CNC?

[CapnFailBoat]

Not sure if this matters but i had planned on using autodesk 123d or fusion360 for my modelling software of choice. I'm more familliar with solidworks but that is extremely out of my budget. Does anyone know if either of the autodesk softwares has g code from model built in?

 

[Tritium]

Possibly this will answer your question:

http://www.shapeoko.com/forum/viewtopic.php?f=16&t=1489

Thurmond

 

[pressalltheknobs]

This may be more your speed. uses a dremel.

http://www.instructables.com/id/How-to-Make-a-Three-Axis-CNC-Machine-Cheaply-and-/

 

[Burly]

This may be more your speed. uses a dremel.

http://www.instructables.com/id/How-to-Make-a-Three-Axis-CNC-Machine-Cheaply-and-/

The link to the article says the above machine can be made for under $600.
The usable working envelope is 10" X-axis, 14" Y-Axis, 4" Z-Axis.

Maybe it's just me...but, that picture in no way looks like $600 well spent.

I guess the question is, do you get more satisfaction out of tinkering, salvaging parts, and not worrying about the functionality of the end product? Is building the CNC more of the goal then actually using the CNC productively?

Or, do you want something that at the end of the day, is a productive tool? If using the CNC to make things is your
goal, then by all means, buy a kit or parts for a proven design that can essentially be bolted together.

I can almost guarantee that you'll get $200 into a "one-off, homebrew" CNC project, only to find that you're not even half-way there. Then you'll have to add more bits and pieces...and suddenly you're $500-$600 into it...and have a non-functioning or pretty sketchy setup in the end. It will look more like the above picture. It may work...or...it may take so much "care and feeding" to make it work...that you'll loose interest (and be $500 or $600 in the hole).

I have been following this hobby CNC space for a number of years (check out CNCzone.com forum). Up until now, it has been either too expensive, or too hit and miss for me to even consider jumping in. I've seen many CNC projects that have burned through lot's of cash...with little to show in the end.

It's only very recently that the parts and kits have come to market that makes this more affordable...but not yet at the $200 price point. Instead of blowing $200, start saving...and bank on proven parts and designs to get something decent. Or start out with just the kit's mechanicals for about $400-$500. Then slowly acquire the control board, motors, power supply, and rotary cutting tool in the form of a Dremmel, Spindle, or compact Router.



This one above is $500 for the mechanicals.
http://openbuildspartstore.com/c-beam-machine-mechanical-bundle/
Usable working envelope is 13.5" X-axis, 11" Y-Axis, 6" Z-Axis.
I think you'd agree, it certainly looks better than that particle board contraption in the top picture.

After the mechanicals, you need:
3 Nema-23 Stepper Motors is another $60-$70 bucks total.
A 24Volt power supply.
A driver board.
A Bosh Colt or Dewalt 611 Router.
The spindle mount for the Router comes with the original mechanicals kit.

I'm not trying to burst your bubble.
I too would love to get an affordable but capable CNC machine to use in building all sorts of RC plane and copter stuff.

But, saying you want to build a $200 CNC machine a little like saying you want to build a $100 DIY FPV QuadCopter.
Even if you could find $8 motors and $5 ESC's...four sets of those would bring you to $52.
That leaves $48 to purchase a Receiver, a Flight Control Board, a Board Camera, an FPV Transmitter, 4-Props...and last but not least...an actual Quad Copter Frame to mount everything on.

 

[CapnFailBoat]

Well all of this is really just for the experience. Sure I'd like to have something atleast semi reliable at the end of it. Not really looking at any wood based designs. Was really looking for the kind of stuff made of the extruded aluminum and the like. This is really just research at this point. If it turns out that my $200 entry price is unrealistic I'll just have to save up like you suggested.

In my head the ideal solution is I buy the frame(mechanicals?) and buy the other components as I save, maybe getting the budget to just get it working and maybe upgrading them as i go along.

I thought I had all the needed parts down at this point, what's a control board and what does it do?

 

[Burly]

You can never do too much research.

I don't think you can beat these Open Builds Extruded Aluminum Systems.
Especially the recently introduced kit they now offer for $500.
All three axis's are driven by their new actuator system with Acme Screws...there are no belts.
This is much more accurate than the belt systems used in the X-Carve kit by Inventables.com.

At your leisure, go out to the Mark Carew You Tube Channel an watch all the videos(26).
https://www.youtube.com/user/OpenBuilds/videos
Start with the oldest and work your way forward.
I think you'll come away impressed by the advancements over time, and the thought and quality put into his whole design rationale.

View the 2 part series on building the OX...it shows lots of concepts.
https://www.youtube.com/watch?v=nS32b55ouQk
https://www.youtube.com/watch?v=8JufQ-xCvdg

As I said, their latest $500 Kit has the latest and greatest drive system they offer called the C-Beam Accuator.
It has this drive on all three axis's
https://www.youtube.com/watch?v=nTzX6c8_8l4

You asked about Control Boards. Control Boards run the stepper motors. Stepper motors advance forward and backwards in tiny steps. The control board sends a pulse of electricity to the motor and it advances one step...there is also a direction wire that controls the direction of movement.

In the old days the Control Board was connected to a PC's parallel port. Today the parallel port has gone the way of the Dinosaur. OpenBuilds sells an xPRO V2 CONTROLLER STEPPER DRIVER BOARD that runs off the computers USB port. Inventables.com has their own offerings. I personally would stick with a PC to USB controller board setups.

Any controller will work...some have better software support, than others. Choosing the controller and software is a totally separate issue than choosing the mechanicals.

I'd also watch all the videos on the Inventables.com website.

Look at the CNC's that were offered on Kickstarter. The more machines you see at differing pricing, the more you will be able to judge the quality of the design...and give you an idea on what you should be paying.

You can pretty much separate the mechanical part of the buying process from the electronics.
The electronics (motors, control boards, software (hopefully free)) is where you can try to get some deals.
For the mechanicals...you really get what you pay for...and it's easy to get ripped off.

 

[CapnFailBoat]

Is there a difference between a control board and a driver board? Is the GRBL shield a form of control board?

 

[ScottyWarpNine]

View attachment 51079

The link to the article says the above machine can be made for under $600.
The usable working envelope is 10" X-axis, 14" Y-Axis, 4" Z-Axis.

Maybe it's just me...but, that picture in no way looks like $600 well spent.

That article is quite dated. Not sure when it was created, but the top comment is 7 years old! A machine like that can be easily built for half that price today.

 

[ScottyWarpNine]

Is there a difference between a control board and a driver board? Is the GRBL shield a form of control board?

The Arduino that the shield is connected to would be the control board. The Shield itself is just an interface that allows you to attach the drivers to the Arduino, unless it has the drivers built into it, which could be the case depending on which shield you're talking about. A quick Google search brings up some that have integrated drivers into the shield, and others that have sockets that Pololu drivers are inserted into.

 

[CapnFailBoat]

So the components list as I know it is

Mechanical
-physical frame surrounding the work space and supporting the rest of the system
-rails
-movement mechanism; lead screws, belts, pulleys, etc.
-fasteners, bearings, and spacers

Electrical
-power supply
-stepper motors, minimum 2
-control board; arduino/grbl or other control/driver board
-limit switches (optional?)
-spindle; dremel or other motor with accompanying chuck and bits
-wires to connect everything
-method to connect all of these together and to the controlling PC

Software
-design/drafting/CAD software; autodesk 123d, Corel Draw, google sketchup, etc.
-method to create g-code from resulting designs (integrated into some of the above listed software)
-drivers for the stepper motors if not built into the control board (I am assuming drivers are some kind of software that translates g-code to physical motion, right?)
-working PC to run all of this from

As far as I know at this time this would be EVERYTHING that someone would need to put together a working CNC cutter. Am I missing anything? If this is it then I plan on doing more research into the individual components and options of each from here.

Burly, I do agree that the C-beam build from open builds is very clean and robust looking and one of the more affordable options that i've seen. The only other thing close I've found so far was the zen kit, http://www.zentoolworks.com/product_info.php?cPath=14&products_id=132, made of pvc. The c-beam set up looks like I can just get the c-beam actuators and make a frame of my size choice around them so there's a plus.

 

[ScottyWarpNine]

-drivers for the stepper motors if not built into the control board (I am assuming drivers are some kind of software that translates g-code to physical motion, right?)

In this case a driver is not software. It is a physical electronics Chip or board that turns low power pulses from the controller to coordinated signals that are strong enough to make the stepper motors move.

