Crusarius Home built CNC

[Crusarius]

So I have been talking for a very long time about building a CNC for all sorts of uses. I started out with plenty of CNC knowledge from all the different jobs I have had in commercial cabinetry. However, building a CNC is very different from running and programming one.

So I started out with my first CNC. a very basic unit that involved a 3d printer and 3/4" EMT conduit. The 3d printer models and most of the hardware came straight from V1 engineering. I tried to build it to be a 5x10 unit. I quickly found out that EMT conduit is very bouncy. I played with some different sizes and setups but was never really happy with it. 

It was definitely a fun learning experience in the most basic of functionality. The control board I purchased connected to a laptop and the software ran the CNC. There were no separate drivers for the stepper motors, it used nema 17 steppers. The nema 17 were very small slow and underpowered for what I was hoping to do.

Here is a picture of my first CNC.


 

This project ended up getting shelved for several years. I finally ended up having some more time on my hands and the ability to learn quite a bit more about a simple CNC. So here is the next steps to my CNC dream.

Stay tuned.....

[Crusarius]

After the initial CNC build using the basic controller with built in stepper drivers, I decided I needed to be able to run more power to larger stepper motors to get the speed and torque desired. I also felt that having to have a computer hooked to the CNC to perform any task was a little silly. Especially since computers don't typically like to be below freezing and my shop is not heated.

So I found a nice standalone CNC controller that gives me the ability to save a program to a flash drive or the controller and run it. It eliminates the potential of lost data over a serial connection.

Picture below shows the first shipment of parts for the build. The standalone CNC controller, 4 nema 23 stepper motors, and 4 drivers. The thing with the round knob is a manual pulse generator, it is used to manually move the CNC.



 

When I began designing the CNC my plan was to use nema 34 stepper motors. Once I realized how big those monsters are, I decided to step back to using a nema 23. The steppers are 425 oz./in. motors. They are rated at 36 volts and 4.2 amps each. These are the biggest nema 23's I could find. They are still huge and weigh quite a bit. But not nearly as bad as the nema 34's.

Since my main shop is only 20x20 I needed something that was big! Yet small and easy to store. I decided to use 2x2 aluminum box for all of the rails and 608-2rs bearings. The plan was to make the entire unit easy to breakdown and store. Plus I was after low cost.

Since I have the 3d printer I also decided to take advantage of that. So I sat down with AutoCAD and started designing. I would not be surprised if I have over 200 hours of design in this thing. Anytime someone tells you designing something from scratch is easy, don't believe them

So onto the design...

[Crusarius]

The key design elements included:
Low cost
Easy breakdown and store
5x12 capability 
No PC required
The ability to machine 1/2" aluminum plate

Y AXIS SKATE VERSION 1

So, with those design elements I started playing. Here is the first version of the Y axis skates.



 

The flange on the side with the dual bolt patterns is the stepper motor mount. My plan was to use HTD belts for motion. The legs on the top were planned to mount the gantry support. The skate was made out of 3 different pieces bolted together using the bearing bolts. 

I was really happy with the design and how well it rolled along the aluminum. Then I slipped when I was playing with it and dropped it on the nice concrete floor. I quickly realized that .25" thickness was not enough to keep it from breaking when it hit the floor.

Here are the 3 pieces. By breaking it into 3 pieces I was able to make the parts in a shorter period of time without using bridging. If I was planning on making a small light duty CNC these skates would have been great. The stepper motor you see is a nema 17. It was just there for testing purposes.



 

[Ljohnsaw]

 :P

[Crusarius]

Y AXIS SKATES V2

After the drop test failure, I decided to redesign the skates to make them more robust. Instead of using the .25" thickness I changed it up for .75" thickness. Since the bearings measure .8661 diameter the .75 allowed me to place holes in the center and still have the bearing protruding both faces. This makes the design quite a bit easier and allows me to make a single part that works for left and right.



 



 

Ignore the bolt lengths, it's all I had in stock

These new skates were made from 6 pieces. They were quite a bit stiffer than the first design and passed the drop test more than a few times. That plastic is slippery.

I am pretty proud of the design to minimize material and printing time, so here is some more details on how the parts were designed.

