The M&F group buy was a gas. It was great fun to study favorite handle shapes, absorb them, work obsessively to come up with a classically traditional but also unique design, figure out how to convert that into a CAD file I could hand off to Lee, and then have him turn that into prototypes exactly matching the drawing, followed by production of over 100 brushes for B&B group buy participants and some others including family and friends.
But as satisfying as the whole project was, it also created an itch. I wanted to be able to make brush handles myself. Not in large quantities, but in the same way, using CNC to control a lathe and enable high fidelity execution of a design in fine detail.
So in early January my son and I made a road trip from Kansas City to La Crosse, Wisconsin, to buy a 25ish-year-old, bench-top, CNC lathe I'd found. Then I got side-tracked and didn't have time to do anything with the lathe until a few weeks ago. The learning curve for me has been humbling. I spent a lot of time hanging around in shops when I was a kid, ran heavy equipment on the Santa Fe railroad during a two-year layout from college, and had summer jobs working in an iron foundry that made combine cylinder bars and a steel fixture plant that fabricated industrial furniture. But I'd never operated a lathe and I didn't really know what CNC meant. There turned out to be a lot more challenge in both than I anticipated.
After much study and preparation, I spent several hours on Father's Day (with some help from my son) trying to turn my first handle shape from a piece of 1.5" oak dowel. Working with different coordinate systems, various set-offs, and tool-path compensation confused hell out of me. I stuck with it, however, and finally managed to make a piece that just barely suggested the Chief. Then somehow my settings got screwed up, and all I was able to produce the rest of that day were tool crashes.
But yesterday I finally figured out what I was doing wrong with off-sets and compensation. I started over, mounted a fresh piece of maple dowel, and let her rip. The result in the form it came off the chuck, un-sanded and un-cut, is shown in the bottom photo, between a group buy Chief that I dissected and the Father's Day mutant creation.
First below is a segment of the G-code from the start of the program, and following that is a backplot.
I still have a long way to go, but it finally feels like I've made a start.
But as satisfying as the whole project was, it also created an itch. I wanted to be able to make brush handles myself. Not in large quantities, but in the same way, using CNC to control a lathe and enable high fidelity execution of a design in fine detail.
So in early January my son and I made a road trip from Kansas City to La Crosse, Wisconsin, to buy a 25ish-year-old, bench-top, CNC lathe I'd found. Then I got side-tracked and didn't have time to do anything with the lathe until a few weeks ago. The learning curve for me has been humbling. I spent a lot of time hanging around in shops when I was a kid, ran heavy equipment on the Santa Fe railroad during a two-year layout from college, and had summer jobs working in an iron foundry that made combine cylinder bars and a steel fixture plant that fabricated industrial furniture. But I'd never operated a lathe and I didn't really know what CNC meant. There turned out to be a lot more challenge in both than I anticipated.
After much study and preparation, I spent several hours on Father's Day (with some help from my son) trying to turn my first handle shape from a piece of 1.5" oak dowel. Working with different coordinate systems, various set-offs, and tool-path compensation confused hell out of me. I stuck with it, however, and finally managed to make a piece that just barely suggested the Chief. Then somehow my settings got screwed up, and all I was able to produce the rest of that day were tool crashes.
But yesterday I finally figured out what I was doing wrong with off-sets and compensation. I started over, mounted a fresh piece of maple dowel, and let her rip. The result in the form it came off the chuck, un-sanded and un-cut, is shown in the bottom photo, between a group buy Chief that I dissected and the Father's Day mutant creation.
First below is a segment of the G-code from the start of the program, and following that is a backplot.
I still have a long way to go, but it finally feels like I've made a start.
Last edited: