Second in a three-part series (Part I; Part III) about upgrading to a new CNC, in this post I’ll share a bit about all the subsystems of my new machine, and a few pictures of the pieces (mostly) working together.
Frame from Montreal, Canada
One of the most important parts of the system is the frame, since its size, strength, rigidity, and motion control mechanisms place limitations and requirements on just about every other part. I purchased a z-axis upgrade from the Montreal-based supplier CNC4Newbie a few years back and was very pleased with both the product and the support, so I decided to use their New Carve frame.
It has much larger aluminum extrusions than the X-Carve, is specifically able to support an 80mm (i.e.: more powerful) spindle, and yet is only around 100#, relevant for my weight target. I purchased the 52”x35” size, allowing me to fit part of a sheet of plywood into the machine and still leave a few inches for the components of an automatic tool changer.
The frame came partially assembled, but it still took a few hours to get it all together with the motors.
The frame is mostly assembled here
CNC Controller from New South Wales, Australia
Though the controller itself is now buried under a mountain of wires and difficult to photograph, the screen - which shows tool path for an aluminum control panel that is to become part of the CNC wiring box itself - gives a tiny glimpse of the capabilities the controller enables.
My original thought was to cannibalize controller from my prior machine, but that would have had several limitations on stepper motor power, support for additional axes, support for a tool changer, and just general expandability. And, of course, I wouldn’t be able to sell the cannibalized machine!
Prior to my research, I had only heard of Mach3 as an alternative to GRBL-based controllers, and I found options at the two extremes: raw circuit boards with limited or no documentation, or expensive bespoke custom builds. A conversation with a fellow CNCer in New York pointed me to the Masso controller, which has the right amount of flexibility and capabilities, and won’t break the bank.
Stepper Motor Drivers from Los Angeles
Six stepper motor controllers: one for each of the five axes, and a sixth to pair up on the two-motor y-axis.
With a new controller, I now also need to get stepper motor drivers that convert the direction and step pulses from the controller to signals for the motor windings. The options are based on what degree - if any - of micro-stepping I want - and max motor power in amps and volts. Micro-stepping subdivides the typical 200 steps (1.8 degrees per step) of a stepper motor into finer steps, trading holding torque for accuracy. And of course, amps and volts translate to power. Also, though there are a wide range of inexpensive imported drivers, they seem to have mixed reviews, so I decided to invest a bit more in high quality domestic components here with the Geckodrive G214V
Stepper Motors from China
I find the stepper motors for positioning the axes from Automation Techologies Inc. The difficult part here is to find exactly the right motor: frame and shaft size; torque; and power requirements. The motors I ultimately picked have 2.3x the power of my prior X-Carve motors.
While researching for this post, I also came across this interesting closed loop high-torque 1/4” shaft NEMA23 motor, which had I found earlier, I might have tried since the price for this system is about the same as the Geckodrive driver & stepper motor that I purchased separately. It’s past the point of me using it, but others might find this useful.
Note that after troubleshooting the steppers, power supply, and drivers for months to address an occasional missed step, I ultimately replaced those with Teknic servo motors for the X / Y / Z axes (leaving steppers on the A & B axes); I documented that process here.
Spindle from China
The unassembled Mechatron STC80 automatic tool changer in green on the left, the 2.2kw / 3HP 80mm spindle in silver on the right, and a whole bunch of supporting hardware.
After crunching some numbers, I see that the largest tools I might be able to put in my machine would be able to remove enough material to require a 2.1HP spindle, so I opt for a 3HP spindle to provide a little headroom. The other choice here is air-cooled or water cooled; a water cooled system adds complexity and weight of a reservoir and pump to my already-complex lift system, and since I won’t be running the machine 8 hours a day, air should be more than adequate. The one limitation this imposes is that the internal fan may not move enough air to keep it cool at lower speeds, but this is a reasonable tradeoff.
Here as well, I find what I need from Automation Techologies Inc.; a better known (to me) online store had a few reviews on their spindles that basically said “this spindle is just the same one as from Automation Technologies, but more expensive”, and a few folks on the Inventables forums had mentioned this source. Additionally, they got back to my email queries about some of their products within a day. So I purchased this spindle, and this matching variable frequency drive made by Huanyang of China.
The Huanyang variable frequency drive (VFD) spindle controller is in the lower right; you can also see the partially-built bespoke wiring case for all the electronic components of the CNC.
Automatic Tool Changer from Frankfurt
This was the hardest piece to figure out and source, primarily because so few mid-level CNCs have an automatic tool changing system. But I did find three options:
Mechatron: The STC80 is an add-on to any 80mm spindle from a company in Germany that focuses on high performance spindles and tool changers more so than mid-level components.
Automation Controls Inc: They offer a spindle-ATC combo. The advantage of this is that the two components are matched, enabling me to use the larger collet of an 80mm spindle, rather than the smaller collet of the Mechatron. However, the spindle itself is water-cooled, which I was hoping to avoid, the documentation is light, and I do not get the sense that the company is an expert in ATCs, only motors, so - without reviews or evidence of other successful users - I’m less confident with this.
DIY: Frank Hermann has developed an ingenious approach to a build-your-own ATC using low cost elements sitting around. Though I may be able to work around its limitation of only handling a single size shank via collet reducers, it seems like it will add another level of complexity to building a machine that is already going to be quite complicated.
I went for the Mechatron route.
A test fit of the spindle and ATC along with the Mechatron-supplied tool holder magazine (which goes thru a few iterations before I arrive at a system that works).
Putting It Together
As the pictures suggest, I have most of the pieces assembled now. It took far longer than I anticipated, partially due to just waiting for delivery of parts as I realize I need one more custom piece of hardware, etc. But also, upgrading the machine forced (or enabled?) me to upgrade just about every other supporting system in the workshop as well, from a stronger lift system to a solid pipe (vs. flexible hose) dust collection system, to an air compressor system that provided a direct warning of low pressure.
In the next post, I’ll share some learnings from the struggles of putting all the pieces put together, and run the machine thru a few tests.