STICKMAN'S LETS LEARN CNC THREAD

Stickman393

Well-known member
Alternate thread titles:

CNC for the barely conscious
I'm afraid of CAD
WATCH ME MAKE STICK FIGURES IN ALUMINUM AND ZINC GUYZ

First things first: I picked up a desktop 3018 CNC mill/router. I picked it up for 3 bills on amazon: This thing is extremely limited in it's work area: no one's going to be making any top-side input/output/power routes on one of these, it's limited to a height of something around 70MM. I may be able to swap the collar for a shorter one I picked up, but there's no way I'll be able to do the top jacks with this one unless I make some serious modifications.

Which...I already own a normal mini mill and a drill press. If I have to manually line up my holes, THATS OK. It's got a decent spindle on it, and seems to be a decent small scale all aluminum chassis.


So...I'm gonna be putting this thing through it's paces. I JUST figured out how to take vector art, turn it into Gcode, and get it properly interpreted by the machine. There are quite a few pitfalls on this path, and I'll be doing a little write up in the near future here.

That being said...I'm gonna start EASY here. 2D engraving. Use the CNC with a 60 degree Vee bit to mark my holes and drill afterwards.

Later on, I'll probably see if this thing has the oomph while using a small end mill to cut out my controls.

I'd like to use this thing to generate a marking template for the top side jacks, and I'm definitely going to be using this to make up some single sided PCBs.

MAYBE, EVENTUALLY, I'll get into 3D design. But by that point I may need to upgrade my machine...probably by adding ballscrews and steppers to my Harbor Freight rebranded Sieg X2.

But for now...a preview.

IMG_20210815_004307_50.jpg
 
You had me at STICKMAN.
It appears to have enough motor to work with aluminum well.
Does the motor adjust vertically inside the housing?
 
You had me at STICKMAN.
It appears to have enough motor to work with aluminum well.
Does the motor adjust vertically inside the housing?
Sure does! I've got it up and level with the bottom of the clamp right now.

I'm probably going to end up pulling all the screws one by one and applying red loctite too. Something tells me these are gonna wabble loose real quick.
Egad, such a CAD...


Want all the toys tools.


Darn, I was going to call my Calamity "Rude Awakening".
This is actually V2 for me...:alien:

I was using some crapppppppy J201's from evilbay on my first version. It didn't work out. I stickman smashed the board. Now I've got a collection of about 55 j201s, and I'm using my little JFET tester to select for ultra low IDSS...oh, and transistor sockets. Started doing that on all my builds now.

Lessons learned...after running my first CNC program...

IMG_20210815_123840_50.jpg

Well...close. Pretty close.

Immediately noticeable flaws: those circles aren't very circle-y. The "I" in "Grind" kinda blended into the "N", and the exclamation point in the "AHHH!" blended into the last H.

Letter spacing is important. We'll go over that later in inkscape.

Also, I was using a 60 degree carbide V engraving bit...and I believe I set the depth of cut a bit too deep. So deep, so so deep, eh, (Clipping...anybody? Love me some Daveed Diggs)

I'll get into the nitty gritty soon...but for now, behold the mildly misshapen goodness.
 
Aite...little update for today.

It's interesting how little unseen things can trip ya up. With that in mind, I present...the uneven 125b thickness.

I should have taken pictures...this little amazon box that I was using sloped down from top to bottom a hair under .25mm. Damn...

A quick pass with my mini mill and a fly cutter fixed that right up. But for future reference I'm gonna look around for a good, cheap source for flatter enclosures.
 
I guess you can’t assume everything is square and tru.... I can see how that would be easy to do. It’s just not something you initially think about... untill something like that happens. But at least you figured it out and know how to deal with it now.... good stuff bro.
 
Today's lesson: Rigidity, feeds & speeds.

I've been reading into the differences between High Speed Machining and Traditional Machining.

The basic idea: each cutter and material combination has an optimal range of "feeds" (meaning how quickly the cutter moves across the workpiece) and "speeds" (the RPM of the spindle).

Too fast, you end up generating chips faster than your cutter's flutes (EG, recesses) can clear them, which generates friction and heat. This will break your cutter bit real quick.

Too slow, and you end up nibbling away at the material, pressing *into* and burnishing the material instead of cutting into it. This will generate friction and heat, could potentially harden the material that you're cutting before you even cut it, and will lead to bit failure.

Ideally, you want to take as deep of a cut per pass as possible, to utilize as much of the cutter's edge as possible, but you MUST balance that with your feeds and speeds. A deeper cut will require slower feeds and speeds, a shallower cut will require faster feeds and speeds.

Notice how I said "ideally"? Well, that refers to traditional machining...if you have a cheap CNC3018 KIT like I do, you DO NOT have an ideal machine.

In particular, there is considerable slop on the z axis and a little in the x axis. These cheapie CNCs are NOT very Ridgid.

I will be addressing the Rigidity issue later on with some modifications to my machine...but there's only so much that I can do in that regard.

So what is High Speed Machining? Simply put: it is a technique that was born out of necessity. The big, super rigid machines that could handle the big jobs using traditional machining techniques can only do one job at a time. Smaller, less rigid machines would sit idle for long periods while machinists would wait for a job to finish.

Someone eventually said: "hey, why don't we use one of these smaller machines to do these jobs, but reduce the depth of cut and increase cutting speed?"

After everybody got done laughing at him, he went and set up a smaller machines out of spite, because fuck those people. He was able to make the smaller machine produce, at which point someone else said they had the idea first, took all the credit for the be discovery, got a pat on the head, and the shop owner made a bunch more money.

Ahem...

So, aside from satirical stories, HSM has great utility when using a small CNC3018. Small cuts, fast speeds, fast feeds.
 
IMG_20210818_160734_50.jpg

My first two passable attempts.

The rude awakening is likely gonna get a light coat of spray paint and a gentle sanding. The acrylic paint pen I was using flakes off, but I wanted to do this to add contrast.

I've been fighting a little with my machine...it has a slight preference to cut deeper at higher X values. I likely need to adjust the bed a little bit...were talking fractions of a millimeter here, but enough to be noticeable.

The newest one is for a VIIB. Why? Vibrato...trill...what else trills? Oh, cats do. ohh! I'll name it after my little buddy, KK, may he rest in peace.

Plus, best damned cat name ever.

The exclamation point is a little fucked, I'm using a new vise that's a bit taller and I forgot to reset my retraction legnth. Woof. It came out decent, though.
 
TODAY: simple CNC software, free edition:

FOR a powerful vector-based image file creation tool: INKSCAPE. The Hershey Text extension is a GREAT tool for creating true single-line fonts.

Single-line fonts, you say? Yes, most fonts translated into vectors will trace the OUTSIDE of the letters, leaving you with hollow, bubble letters that take up a lot of space and may not be the aesthetic you're after. Single-line fonts can be traced using...you guessed it...A single line. Great for engraving purposes.

FOR a simple, intuitive .SVG to G-code translator software, carbide create free is astoundingly good. You can even do the design elements in carbide create, but honestly I think it's worth spending the time learning Inkscape: it's far more flexible for the design stage.

For controlling my CNC3018...right now I'm using Candle. Free, open source, interfaces with GRBL controllers easily, transfers all data over USB.

I DO plan on exploring other software options as my desires to create more complex artwork featuring stick figures grow...I do find certain elements of Inkscape to be awfully limiting, and I have to say that if I could get my creations in affinity designer to import into Carbide Create more easily I'd be thrilled...

That's it for today...I'm hoping I can get some kind of beginning-to-end step by step guide. That may come with time...
 
Well my friends, time for an update.

My CNC has been out of commission for about a week and a half...I've been working on figuring out how to make a manual control for my z-axis and I've been coming up short as of yet.

But...I was able to make an important upgrade to my setup: squeezing every little bit of ass out of my spindle.

It's NOT convenient. Not at all. But the 360W 48V power supply I slapped on there is letting me top out my spindle speed and making use of all the motor has to offer.

Before...well...the stock power supply was shared between the spindle and the steppers. Not an ideal setup...sharing power supplies opens a whole bag o worms. Better to keep the stock PS on the controller...leaves plenty of headroom for the steppers, instead of trying to deal with a 200W spindle and 3 roughly 30 watt steppers.

BUT...it kinda sucks right now. Looks like the overload protection on my new PS is particularly fast acting, so I have to give the motor a soft start by turning the power on and using the speed potentiometer on my spindle controller to get the motor up to full speed. That...and I have to multi task. I have to press "Send" on the computer as I ramp up the speed potentiometer...yeesh. gotta fix that. Maybe I can use a maximum speed output signal from my GRBL CNC controller to trigger a 24vdc relay...and use the normally open contacts to connect power to the spindle. Still need to figure out how to give this thing a soft start though...might just need to buy a better motor controller.

Although...holy hell. A quality power supply makes an immediately clear difference. That spindle spins with authority.

Next on the list...I gotta figure out how to configure my starting Z height. I've got a probe that I plan to use... To be continued.

Here's a pic of the madness, and my latest design for a Boss Hyper Fuzz clone, PCB courtesy of AION.

IMG_20210831_173216_50.jpg IMG_20210831_175804_50.jpg
 
Last edited:
Alright, Ive been experimenting a bit recently.

First off...double sided tape is your friend. I was using a vise for a bit...but for engraving work, it's not really needed. With some good double-sided woodworking tape, my enclosures stay PUT while I'm milling. No problem.

I'm still running into rigidity issues with my setup...the z axis is just...gah...a pain. I just got some aluminum stock for making new vertical rails...hopefully an upgrade there will allow for better performance.

One thing for certain...don't go over your skis. I tried increasing my depth of cut slightly on this latest attempt, and it did NOT go well. Too bad. Now, though, I've come up with a backup plan for moments like this.

IMG_20210906_165622__01_copy_1619x2159.jpg

IMG_20210906_165640_copy_1785x2381.jpg
 
This is an interesting read. Each project gets progressively better as you get more in depth with the machine. I'm on a bit of a parallel track with my laser.
 
This is an interesting read. Each project gets progressively better as you get more in depth with the machine. I'm on a bit of a parallel track with my laser.
Schwweeeet! What's the specs on your laser? Been considering one of those...
 
This is an interesting read. Each project gets progressively better as you get more in depth with the machine. I'm on a bit of a parallel track with my laser.
Start doing writeups. I have the 3018 with the spindle and the laser and want to start getting into using both. I really want to see if I can use the laser to burn off the powdercoat.
 
  • Like
Reactions: fig
Alright! Progress!

I've been looking into ways to fix up my depth problems when engraving out-of-level surfaces.

And lo, I have found ye.

A free program called open CNC pilot will take the G-code written by carbide create and adjust the Z depth based off the actual dimensions of the work piece

It does this by probing a matrix of points. Connect two conductors to the z axis probe connector on the GRBL board, connect one to the enclosure and one to the actual bit.

It cycles through the point matrix and maps out the height of your workpiece...and then allows you to adjust your cutting depths dynamically, depending on the position of the cut.

It replaces candle as a GRBL controller too...pretty slick.

IMG_20210913_153257_copy_1709x1281.jpg
IMG_20210913_154107_copy_1878x1408_1.jpg IMG_20210913_153303_copy_2218x1664.jpg
 
Viola...the thermionic punisher.

Erm, tube driver. PCB courtesy TH custom effects. That window took like 30 minutes for my little 3018 to mill out. Woof.

Gonna attack the enclosure with a wire disc, maybe spray some bleach on it and let it soak up the sun. Then hit it with a blowtorch. Throw it off an 8 story building. And then a bomb. Fuck it, let's napalm that shit.

Little spray paint and some coarse steel wool to really make the engraving pop.

Not sure exactly what I'm gonna do with the window. Yet. I've got some 2MM thick glass slides coming in...that would likely need some padding and support in order to keep it from breaking.

Although..I have considered keeping it open-air. Would CERTAINLY help with keeping the enclosure at a reasonable temperature. Regardless, I was planning on throwing on a few threaded sintered bronze filters on the sides...having the top open would really help with inducing a convective draft through the pedal enclosure.

Although...quick math here...the pedal draws between 300-500ma at 9V input. So, it's about a 4.5 watt device, and produces about 15 BTU/hr. Not insane...but yeah, a little ventilation will go a long ways here. IMG_20210913_203812_copy_1825x1368.jpg
 
Back
Top