Ultra Low Cost CNC

[stewi]

Moving from another thread:

I like to contribute with my plan of a 4' by 8' router.
Since I like to keep my garage mostly for car parking, I plan to take the machine apart and set it up on the concrete floor, when needed.
Also, by removing the bridge with the X and Z axis, I will be able to convert the two posts for CNC hot wire cutting.
The linear travel with a GT2 timing belts, I developed for self service kiosks, does work better (less play) and costs much less than rack and pinion.
The wheels are custom made and cast rather than turned on a lathe (I don't have a lathe, only a drill press).
All plates are either plywood or 5/8" particle (floor) boards.
Since I have already motors and driver, I plan to spend less than $400 for the rest.

Upon Ahren's concern about belt stretch and wheel deflection, I responded:

Ahren, thanks for your feedback.
I do agree that the belt stretches, which can cause vibrations or worse resonance with the steppers and router. I intend to lay down “stoppers” on the aluminum tube, which I will likely cast out of epoxy using the belt as a mold. The belt will then only stretch between the two wheels. There are many CNC machines out there using a belt drive as a primary drive and for good reasons. They also stretch somewhat in this application, but with 99.8% grip surface a belt drive is almost play free.
The wheel in the picture will serve for a plug in a silicone mold I’m making. I can cast the wheels out of a harder material than the aluminum (see thread about epoxy granite), however, I will implement a bronze bearing in the core.

I feel quite at home in the thread for open source CNC machine, with my intention to share over 30 years professional experience with servo drives, robotics and linear motion and that in the old and new country.
However, I’ll start a thread for ULC (ultra low cost) CNC machine. I’ll share drawings, BOM, data sheets and calculations.

Regards,
Stefan

[stewi]

Before I dive too deeply into the machine design, I would like to have a word or two on the motors and drivers.

I got these motors out of a trash container from a large German company, which may be a bit unfair to use in this ultra low cost application, since these are Swiss made motors and quite expensive:
Escap stepper http://www.electromate.com/db_suppor...edy_Line48.pdf

Once I have the machine running, I’ll measure the actual loads and will downgrade the motors likely to a 400 oz*inch holding torque.
For all more interested in the motor physics may read this:
Stepper info http://services.eng.uts.edu.au/cempe...s_ch10_ppt.pdf

I searched quite a bit on driver boards and was close to develop my own board until I found this one:
HY-TB4DV-M driver board. Vendor calls for professional designed board. I’m more interested in the optical isolation of the 5V signals to and from my computer
“50 Full closed-type optical isolation to protect the user's computer and equipment
60 Professional design, two-stage signal processing, super anti-jamming”

The board has no common ground and the mounting holes have nothing to do with ground either. If the board is a copy clone from China and the original board designer reads my thread, than I do apologize for my critical words.
All I can say, it worked well for me in the past and I haven’t blown up a board yet either.
You may buy this board from China or from me, tested, for a little more money.

More commonly used in the market are break down boards with parallel port or USB connection and driver boards, one for each motor. Advantage, you can hook up more powerful driver and motors, USB is certainly also a big plus, but it is more money than the single driver board, you can buy in 3 and 4 axis version.

Also, motors could operate at a higher voltage of 48 VDC with better torque, but the driver board calls for 36 VDC max (which I haven’t tested yet) and I do most of my applications with 24 VDC.

Although there is open source CNC software on the market, I settled with Mach 3 software, which can be used free with limited G-code. ArtSoft USA - Home of Mach3 and LazyCam
It is $175 for a license and I find it a well spent investment.
There are many free add-ons in this software like letter engraving in almost any font, but most importantly, I can configure the port and pins of the printer port, that it works for the driver board of my choice.

I’m using a 4 motor configuration in all my designs, operating the longer Y-axis in master slave configuration, or for CNC hot wire cutting all 4 axis operate separately in 2 y-axis and 2 z-axis set up.

Stefan

[roy_okc]

Stewi,

I'm presently wrapping up a ~18"x48" capacity build. I designed it small to fit in my workshop. However, I see the possibility of wanting to occasionally wanting to work an entire sheet of material. This would be perfect for that, able to do in my garage or even driveway and then store it away in minimal space. I'll be watching your design.

Thanks,
Roy

[stewi]

Thanks, Roy.
I do consider drilling some holes in the concrete in order to assure the gantry is plum and square each time I set it up in the same place. You may have to do this to your driveway as well. A small price to pay for machine mobility.
I made quite some changes to the drawing.
First, I closed the boxes with the Y-axis tracks. The exposed ribs could interfere with motor cables. I can likely not effort the extra price for cable chains and will have a few cables looping on the floor.
I also changed the set up for the Z-axis and have a spindle push up the weight rather than pull it up. The main reason behind is, that I have a perfect circulating ball screw, I like to clone in a mold and make a low cost epoxy granite screw out of it. The impact strength of the epoxy should be sufficient, but I'm concerned to hang my Z-axis on it.
Else, I spend quite some time in optimizing the wrap angle on the pulley wheels.
If I don't use the belt as intended with the motor on one end, an idler pulley on the other and the gantry strapped to both open ends of the pulley, than I do have a smaller wrap angle around the pulley. If I raise the motor position than I get more wrap on the pulley, but forces on the belt will also point upwards and not lateral, ergo, it is less efficient. I could also use a larger pulley wheel to get more teeth in action, but would also loose torque with it.
I can see a few higher math integrals in this problem and displace this problem for later.
I,m currently building molds for the wheels, ball screw and other parts. Pictures will follow.

[stewi]

' making slow progress by casting the wheels for the linear tracks.
I made one perfect model in order to make flexible molds for a small production of casting 3 wheels at the time.
The product I'm using here is OOMOO 30 a liquid rubber compound. After 4 hours wait time the model came out of the mold just fine, despite the deep groove for the belt.
For casting, I'll be either using Alumilite Casting Resin and Silicone Mold Making Rubber
or Smooth On Onyx Urethane Plastic Resin for Casting into Rubber Molds by Smooth-On with a hardness of 80 D shore. That is about as hard as aluminum.
Upon casting, I'll have the inner bronze bearing placed over the shaft.