here, 200 mm/min might work but 100 mm/min will insure through-cuts…

[attachment file=36426]

Then cut the parts. These all look nice and fall out on their own. There’s a slight crown in this piece of cardboard that I usually can just ignore (focus is reasonably good over range of 6mm or so) if it’s not too severe or abrupt…

[attachment file=36427]

[attachment file=36428]

[attachment file=36429]

[attachment file=36430]

– David

[vicious1 (Ryan)]

That looks cooler in cardboard than in wood!!! To the Gallery it goes!

[jeffeb3]

@dkj4linux

Any ideas?

If you think about power, then if it’s perfectly focused on the top, then all the power is going down into the hole. If it’s focused 15mm lower, then a lot of the power is still on the surface. The amount that ends up in the hole is probably pretty close to the amount if it was focused 15mm off.

It seems like the cut would be slightly larger as it went further down, and then it would end up spreading out again, and get out of focus again, but focusing lower without widening the hole wouldn’t get any more power down in the hole.

I wonder if you could make cuts wider at the top, before focusing lower. If you could consistently cut 4mm thick with it focused on the top, I wonder if you could cut to 8mm or more if you made the first 4mm double wide, and then cut in the middle of those two paths after 4mm deep. I don’t know a good way to do that in CAM, but it would be easy to make some manual gcode to test it. If it worked, we could try to find a way. Actually, I wonder if the tricoidal milling would do that.

[dkj4linux]

I think you are right, Jeff. Good call.

While best focus certainly concentrates the power at the surface, makes the narrowest kerf, and maximizes penetration, it leaves an “aperture” too narrow for a now-“defocused” beam, on successive passes, to pass through. Intentionally widening the kerf at the surface, as you suggest, makes sense then… purposefully creating sufficient “aperture” for the beam’s energy to pass through on successive passes. Not a problem with a collimated beam certainly… but definitely a problem with the focused beam we have here.

Thanks! I think that explains what I’ve been seeing quite well.

– David

[jhitesma (Jason)]

That makes a lot of sense.

What I tried with “descending focus” was doing more or less what David does to find optimal cutting…except I wasn’t trying to get a full cut - I was looking for the cleanest cut and sharpest line. Then I’d run my test cuts off the edge of the material and measured how deep it cut.

I then tried doing multiple passes dropping the Z in various amounts - full depth of initial cut, half depth of initial cut, 1/4 depth…but none of them seemed to help and just leaving the Z constant seemed to result in equally deep additional passes as additional passes with the Z dropped.

The beam width description certainly makes sense as to why that would be since I was looking for the cleanest (hence smallest) cut at the top of the surface.

I really need to pickup a three element lens…I went with the G2 on mine (I’m using the “standard $100 2.8w laser”) since I wanted to cut balsa and figured the more efficient lens would help. But it sure seems everyone I know running the three element lens gets noticeable better cutting performance at the same power levels. So it seems the higher efficiency of the G2 isn’t enough to make up for the larger spot size. Kind of like how a good antenna is better than more power in radio - tight focus is better than more power with a laser Though until I remember to pickup a three element for myself to test with and compare the same laser/driver/power setup with both lenses this is just speculation - even if it makes sense

[dkj4linux]

@bill,

You were serious about editing the Wikipedia page for “152”!!!

How cool is that???

Thanks,
David

[jeffeb3]

[citation needed]

[billsey (Bill)]

@jeffeb3

[citation needed]

Sheesh, citation added.

[vicious1 (Ryan)]

Hilarious!

[dkj4linux]

An update on my MPCNC-inspired CoreXY laser engraver… still a work in progress but moving once again!

Please recognize that this machine is too large/heavy/clunky for best corexy operation and is NOT a general-purpose CNC machine or MPCNC replacement – but as a dedicated laser-toting machine, with no real tool-forces to deal with, this machine seems to operate quite well. It’s been a fun project… and I’m quite pleased with the result. Development began about the same time, or slightly before, as the ZenXY project and we’ve all learned a lot about CoreXY since – so please don’t be too hard on me… =8^)

This project has sat idle for several months and I’ve had some difficulty locating all the original STL files – some were done in Tinkercad originally and the rest were done in Onshape – but I’ve rounded up most of the parts needed to build the MPCNC-inspired CoreXY laser engraver and uploaded/published it to Thingiverse

https://www.thingiverse.com/thing:2375050

I had to regenerate several STL’s (I haven’t sliced/printed these exact model files) but I think these should be usable. The videos are showing the original configuration but operation hasn’t changed. Several parts are stock MPCNC/LowRider parts or mods thereof… the NEMA17 corner motor mounts are stock LowRider parts IIRC and the bottom-most three corner parts are stock MPCNC parts (Thanks, Ryan!). The Marlin firmware (and the stock MPCNC and LowRider parts) are from this site and only a few lines in the Configuration.h file need modification to get CoreXY operation and proper steps per mm on each axis.

The Z-axis linear slide assembly shown mounted on the tool-plate is found at https://www.thingiverse.com/thing:2813896. Thanks to John Mulac (3DPRINTINGWORLD) for a really nifty mechanism!

– David