Here is the end plate and the cross bars. if you look carefully, you can see I added a standoff for the bearings. The cross pieces lock into the end pieces and then the bearing bolts lock those together.


 

Here is the motor mount. That is a single part printed 2 times. and placed together. That one took me a while to figure out but in the end, it ended up working very well.


 

The two tabs slide over the top of each other and get inserted into the slot in the bearing end plate.


 

Once the motor mount is installed the cross bars are installed, then the bearings are bolted on.

The 4 bolt holes on the top are to connect the gantry support piece which another 2x2 aluminum tube fits into and is then bolted in place.


 

Once again I was very happy with this much more robust design. Of course, as I attached the stepper motor I realized that the stepper was a little longer that I thought it was.

I ended up having to notch the gantry mount to clear the end of the stepper motor. (mental note, do not trust the part dimensions from the sellers page).



 

I also ran out of black filament during the making of the new gantry mount.

So that is the complete Y axis skate story   Stay tuned for the next parts another time. I think its late and time for bed. 

Thanks for reading.

[Walnut Beast]

It's looking pretty sharp 👍

[maineshops]

Good looking build. Be sure to use some bellows covering for your rails. My round rails collect chips. I used 425 steppers too. Do the job very well but I'm 24 x 36 in over all. Dan

[Crusarius]

I was thinking about just adding a sweeper. I have definitely noticed a lot of buildup from the little bit I have done.

[Crusarius]

X AXIS

X axis was very similar to the Y axis except I ended up wrapping the tube on 3 sides with the skate and then the fourth side with the Z axis plate. This gave me a little more resistance to rocking since I am only using a single 2" tube for the gantry.



 

The motor mount is on the surface because I oopsed and did not make the slots wide enough for it to fit inside. I did remedy that and reprinted the parts. 

This picture is showing the front side where the Z axis will be mounted. I am glad I ran out of that red filament because I was not impressed with it at all. you can see where there was insufficient layer adhesion, and it broke where I installed the bearings to guide the lead screw. 



 

This is my box or first gen parts. I have more than enough to build another CNC here. maybe I should use the first V1 engineering version and make a little one?



 

[Crusarius]

Z AXIS

So with all of the X axis parts all reprinted and functional I moved onto the Z axis. This is how it started



 

Trying to keep a low profile machine for space reasons, my plan was to run a belt to the lead screw for the Z axis. Unfortunately, I did not design to match a standard belt length. I tried to make my own but that turned out to be quite painful being stabbed by all of the steel belts and not really worthwhile with the setup I had.

The belt was cut then I ground the outer shell off and then the splines on the mating piece. I glued the tabs together, put a staple in it and then wrapped some sewing thread around it. I may have been able to get away with just simple contact cement. I did not have any.



 

I was actually pretty surprised at how well the belt loop worked.

After abandoning the belt for the Z axis I machined an aluminum sleeve to direct connect the lead screw and stepper motor shaft. The stepper motor has a 10mm shaft and the lead screw is 8mm. Very simple sleeve with set screws.

I added a set of standoffs to the top of the existing gantry and bolted the stepper motor in place. ended up working out pretty good. 

I also ended up remaking the bearing mount plate since I did not account for the diameter of the lead screw nut and it was interfering. Once again trying to keep things as small as possible came back to bite me in the butte.



 



 



 

The router mount was just temporary since I was so anxious to see if it would cut. So excuse the hackjob

After getting all 3 axis mechanically solid I started realizing how much wiring was going to be a nightmare.

To be continued.....

[21incher]

Looks nice but why 3d printed.  There's a lot of vibration  to absorb and forces that I doubt printed parts will stand up to over time. I thought  you have a small milling machine. I find  it takes less time and is cheaper to use aluminum.

[Crusarius]

THE WIRING NIGHTMARE

So when I started designing this thing, I was not smart enough to start with the finished product and go backwards. If I had, I would have known that I need a pretty big control box to fit 3 power supplies, 4 stepper drivers, and a controller. I needed 21 wires just for the power supplies. I also never thought of home switches or limit switches during the design (these will be talked about later). and id not account for powering whatever spindle ends up on here.

