Atomotion is a flexible and powerful modular control system that makes it possible to upgrade existing robots, 3D printers, CNC machines to be more intelligent and connected or build new systems that are more complex and powerful with features like servo control and extensive sensors. It will be based on open sourced interfaces for simple customization using your own module designs. What will you build?
Last year, our team started the development of a new 3D printer called the Apparatus ARC.
In order to make it both more user-friendly and cost effective, we decided to use a full feedback servo setup using brushless DC motors and inexpensive magnetic encoder sensors. To do this we needed a fast microcontroller with a lot of IO. We found is that our options for making a prototype were either not able to handle the full requirement, not flexible enough for different types of motor setup, or were just too expensive to prototype with.
We compromised by doing a more limited prototype with a board that used the K22 microcontroller (ARM Cortex M4) and then designing our own custom board with a version of that same controller with more IO available to create our full prototype.
We’ve had good results with using it so far, this video shows system calibration and homing the Z only using the feedback on the belts and from the motors.
•K22 Freescale/NXP ARM Cortex m4 Microcontroller with Floating Point Processor
•Wifi Module
•Full Size USB-B Connector for Direct Connections
•Micro USB-B Connector for Flashing (Possibly MBED Compatible)
•OpenCNX 164 Connector
•K22 Freescale/NXP ARM Cortex m4 Microcontroller with Floating Point Processor
•i.MX6SL Freescale/NXP ARM Cortex A9 Microprocessor with Graphics Accelerator
•Wifi Module
•Ethernet via RJ-45 Connector
•CAN Bus via RJ-11 Connector
•Full Size USB-B Connector for Direct Connections to the K22
•Micro USB-B Connector for Flashing (Possibly MBED Compatible)
•OpenCNX 164 Connector
•Supports 4X OpenCNX 34 Compatible Modules
•1X OpenCNX 164 Controller
•Power Input up to DC 35V 32A
•Power Conditioning and Voltage Regulation for All Onboard Power Supply
•Supports 8X OpenCNX 34 Compatible Modules
•1X OpenCNX 164 Controller
•Power Input up to DC 35V 64A
•Power Conditioning and Voltage Regulation for All Onboard Power Supply
•Four Toshiba TB67S109AFTG(QFN48) Stepper Driver ICs (3A Drive Current, 1/32 microstepping)
•One K02F Freescale/NXP ARM Cortex m4 Microcontroller for High Frequency Stepping and Motion Control
•Four 3-pin Limit Sensor Connectors
•OpenCNX 34 Dual Compatible
•One Toshiba TB67S109AFTG(QFN48) Stepper Driver ICs (3A Drive Current, 1/32 microstepping)
•One K02F Freescale/NXP ARM Cortex m4 Microcontroller for High Frequency Stepping and Motion Control
•Two 3-pin Limit Sensor Connectors
•One I2C Header
•Two Input Pin Header for AB Quadrature or Other Sensor Feedback
•Supports Servo Control Using Quadrature Sensor (Sensor is separate to the module.)
