CNC adventures

[jhitesma]

I've wanted some kind of small CNC and/or 3D printer for a long time. They both have a lot in common mechanically so I figured first step is to get some kind of X/Y positioning table working.

I decided to go with the classic "DVD Drive CNC" as a first step since I have piles of old CD/DVD drives laying around.

First night I took apart 3 drives...none of them had stepper motors, they all used DC motors with optical encoders for positioning

Second night I took apart 3 DVD drives and 3 of them had steppers...though one of them was a rather funky setup. But with 2 steppers from DVD's I had the basic mechanics to build an X/Y table at least!

A few weeks ago when I first decided to tackle this for real I ordered up a couple of "easydriver" stepper motor drivers off ebay. They were only $2 each with free shipping so I got 5 even though I should only need 3 for this project.

I also dug a few larger steppers out of an old copier/printer collator a friend had given me and tested with them first.

Wouldn't you know it...first test I hooked power up backwards to the easy driver and let out the magic smoke. Good thing I ordered extras! Second try I was able to get the motor to respond, so I knew I was on the right track!

Last night I wired up 2 easydrivers, found a version of GRBL that was modified to work on the arduino mega (GRBL normally runs on '328 based arduinos but I don't have any UNO's and didn't feel like using one of my pro-minis since I have 3 mega's in various states of usability here from my experiments with multiwii.) Got GRBL flashed and responding...but it was 1AM and I didn't have a power harness for the motors still.

This morning I quickly whipped up a power harness, hooked a lipo to it and....it worked. Running off a 3S it worked quite well...but the steppers got quite a bit hotter than I would like. Ran it off 2S instead and the motors still seem to respond fine and now they stay nice and cool:



(note under my monitor the collection of vibration dampers I scavenged from the DVD drives. One drive actually had some just like you see on mobius/gopro mounts - making me suspect that's what they were originally designed for. The ones in the photo are a bit different of a design but are really nice soft material so I'm probably going to try and make a mobius damper out of some of them.)

And proof that it's working:

Next step is to dig up some hardware and make some kind of fixture to mount these to and I should have a fully functional X/Y table! Then tear apart a few more drives to get myself an X axis and I should be able to mount up a pen and have something that works.

It's only about a 38mm x 38mm working area. So I won't be doing anything super useful with this. But it's a good almost free first step and all the electronics will work to drive larger more powerful steppers on a larger frame when I'm ready to make something a bit more useful.

The big question is what kind of tool to put on here. I'll use a pen first since it's simplest. I'm tempted to pull the laser diode out of one of the DVD burners and make a small engraver...but my laser goggles aren't safe for that wavelength and I don't feel safe working with unprotected lasers. And new laser goggles aren't cheap.

The little DVD steppers aren't nearly strong enough to do something like lift a Dremel or much else of a spindle, and a drag knife is totally out of the question. But I could pickup a cheap hotend and use one of the larger steppers from the collator to make an extruder and turn this into a micro 3D printer...not sure how many useful things I could print in a 38mm cube sized working area....

But hey, like I said it's mainly a proof of concept before I put money into motion control hardware for something a bit bigger. Eventually I'd like it big enough to draw on full sheets of foamboard and maybe even cut foam. And someday I'd really like to add a nice spindle and start cutting my own multirotor frames.....baby steps...baby steps....

 

[RAM]

Nice experiment. I was thinking of tearing some old drives apart, nice to know they don't all use steppers. Also a good/cheap way to learn about the hardware and software bit needed to control a table.

You still need a z axis for the pen unless you don't care about lines going all over the page.

 

[jhitesma]

Yeah, apparently DVD drives tend to use steppers but most CD drives (even burners) used DC motors for positioning the laser. Which is a bummer as I have a lot more old CD drives on hand than I do DVD drives

I do plan on doing a Z axis...but I may strip down an old 3.5" floppy (they used steppers for positioning the heads) as the mechanics are smaller and lighter so it's easier for the weak steppers from the DVD drive to deal with. Or I may just rig up something with a small micro servo since I have some spares on hand for my FT mini's.

Then again if I go the laser route Z axis isn't as important since you can just turn the laser on/off. You still have to adjust the laser for focus - but that can be done manually before starting the operation. I need to see if I can find a cheap but powerful enough laser diode in the range my goggles are safe for...nothing that could cut foam - but something that could do some woodburning or paper cutting just for fun. Or just go with the diode from a DVD burner or blu-ray ( I think I have one blu-ray drive that's dead and collecting dust) and put it in a box - but I'd really prefer something I can enjoy watching while it works

 

[jhitesma]

Well, that settles that. Just double checked and blu ray uses 405nm lasers...my goggles are good for 190-540 so I am covered for a blu-ray diode ... looks like tomorrow I'll be digging through my junk pile to see if I do have a blu-ray in there. If not...may have to hit the thrift stores...and if not that then I can always pick up a diode on ebay cheap.

Much cheaper to find a high powered laser within the range of my goggles than to buy new goggles that I'm willing to trust my eyes to

Looks like this will be an engraver after all...means I don't really need to bother with a Z axis....

 

[jhitesma]

Well, my quick and dirty frame still needs a lot of work, and foamboard isn't the best for a table surface (since it's not very straight/flat) but....



For now I just zip tied a pen to it, and this was just done using universal gcode sender manually to find the extents. Looks like about a 1.5" x 1.5" work area as I expected.

The stair step was from me moving it in 5mm increments.

I tried using makercam.com to generate some gcode. I made a little 1" square with a J in the middle...but wasn't sure just how to generate the toolpaths for a pen vs. a cutter. I gave it a go anyway...but the pen never moved more than about 1/8" max...so something is off in the scaling. I can input manual gcode and it responds correctly so I'm pretty sure I've got grbl setup correctly. But makercam is taking some getting used to. I've also seen ways to use inkscape to generate gcode for engravers so I may try that instead.

I need to finish the frame though, add some more nuts on the bolts so I can fine tune the alignment, and make a better work table (I've got some acrylic I'll probably try next.)

Then figure out what I'm doing wrong generating gcode. If that all works next step is to either add a z-axis or a laser...and I'm leaning more and more towards laser

I've seen that a 200mw 405nm laser can apparently cut foamboard so that's really tempting to try. Those are the modules used in blu-ray players so if I can find a cheap blu-ray drive to scavenge I'll be set since that's also within the range of what my safety goggles cover. If not...I can take the module from one of the DVD-R drives I took apart. Those are in the 650nm range which isn't covered by my goggle though so I'd have to put it in a box and use a webcam to monitor it - or plunk down $70-$90 for another set of safety goggles.

Or I could just buy a 405nm or 445nm module on ebay for <$70. 445nm is cheaper at higher power...but because of how they're made they don't have as tight of a beam so the power is more dispersed. So a lower power 405 actually performs better than a higher power 445 because it concentrates the energy more effectively.

Even if I use the module from one of my drives I'd have to buy a holder for it so it may not be much more expensive to just buy a module. I can do some small scale tests with this setup to see if I can indeed cut foam with one as claimed...and if so...then I can start working on a larger mechanical system to eventually build something big enough I can cut foamboard with it.

Short term I'd be super happy if I could engrave on CF so I could etch my name and contact info into my multirotor frames cleaner and more accurately than I have been with a handheld mechanical engraver

 

[nic872]

That awesome, you can get some stepper motors from old scanners also you can use these two pages http://www.instructables.com/id/Build-Laser-Cutter/, http://www.instructables.com/id/Laser-cutter-start-slicing-stuff-for-under-50-dol/ , i have been wanting to make one but haven't yet. this looks like it will end up awesome keep us up to date with this please.

 

[jhitesma]

Yep, I have bigger steppers I took out of an "old" (2005 so not that old) collator and I've got a big refrigerator sized HP color laser printer (2007 vintage - works but everything comes out with a blue cast and the expendables are down to about 10% life left and replacing them costs more than the printer sells for now.) waiting to be stripped. Been keeping my eyes open for any suitable scanners to strip but haven't come across any yet.

Thanks for the links, seen those both already - read just about every laser/cnc/plotter project on instructibles

After posting that last night I got the gcodetools plugin for inkscape working and confirmed that my issue was figuring out makercam. I was able to generate gcode for a J and for a little freehand squiggle. But...my config is upside down and backwards so I get to fight with the bitwise mask in the grbl config again. Apparently it was correct originally and my attempts to "fix" the directions actually made them both backwards I was just mixed up as to which way should be positive on each axis

Was pretty sweet when that first J came out though even if it was upside down, backwards, and using a toolpath more suitable for a spindle than a laser!

 

[jhitesma]

So over lunch I got the axis directions fixed...figured out what I was doing wrong with homing...and....



Success!

Most of the irregularity in the spiral was from my quick and dirty freehand drawing in inkscape - though some of it is from the pen hanging due to the uneven workbed and poor alignment between the X and Y rails.

Bit more work on the chassis and bed and it should be working a lot more reliably.

 

[jhitesma]

So after success at lunch yesterday I hit some snags. When I tried to repeat it last night I just kept having issues. My X axis kept jamming and everything was just...off....

Finally I thought "Hmmm...maybe the 3s pack I'm powering the steppers off is getting low." Yep, that 1300mah pack was 950mah down and the voltage was sagging. No wonder!

Today I tried it again though with it charged...and still had problems Thankfully one of the drives I scrapped was an external and had it's own 5v/12v power supply which still worked. Wired that up and....it still was hanging on the x-axis Well, I had turned the stepper drivers down as low as they'd go on current so I turned the X back up again...and eventually got it moving well. But I have a feeling that the X axis stepper is in a sorry state. Some of these drives have been sitting outside for a few years so I'm not shocked.

But...the potential is here so I built a new base for it that let me get things a bit more stable and square:



It's just a 4"x6" chunk of wood with a few holes drilled so I could mount bolts through it. The bolts are attached to the wooden block but stick out and there are 2 more nuts on top that secure the Y platform while allowing me to adjust it's four corners individually so I can level it better. The whole setup is MUCH more rigid now and stays aligned a lot better!

Now I'm trying to get the hang of generating g-code. Keep having problems with the home position winding up way out of place. But I still managed to get a fairly complex drawing of a penguin (my daughter loves penguins) to plot while I recorded it:

Sorry for the poor quality video. It's a pain to get the angle setup with a mobius and the lighting back here is pretty dim at night. But there's some proof of it working!

 

[jhitesma]

Few more tweaks and I've got the X axis working more reliably...did another video with a better camera angle, but still pretty lousy lighting:

Took apart 3 floppy drives looking for a suitable mechanism to use for a Z axis...no luck, they all had stepper motors, but they way the mechanics were implemented they wouldn't be easy to adapt to a Z axis. So if I want to keep the pen I'll have to come up with a better method for that.

Decided to stop risking shorting something out and mounted the stepper drivers and arduino to the frame and then did a bit of cleanup on the wiring as well:









Detail of how I mounted the bottom mechanism so I could adjust it's position. I messed up on cutting the base and didn't have room for the holes - but that means you get a cutaway view of how I did it


The arduino is just zip tied to a piece of thin acrylic I had on hand and then mounted to the screws that hold the X mechanism in place. Took me 3 times to make an acrylic piece that would work. First one I cut too short due to a mismeasurement. Second one shattered when I drilled the holes, this one broke a bit but it's not structural and I wasn't about to cut up a 4th piece of acrylic at that point


Still need to make a better worktable....

 

[Balu]

Why does it pause temporarily while drawing? Is it sending Z axis commands but they obviously can't be processed?

 

[jhitesma]

Yep, I think there may also be a few pauses between cutting operations in the gcode...I haven't really looked too closely but it's definitely something in the gcode.

I finally figured out why my X axis was giving me so much trouble. Turns out it was my single zip tie pen mount. It was pulling the pen into the slider that runs along the screw from the stepper and jamming it up. Switched to 3 small zip ties to spread the load out and it's working much smoother at lower amperages now!