So if any of you decide to build a CNC in the future I highly recommend you start at the finished product and work backwards when you are designing it.

I am not sure if I should be showing these next images since they are full on sketchtacular. But hey, if my mistakes can help someone else why not. Have I mentioned how much fun it was to build this thing?

Already on with sketchtacular 

I tried so hard as I was going along to keep the wires as clean and neat as possible. I do have to say I did pretty good compared to what I usually do.

Here is a picture just putting everything together testing it to make sure it worked.


 

One of those power supplies is powering the controller with 24v. the other is a 36v powering the stepper motor and driver. You can already see the mess of wires and I only have 1 axis connected.

At this point since I had a functioning axis I had the urge to play. I have heard to many people bashing on belt driven CNC's saying they are not accurate or repeatable (kinda like the cantilever head bashing) So I decided to do a quick test. The video only shows a few times but I ran it over 30 times and had very good results.

This was the setup I used.


 

(140) Checking repeatability of the new CNC. - YouTube


OOPS missed kids lunch, brb.....

[Crusarius]

ok, back again...

Most CNC's are designed with a separate control box that is mounted to the table. This works fine for smaller tables. 

The design of this entire project makes it so I can make the cutting area as big or as small as possible. With that in mind I had to decide how to mount the electronics in a way I did not have to worry about all the extra wires I would need for stationary mounted box

I decided I wanted the control box mounted on the gantry. This meant I would only need to have extra wire going to the X/Z axis.

So here is the first version of the control box.


 

Everything is mounted to the MDF and ready for wires.

And here it is with all the wires.


 

YUCK!!!

Also remember that wires take up alot of space. the 10mmx10mm drag chain was not big enough.


 

So in the process of cleaning up the wiring I picked up a bunch of these spring connectors. Hopefully they will work well to clean it up. But right now I have a functioning CNC that I am having some fun with. 

First cuts
(140) First test cut on homemade CNC - YouTube

During my first cuts I noticed I had some rocking in the 3d printed skate. So I decided I wanted to make something a little more robust and have some adjustability. This next video shows me cutting a test piece to see if my idea was going to make it work any better.
(140) First parts I cut on the new CNC. These parts are to modify the new CNC. - YouTube

Since I was making sawdust and having fun I couldn't just leave it at that I decided to put a 2.5" cutter into the router and see how it did surfacing.
(140) Flattening slab - YouTube

The results would have been much better if it was fastened down better. But I was still very happy with how it worked.

[Crusarius]

Looks nice but why 3d printed.  There's a lot of vibration  to absorb and forces that I doubt printed parts will stand up to over time. I thought  you have a small milling machine. I find  it takes less time and is cheaper to use aluminum.

I wanted to build a test unit to see how it would work. The 3d printed parts worked well to prove my concept and figure out what I liked and didn't like. I wish I had done that with my sawmill

I actually do have a 10x42 mill and am currently in the process or replacing the plastic with aluminum. The updates will continue as I get more done on it.

Since I finally got the setting dialed in on my printer, I feel they would last pretty well. But it also depends on how much I try to push the machine.

[Crusarius]

If anyone is interested here is the wiring diagram



 

It does not show all of the individual wires but gives you an idea what is involved.

[Hilltop366]

 

[petefrom bearswamp]

Fun read for me, but being old and not savvy in CNC nomenclature MEGO

[Crusarius]

What do I need to fix pete? I like to keep things simple so anyone can understand them.

[rusticretreater]

Your diagram shows power supplies 1,3 & 4. I guess #2 power is not good enough for you. ;D

[Crusarius]

I was wondering who was going to notice that. I saw it right after I posted it  #2 is reserved in case I need to separate the controller power supply. The controller needs 2 - 24 volt inputs. right now they are both being run off of one power supply. So far I do not think I will need another one.

[Crusarius]

Well, I finished machining some upgraded parts today and decided I wanted to play with the CNC before it gets torn apart again.



 



 

So the first cut I made of the sign I did not realize how dull the router bit was. I filled the shop with smoke. It did not cut very good at all. So I just left the sign in place, changed the router bit, and then restarted the program. What a difference a sharper router bit made.

I am also thinking it may be time to actually build a real table for this thing. I had some jumping when it was cutting and I think its because I just have it clamped to a very wobbly table.

All in all, I am super happy with the results! The sign took roughly 15 minutes to cut using a 1/4" 2 straight flute router bit. I probably could have taken a deeper cut with the sharp bit but I am still learning and being cautious.

(141) More testing with my home made CNC - YouTube

[Walnut Beast]

That is pretty slick 💪. Nice video 

[Crusarius]

Thanks walnut beast.

After watching the video several times and seeing the entire machine perform a real world task I am working on some upgrades to make it a little more robust. Stay tuned...

[Crusarius]

Ever get an idea in your head and you just have to go with it whether it is a good idea or not? Well, I think that is why I made version 1 of this CNC all 3d printed. After playing with the CNC for a bit with the 3d printed parts, I started wondering about longevity. I noticed my joints seemed to loosen up after some use and I was seeing more potential for movement. So on to the next version....

One of these days I may be happy with the CNC I build. So here we go, Version 2 or is it version 356?

So 21incher asked me why I used 3d printed parts when I have a machine shop and can do much better. The reason was I had an idea in my head and just had to go with it. It felt right at the time, but unfortunately there were just to many variables that could add to to much play and less accuracy. With that said I thought alot about what he said and here it is.

I happened to have a 30x48 piece of 3/4" aluminum plate laying around. First step I needed to do with it was to cut a 4" strip off the end of it. Let me tell you, this was not an easy task. I have a metal cutting compound sliding miter saw that will do 12" cut. so that gets me 24". Well that extra 6" in the middle needed a sawzall. I used to love my sawzall, now it is the last tool I ever go for if I can avoid it.

So the blade on my chopsaw was pretty hosed. it burned more than cut. I finally broke down and tried out my harbor freight circular saw blade sharpener. It actually worked really good. Made cutting the plate much easier.

I finally had all of my rough billets made. After getting the rough sizes I put them all in the mill and made them all the same size and true. Working with square clean materials is so nice!!



 

Once the billets were made I actually took some time to make up a detailed set of drawings instead of shooting from the hip like I typically do. This definitely made things alot easier for consistent machining.




 

Most of the machining completed.



 

First concept... 1x1x.120 wall tube. Readily available and easy to work with. 



 

Unfortunately, getting it square to the rails was an entirely different experience. It was definitely solid and more than strong enough but was a royal pain trying to keep it flat, true, and square. Since I still had that 3/4" aluminum plate, I decided to use that instead. Can't get much flatter



 

In the process of upgrading the entire frame I also decided I wanted to run dual beam for the gantry. I felt the single 2" allowed to much opportunity to twist. Plus looking at different spindle ideas I realized most of the ones I was interested in were 13" long. I really did not want to have a 23 pound spindle dangling in space like that.

I used parts of the 1x1 tube to connect the motor mount and the triangle together. This gave me some adjustability to square things up when I get a real table under it.



 



 



 

[Crusarius]

You may remember those wood blocks I cut as a test using the version 1 CNC. Those were test pieces to test the accuracy of the 3d printed machine and also to prove a concept for the upgraded machine.



 



 

Both were successful. Here is a picture of the back side of the gantry with the machined parts.



 

I finally broke down and ordered a spindle for this machine instead of the router I was using. My router is an old porter cable single speed. It is a great router and works great but for this application I really wanted variable speed. I am designing this to work with cutters as small as 1/8" all the way up to about 3".

This spindle is roughly 13" long and 3" diameter, it also weighs about 23 pounds.



 

It is sitting on the plate that will connect the bearing plates. Hopefully that Nema 23 stepper motor will have enough power to lift that behemoth. When I was mocking everything up I noticed that the cast aluminum spindle mount is just a rough casting. there is not a square edge on it. I will need to set that up in the mill and machine some alignment surfaces on it so I can align it using a square instead of using hopes and dreams which have worked for me in the past.

Wow, this building a precision machine stuff is time consuming!

This is the entire Z axis setup and almost ready for testing.



 



 

Thats as far as I have gotten so far. stay tuned for more adventures of making something from nothing with your host..... ME!

Thanks for tuning in to the disorganized chaos from inside my head