•OpenCNX 34 Compatible
•Two 2-Pin High Current MOSFET Drives for Heaters (@ System Input Voltage, Maximum 8A Combined)
•Two 2-Pin Low Current Drives for Fans (12V, 0.5A Each)
•Two Thermistor Connections for Temperature Sensing Up To 280C
•OpenCNX 34 Compatible
•Two 2-Pin High Current MOSFET Drives for Heaters (@ System Input Voltage, Maximum 8A Combined)
•Two 2-Pin Low Current Drives for Fans (12V, 0.5A Each)
•Two Thermistor Connections for Temperature Sensing Up To 280C
•Two Thermocouple (K Type) Connections using 2X MAX31855 IC (up to 1260C)
•OpenCNX 34 Compatible
•General IO module:
•I2C interface
•2 PWM Out
•4 IO Pin
•2 Analog Pin
•3.3V plus Ground
•5V plus Ground
•12V plus Ground
•OpenCNX 34 Compatible
•Logic Chip Based Laser Fire Safety
•Combines Chiller Status, 2X Door Switches, and Input from Controller Module
•PWM or Analog Drive for Most External Laser Power Supplies
•Current Controlled Drive for 5V Laser Diode (Max 1A)
•OpenCNX 34 Compatible
We’ve gotten a first iteration of the system planned and prototyped. We now have a clear roadmap for what we think is a final 1.0 design, but we want feedback from you! To create a production ready design for both the Controller Module and the Controller Pro Module, we need to complete around $35,000 in pre-orders. Less than that, we still will be able to complete the Controller Module first and will do the Controller Pro as we have additional sales. We’ve done most of the design work for the Function Modules that we are showing, but will do a final schematic revision and PCB layout based on the V1.0 of the specifications. The Controller Module is a more simple version of our existing ARC controller, so the final design and layout will go quickly. The Controller Pro Module includes the i.MX6SL, which will be a new processor for us, but is widely used and well documented. The tooling for each casing is just a single mold, and will cost us less than $3,000 to produce. We have local connections in China to either run a limited production run or for a full scale production, so we will be able to go ahead in any situation.
Software is the other part of this, since no microcontroller is useful without it. Our plan is to contribute a Smoothieware porting to that project to make it compatible with our hardware. If the project is successful enough, we will consider developing more advanced software tools to make creating robots or CNC machines much more simple and robust. Things like system kinematic modeling from CAD models or self-configuring systems based on the modules that are loaded.
We think that the Atomotion system will make a difference in how people prototype new projects and products that require more complex automation. 3D printers, laser cutters, general CNC machines, and robots are now using simple solutions that limit what they can do, the next step is to make them more intelligent and accurate with better features.
Our team has experience with the full process of product development, manufacturing, and fulfillment. We have shipped around 2000 3D printer kits as former founders of MakiBox, we helped Ambi Climate with the production of several hundred beta units for their launch, and we have previously been involved in product development and manufacturing for several consumer product companies.
Some people may be concerned with what happened with MakiBox since there were about 1000 customers left waiting for their orders. It has taken us two years to recover from what was a very bad situation. As Apparatus, we fully intend to ship ARC 3D printers to any customers that did not receive a MakiBox kit that they ordered. Our investors support doing this from sales profit. By supporting this campaign, you also help us to get another step closer to being able to realize that promise.
We’ve already mentioned it, but, yes, this is the same founder as MakiBox. That was the early days of crowdfunding and 3D printing. We got caught up in the enthusiasm of the day and made mistakes of both trying to do too much R&D and also allowed for too much growth without getting caught up on outstanding orders. The short version of the story is that we ran out of cash a month before we would have finished all outstanding orders and a few of our investors made it impossible for us to recover the operations. Instead of just hiding or avoiding this, we are working to take care of those customers that were caught out. The collapse of the company left us in pretty bad shape both financially and mentally, and it has taken us two years to rebuild to the point we can start to make good on that intent.
That same trouble was a very good learning experience. Since then we have continued to explore how to make manufacturing more accessible for smaller runs, and in the long term, we plan to work on new technologies that help close that gap. The Atomotion system is a tool we need in our toolbox to do that. We’ve experienced both the best and worst sides of building a business, so we can say that we have what it takes to finish this project and get it into your hands this year.
Being located in Hong Kong and having been in the consumer product industry for 16 years, the manufacturing part of this project is very manageable. If we need to do a smaller production run, we will probably do it with a smaller SMT PCBA shop in Shenzhen, the same that did the prototype run for the ARC driver board. For larger volume production, we will use a local factory owned and run by Swedish companyNote.
The biggest risk is with the Controller Pro module since it is using a new processor to us, but our experience is that Freescale has extremely good documentation for their products, and our team has experience with boards of this level of complexity. Being located in this region, makes it easier to manage this risk and to get prototypes made quickly and in a very cost effective way.
If you would like to help, please spread the word: