CNC | Maker Portfolio

MakerCNC

Problem: Commercial-grade CNC routers and mills are
expensive and unavailable to makers and robotics teams.

Goal: Design a CNC based on commercially available
components able to process woods, plastics, and nonferrous
metals with superior capabilities to similarly-priced
hobbyist CNC routers.
Solution: MakerCNC is an adaptable CNC router based on
3D printed parts and off-the-shelf components. The design
allows robotics teams to fabricate their own machines
capable of routing wood, acrylic, aluminum parts in-house.
Additionally, MakerCNC is an attractive alternative to
products like the X-Carve, allowing further improvements.

Design

MakerCNC was originally created to utilize accessible
3D printed parts as a part of its design.
3D printed sideplates, motor mounts, Z-gantry
MGN12 linear rails for each axis
1 ¼ hp Makita router
NEMA 17 stepper motors + 8mm Lead Screws
20x60 aluminum extrusion for structure
600mm x 560 mm work area
680mm x 770mm overall size
Moving gantry design for space efficiency

Materials

The design was assembled using commercially available hardware from Amazon and Misumi, due to accessibility and cost. Skateboard bearings were used for lead screws and common 3D-printer parts such as the 12mm linear rails and 8mm lead screws were readily available. Total costs for the base-model of the project were kept under $500 dollars. Improvements increased the total to $900, highly competitive with other commercially available hobbyist CNCs such as the X-Carve and Shapeoko at the time which cost upwards of $2000.

Electronics

Given the aim of the project, a low-budget controlsystem was implemented utilizing open-sourcedhardware, firmware, and software. An ArduinoUNO with a CNC shield was used for controllingthe motor inputs, I used GRBL firmware tocommunicate with the computer, and CNCJS, anopen-sourced G-Code compatible software tocontrol the machine from my computer.

Successes and Challenges

The first iteration of the CNC was successfully able to process MDF, plywood and other wood derivatives. It also fared well against plastics like Delrin or acrylic (to some extent). The machine was capable of a depth of cut of around 1mm @ 500mm/s feed rate. The level of precision (approx. ±0.05mm) for its use-cases was found to be was satisfactory for cutting parts for robotics and personal projects.

However, I found issues when it came to processing non-ferrous metals like aluminum. Partly due to the budget end mills I was using as well as the lack of rigidity in the machine, routing parts from 6061 aluminum was found to be difficult and tedious with a maximum depth of cut of 0.2mm @ 1000mm/s to prevent the dulling of the blade. This was due to a variety of factors:

Router speeds (20-30,000 rpm) not optimal for processing aluminum
Lack of rigidity from 3D printed parts
Bad cooling / chip-evacuation
Insufficient work-holding
Low-power NEMA 17 stepper motors
Backlash on the trapezoidal lead screws (as compared to ball-screws, etc.)

Changes & Iterations

To address the limitations of the initial design, I kept with the theme of accessibility, partly due to my lack of fabrication equipment, utilizing the 3D prints and the existing CNC to iterate on the initial design.

Add paragraph text. Click “Edit Text” to update the font, size and more. To change and reuse text themes, go to Site Styles.

I designed a 3D printed vacuum fan with a custom impeller to help with chip and dust evacuation (one of the safety concerns) as well as a custom dust-shoe to prevent wayward aluminum or plywood chips.

As the machine evolved, I wanted to create an avenue for further development of the machine, and at that point, my main limiter had been the power output of the router. Though very powerful for its price-point, it was not capable of the same level of precision, lower RPM’s and power output as a traditional spindle.

Thus, I decided to upgrade the machine using a 1.5kW air-cooled spindle as well as more powerful NEMA 23 stepper motors and their accompanying electronics to power an additional ball-screw.

Spindle:

0-24,000 RPM (spindle) vs 10,000 - 30,000 (router)
1.5kW vs 920 WVariable
ER11 collets vs single 1/4" collet
<0.005mm runout vs approx. 0.025mm runout

NEMA 23 Stepper Motors:

425 oz-in output

Ball Screws:

approx. 0.05mm backlash
No need for anti-backlash nut (increased friction/wear)

Final Design

1. Rigidity:
The final design for Maker CNC utilized a more rigid body, relying on ½” 6061 aluminum plates as a base for the sideplates, Z-gantry body, mounting brackets, etc. This allowed the router to achieve higher feed rates, reducing chatter, and increasing its ability to process aluminum, one of the main design considerations and goals going into the project.

2. Power:
An 1.5kW air-cooled spindle was implemented along with its accompanying VFD allowing increased (and real-time) control over speeds, feeds, and power. As a CNC-oriented device, it allows the machine to achieve lower RPMs with higher torque, optimal for non-ferrous metals as well as more commonly processed wood and plastics.

3. Value:
The initial design was able to stay well within the $500 dollar budget, and the increased capabilities did not increase the costs beyond those of the commercially available alternatives, totaling around $900, providing a more powerful and effective platform for fabrication than comparable hobbyist-oriented machines.

Overall the design was found to be successful at its initial goals, capable of an almost 1mm depth of cut in aluminum @ 500mm/s feed rate, as well as very fast processing of woods, plastics, and other commonly used materials in robotics. The design was found to be cost-effective and highly customizable.

MakerCNC provides a powerful alternative for robotics teams and hobbyists to be able to process their own wood and metal parts, giving them greater autonomy over their projects. For teams such as my own robotics team, it enables rapid prototyping at a lower cost and opens up stronger and more accurate means of fabrication.

Improvement and Moving Forward

1. Re-Design Initial Idea:
Though the initial design was fairly successful in achieving low-cost fabrication of materials for my team and I, the costs and structure could have been more optimally chosen to achieve increased processing capabilities at a lower cost.

A fixed gantry design would increase rigidity for the same price, though it would be less space- efficient.
Bulkier 3D printed parts to achieve increased rigidity → decreased chatter / wobble
Utilize inherent strength of 40mm aluminum extrusion for y-gantry, etc.

2. Improve Upon Current Iteration:
The current iteration of MakerCNC still has room to grow. Though ferrous metals such as steel are most likely unavailable without major modifications, changes to the hardware, such as converting all axes to ball screws, improving cooling through mist-coolant systems like the “fog-buster” and improving safety through the addition of a fully-enclosed CNC router.