 

[Burly]

That article is quite dated. Not sure when it was created, but the top comment is 7 years old! A machine like that can be easily built for half that price today.

That's for sure.

Is there a difference between a control board and a driver board? Is the GRBL shield a form of control board?

There are 2 functions...the control function, and the drive function.

The controller is usually represented by a small microcontroller that talks to your PC over a USB connection.
From there the microcontroller sends normal TTL on/off electrical signals to the drivers that will each power a stepper motor.
The drivers are really single "driver chips"(one per motor) that are hooked up to higher voltage (usually 24v).
They take those 5-volt signals from the microcontroller, and switch on and off 24v currrents at 1-2 amps to the motors.

Sometimes these two functions are put onto two separate boards.
Sometimes an Arduino Board or Arduino Clone is used for control.
Then a Driver Shield Board is slapped on top using the standard Arduino header connection.

Here's an Arduino UNO:


Here's a gShield Driver for 3 Motors that mounts on top (into the standard Arduino headers).


Or everything (control and driver) can be combined on a single board.
On the following two boards you should be able to pick out the single microcrontroller chip and the 4 rectangular driver chips.

Single Board TinyG (TI DRV8818 Motor Driver Chips):


Single Board CNC xPRO 2 (TI DRV8825 Motor Driver Chips):


The software running has no idea if there is a single board or two separate boards.

I mentioned the part number of the driver chips because some will handle more amps than others.
I would buy the board with the beefier drivers in order to have the option to drive larger Nema 23 size stepper motors.

When you get to larger size CNC router tables carrying heavy routers and spindles, you obviously need heavier motors running 240v or 3Phase. Here the drivers come in separate boxes, one driver per box, with active fans and heat sinks for cooling.

But in the sizes of CNC we're looking at (up to 2'x4') we can get by with stepper motors running off of 24v DC at 2-2.5 amps max.

Concerning your master list...yes it looks like everything is there.
Concerning your link to the ZenToolWorks machine...don't do it.
It's a very old design and a fixed size...
It's a one off...

Stick with the Open Builds Frame System.
The only other one to consider is the X-Carve by Inventables.
These are the only two vendors I can think of that I'd spend money with at this point.
Maybe I'm missing others...but I can't think of any.

Open Builds.com and Inventables.com are mainstream with a history behind them.
CNC is not like 3D printing, which has lots of offerings.

I think Open Builds is best in design, flexibility, and bang for the buck.
Watch the Mark Carew YouTube videos.

Also, as far as design software...how much truly 3D will you be doing?
If you are doing whole cuts through materials, or plunging halfway through and cutting a slot, that's called 2.5D...much simpler software.

 

[CapnFailBoat]

As of now i had only really planned on 2.5d projects, no 6 axis insanity planned yet. I am familiar with "professional" grade cad systems so I'm not worried about learning how to use the design software.

I had heard a few projects talk of having the ability to "upgrade " to a 3d printer by adding an extruder. I'm assuming that would require a new control board driver set up in addition to extruder and supporting hard ware.

On a side note i already own an arduino uno rev3 and that was my reasoning on leaning towards grbl, just a money saving aspect.

At this point I'm actually more excited about building the cnc machine than the original project that made me consider getting one.

 

[Burly]

I think a combination machine would be a compromise.
A 3D printer really needs an enclosure to keep the heat constant.
If the plastic being laid down isn't at fairly constant temps...I think it will warp.
If you were 3D printing action figures it may not matter.
But, if you were printing precision pieces to fit together, it would matter.
The motors used to drive 3D printers only have to drive a relatively light weight extruder at very low speeds.
A CNC machine is driving much heaver rotary tools at much faster rates.

The two machines are so different that a combo machine I fear would not be very usable.

On the electronics front, maybe you can experiment.
Buy a shield and one motor.
Then use GRBL to drive the motor.

 

[Burly]

Here's something very cheap to get you into a shield that attached to an UNO Board.
The little boards have A4988 drivers by Allegro.
There's one that ships from New Jersey.
There's one from China that includes and UNO(clone obviously).
http://www.ebay.com/sch/i.html?_nkw...ice=c&geo_id=10232&keyword=grbl+shield&crdt=0

 

[CapnFailBoat]

I'm probably going to be kicking myself for saying this but this has all made it seem a lot easier than i was anticipating (not much cheaper though). The last question i got for you guys before i start doing some hard research of my own and get going on that library of video recommended has to do with stepper motors. As far as I've figured out nema 23 is just a stronger motor than nema 17s. Is there a physical difference in size, will a nema 17 fit on a nema 23 mounting bracket?

 

[ScottyWarpNine]

I'm probably going to be kicking myself for saying this but this has all made it seem a lot easier than i was anticipating (not much cheaper though). The last question i got for you guys before i start doing some hard research of my own and get going on that library of video recommended has to do with stepper motors. As far as I've figured out nema 23 is just a stronger motor than nema 17s. Is there a physical difference in size, will a nema 17 fit on a nema 23 mounting bracket?

They are physically different sizes. Nema 17 is a 42x42mm square, Nema 23 is 56x56mm.

 

[CapnFailBoat]

Still doing my video watching and research but it turns out that fusion 360 from autodesk comes with a built in CAM suite that can generate g-code and tool pathing simulations.

And it also has an option for 1 FREE year of use for enthusiasts, hobbists, and startups that make less than $100k annually. Academics get a 3 year free pass. As far as I know you can 'renew' the freebie when it ends. From what i've seen it's just as good as the solidworks that i learned in my courses, a bit different on the layout but all the important stuff is still there. Seems like it would be a great tool for anyone who was into CNC.

 

[Burly]

Yes, that AutoCAD package looks like the ticket.

If you're a mechanical engineer, you've got to be loving this stuff.

You should really get a kick out of watching the Open Builds system videos...taking extruded aluminum that's manufactured with special V-tracks, and adding sets of wheels and mounting plates to construct fairly robust...but inexpensive, linear motion systems.

The Nema system for electric motors is based on having a standard diameter, mounting plate, and footprint for the 4 mounting holes.
http://www.piclist.com/techref/io/stepper/nemasizes.htm

Nema 17 is lighter duty...for 3D printing and smaller CNC setups.
Nema 23 is larger and probably what you'll be using.
You can get various power(higher amps for higher torque) motors in each size.
The lower torque motors are short in length.
As the torque increases, the length of the motor increases.
The diameter of the motors within each class stays roughly the same (some enclosures are circular, some are rounded edge squares).

As a last point, there are two categories of stepper motor/driver setups...uni-polar and bi-polar.
The difference is how the coils are wound, and voltage requirements.
Unipolars only use positive voltages...Bipolars use a positive and negative voltages.

Unipolars only power half the coils at any one time.
Bipolars are better as they utilize all the coil all the time...giving more torque.

In the old days, hobbiest used Unipolars because prices of the Bipolar Stepper Drivers were more expensive than Unipolar.
In Unipolar motor drivers the current always flowed in the same direction.
But a Bipolar motor driver had to have additional descrete electronics to reverse the current.

Today the stepper motor drivers you'll be using (2-2.5 amp) don't employ decrete electronic components.
Today everything is embedded in a single chip.
The is no price difference in moving up to a Bipolar stepper driver and motor.

Bipolar Motors usually have 4 wires...those are the ones you want to get (cleaner connection).
Unipolar Motors usually have 6 wires, two wires are the center-tap wires of each winding.
You can connect a 6 wire motor to a bipolar stepper driver by ignoring the center tap pair.
Now the 6 wire motor is identical to a 4 wire motor.
Keep the center tap wires separated and put electrical tape over each exposed end.

To sum it up, a 4 wire motor can only work with a bipolar driver.
A 6 wire motor can run off a bipolar driver(by ignoring the two center tap wires)...or unipolar driver using all 6 wires.

 

[CapnFailBoat]

Hahaha, been thinking about the configurations I wanted to do and settled on a lead screw for the plunge and a belt and pinion set up for the x and y.

I was originally leaning owards the gshield/arduino set up but looking into the available steppers I'm noticing that most of the nema 23s have a 2.8 a per phase max and the gshield only supports a max of 2.5 a and would require cooling at that point. Would there be any advantage to using a nema 23 over a nema 17 in this set up? Or would the limitations of the max amperage from the gshield nullify any potential gains? My searching up to now has come up with some mixed results on this matter.

Another concern I had was the max working tension of using belts, would a nema 23 be capable of exceeding this? I had planned on eventually upgrading the spindle into something that would be better capable of working aluminum and other relatively soft metals.