Hobby CNC - A documentary (in progress)

[Rainmaker]

This is just going to be a place to collect my adventures in low budget CNC.

So, I admit up front that I have veeeery little experience with CNC operations or equipment. I went to a class in Highschool at one point, where we each got to hit the "go" button on the CNC lathe and make some little aluminum souvenir. I have a little more experience with manual mills and lathes, but it's been awhile. I do have a working knowledge of how these things function, 3D printing has given me a decent base set of skills with hobby grade automated tools, components, software, etc...

Always wanted to be able to do more than just FDM plastic though, so a CNC router has been on my wish-list for quite awhile. Anyway, what pushed me closer to the edge recently was a snowball rolling downhill...
- Bought the Relic cocker pump from Lt. head-shot
- Bought a rental EMEK from djeclypse
- Modeled a "relic" style EMEK body
- Modeled some finger-groove EMEK grips
- Said to myself "Gee - these grips would look way cooler in hardwood..."
- Saw the CNC router thready from rawbutter
- Found a $100 gift card I forgot I had from some by-gone birthday
- Had an extra beer one night, and woke up the next morning with a router en route.

The Comgrow ROBO CNC, el cheapo hobby CNC router



The EMEK grips that I hope one day to hold in my hands



So, out of the box this was pretty easy to set up. it basically came in 2 seperate pieces: The table part, and the gantry part. I just had to attach the gantry (if that's the right term) to the table, and plug in all the wires. and Viola! Powered up! Buuuuuttttt, then I had a few hiccups.
- I couldn't get the spindle to run at ALL at first. Turns out there are a couple toggle switches on the main board, and they both have to be in the UP position in order for the spindle to run. It took me awhile to figure this out
- I couldn't get the manual controller to do all the thigs it's supposed to. The buttons on it are finicky, and hitting the "Back" button while in manual control sometimes makes it turn on the spindle, or change the movement increments....? Still haven't quite figured that one out.
- I kmew NOTHING about GRBL, or its interface, so that took some youtubing and manual reading
- I couldn't figure out how to Home the spindle. There are limit switches on all axes, So I figured it must be doable. Turns out you have to change a line in the GRBL settings to enable homing
- When it hits the limit switches during operation or manual control, it locks the machine, and it took me some time to figure out how to "unlock" it. Turns out in GRBL you have to hit "Reset", and then "Unlock" in that order

Anyway, I was able to get the basic functions down, and load up the sample routine for a first cut...

First Cuts made on the router. This is just a sample thingy they provide on the SD card. Turned out pretty decent


Success!

[Rainmaker]

So, the ultimate goal being to make those nifty grips, I realized I would have to be able to mill two sides of the work piece...

so how do I do that?

Initially I was thinking I would have to make some kind of jig, and make sure the work piece is cut perfectly to the exact size i want, and then maybe clamp a 90 degree bracket onto the bed to act as an indexing point... but all that seemed maybe a bit over complicated? So I went to the good old youtubes, because someone has probably done this before right?

Yes! This video was very helpful. Exactly what i needed. Even uses Fusion 360 for the CAM part.

https://youtu.be/Ce6mYKMAoX8

So, I made this simple shape to try out a 2 sided project test

From the video, i learned that I need to mill holes for my pins that will serve to index the work piece in the right place when I flip it over. And I need to make sure I am aware of the axis around which I will be flipping the part to the other side.So i added these dowel things to help me bore the holes in the right place. Flipping axis shown in red. I wanted to purposely place the object so that it was not symmetrical, since I know my grip panels wont have a line of symmetry, and I need the reassurance that I can accomplish the flip without needing a line of symmetry.



Now into the CAM part. I made 2 setups, and the 2nd one is nearly identical to the first setup except that the z-axis is flipped. Then i added 2 operations to each setup.

Setup 1:
- Bore the indexing holes for the pins
- Mill out the pocket



Setup 2:
- Bore the center hole
- Mill the outside of the object



After that, and a little more research, i figured out how to get Fusion to export the operations as gcode via "Post-processing"...
But that generated a .gcode file, not a .nc file (what GRBL wants)... so i just took a leap of faith and changed the file extension to .nc
And that worked!


So then I started out just trying to bore the holes for the pins into the spoil board.
and that worked easy enough.
But as I was doing this I realized a problem. The Z-travel on this thing is not quite enough to be able to fit the spoil board, and a workpiece of the same thickness, and the z-probe, plus have clearance height enough for the tool head to move up when the gcode tells it to...


So, I tried using a thinner board I had lying around as the spoil board.... weeeeelllllll, yeah about that...



It turns out that this board presented a couple problems, and also revealed another problem with the router.
- Thinner board means I should have re-done the milling routing to NOT bore as deep... yeah, see that bit of aluminum missing from the table just in front of the board? Ooops
- Apparently this board is NOT made of wood, but some kind of fiber reinforced polymer. Which melted. And stuck to the end mill. And over-sized the holes
- Also fun fact i discovered, the big red EMERGENCY STOP button on the front of the machine? Yeah, that doesn't work.

So now I'm coming up with another way to compensate for my lack of z-travel. More to follow.

[Spider!]

Good luck! Spindle on!

[Rainmaker]

So, maximizing my z-travel utilization! I decided to stick with the idea of using a thinner spoil board, and I found a scrap piece of wood in the garage that's about 1/4" thick. That's giving me an extra 0.35" of room to play with versus trying to use a spoil board of the same thickness as the work piece (~0.6" thick workpiece)

Also, I took apart the z-probe and realized it consists of a thinner metal disk bolted inside of a much taller spacer "puck" looking thing. The disk measured out at a consistent height all over, so I decided to use that piece for the z-probe by itself. This reduced the height of the z-probe from about 0.7" down to 0.3"

So between those two things, I gave myself an extra half inch or so of z-travel.

I also played around with different spindle mounting heights, and inserting the end mill at different depths, to kind of come up with a best case scenario.

All said and done, I've got ~1.75" of z-travel.
Then take into account the 0.25" spoil board...
and the fact that after z-probing the tool head moves up 10mm...
I've got about 1.1" of z-travel that's actually usable for milling.
Thankfully that's more than enough for what I'm currently hoping to do, Whew!

Here's a view of the current stack up: Spoil board, work piece, z-probe



Next I decided to try milling out the pocket of the test model, which actually went really well!
I know very little so far about speeds and feeds, so I'm probably doing that part all wrong. The max spindle speed is 10K rpm, everything i'm reading about milling wood says I should use a higher speed than that, but that's all I have to work with, so..... Also I'm running I think about 50 in/min for the feed. which seems to be working fine. I'm taking very small depth cuts, only 0.2mm layers at a time, so this is a slooowwww process. but after about an hour and a half, I had this!!



Woohoo! Now to flip the piece, use the locating pins to ensure it's indexed right for the milling the other side, and GO!

Aaaand stop. Haha.

Once I flipped it and hit the go button, the spindle moved all the way up and hit the upper z end stop switch.
It took me some time to realize why it did this, but eventually I understood that in the CAM program when I made the setup for milling the 2nd side, i DID flip the z-axis, but I did NOT move the origin to what is now the new top surface of the part. The origin was still located where it was in the first setup. I'll try to post a picture that illustrates...



With the origin corrected, new gcode generated, and all my fingers crossed, I started over from the beginning
- New board cut
- New pin-holes milled
- New pocket milled
- Flipped work piece and hit GO on the 2nd side milling routine...

So far so good! It's about 45 minutes in now and running smooth! I hope to update this post in an hour or so with the completed 2 sided test piece!

EDIT: Test piece done! The pocket is shifted relative to the outer hull about 0.02" in the Y direction, so I'll have to figure that bit out. But not bad for a first try!

[Spider!]

How does the final cut work? Is the piece just free at that point, or is it anchored somehow?

[Rainmaker]

I made it so that there was still 0.2mm left un cut. Then just used a hobby knife to cut it out

[Rainmaker]

and another. just going to sit back and sip coffee this time while I wait for the 30 seconds...

[Rainmaker]

last one, don't know how many of these i want.
sip sip...

[shooter311]

That is awesome, I can't wait to see what you make with this! I too struggle with the limitations of FBM printers. Luckily for me, my buddy just bought laser engraver, and a CNC. I'll let him work out all the kinks and then I'll start sending him designs lol!

Good luck, I am pumped for you!

[XEMON]

Im gonna keep an eye on this ... till i break and get one (or a K40?) ...

[Rainmaker]

Thanks guys! It's definitely a learning process so far.

Just added post #2

[XEMON]

So whats the maximum gap between the bed and the collet?

Also, is the bed removable? How is it mounted?

[Rainmaker]

Post 3 just updated. XEMON i think I answer your questions in the post.

Oh, and the bed is part of the structure of the machine. Not removable per se, since if i remove it, there will be just the front and back plates and nothing holding them together.