Compact CNC Lathe

Jan. 2026

It's making chips! I had just enough time to get the machine running for the first time and turn a handful of test parts before driving back to Chico for the spring semester. I finished the spindle assembly, dialed in all the axes, and loaded up some test material. I didn't have time to wire in the encoders for threading, so unfortunately, no threading demo as of yet.

All the toolpaths for the pawns were programmed in Fusion 360. I picked relatively conservative feeds given that the lathe is currently rolling around on some Ikea wheels, and not bolted to a concrete slab. This is why there is some shaking in the video: the whole machine is rocking back and forth!

My plan is to continue developing the machine, and hopefully bring it to a polished state. Full enclosure, (mist?) coolant, tool changer, the works.

Planning

In 2023, I began selling mechanical pencil components, each one meticulously manually machined. I quickly began to struggle to keep up with demand. As I made my first hundred sales, it became increasingly evident that automation would enable me to increase productivity and, consequently, volume. After researching my options thoroughly and receiving quotes from several job shops, I decided that the best course of action would be to construct a teeny purpose-built lathe. This of course, is a ludicrous decision that only a 17 year old high school student would make.

Planning

My design had to work around several key constraints. The new lathe would have to:

• Operate on a standard 120v outlet, and draw no more than 2.4kW.
• Fit through a standard 30" door.
• Easily fit in a small SUV.
• Safely contain coolant, chips and (some) noise.

In addition to the stringent constraints, the lathe would also require sufficient tooling to turn my pencil components. Having manually machined over three hundred components, I understood that the new lathe must possess:

• A threading tool
• A turning and grooving tool
• 2 twist drills

Based on my tooling requirements, I opted to design a machine equipped with a gang tool system, where multiple turning tools are arranged side by side on a T-slot plate. Consequently, the X axis, or cross slide of the lathe, would require a considerable travel distance to provide adequate clearance for each tool.

Initial Design

The availability and price of components significantly influenced the design of the new lathe. I prioritized securing the most expensive and challenging-to-obtain components first. By “locking in” key components, I would save time and money, but sacrifice design flexibility.

Most notable among the early acquisitions were:

• A heavily modified mini lathe spindle with a pair of appropriate angular contact bearings.
• An 80mm wide ballscrew actuator, with two large bearing blocks, and a ⌀20mm ballscrew.
• A 0.7kW servo motor to drive the spindle.
• A three-inch square aluminum extrusion to serve as the machine base.

Working around these components, I designed a frame and motion system. For surfaces that I knew required relatively flat surfaces, I opted for an aluminum tooling plate, which is inexpensive and readily available. I also employed laser cut mild steel where tolerances were less critical.

Initial Fabrication

During the first few months of 2024, I gradually machined, assembled, and tested the frame. Once the frame was securely bolted together for the first time, I partially disassembled it and filled the large hollow sections of the aluminum framing with epoxy granite. The epoxy granite is designed to dampen vibrations and stiffen the frame.

My epoxy granite recipe, admittedly amateur, was based on industry best practices. The final mixture consisted of 80% granite by weight and 20% epoxy. For the granite, I used a combination of coarse granite pebbles and DG. Although sifting, washing, and drying the granite was time-consuming, the casting process itself was quick and straightforward.

At this stage of the project, the scope getting out of control. Recognizing this, I deliberately set aside the enclosure design until I had an operational machine to test.

Moving Forward

This project is still ongoing as of January 2026. Here's a breakdown of what's been completed, what's currently being worked on, and what's next: