I've done some miscellaneous research to help prepare for designing the tool changer. I've summarized some of what I've learned below.
The E3D tool changer uses kinematic couplings to attach the tools. Kinematic couplings are meant to exactly constrain the attachment, without either over or under constraint. E3D gets repeatability of 3 microns using kinematic couplings. While they don't seem to be a very popular topic, there are a number of sources on them, including a whole website by MIT on kinematic couplings. They seem simple enough that I plan on using them for attaching tools.
At this point, I want to make the tool changer fully detach the tools when they are not in use. To connect, I am planning on using pogo pins, spring loaded electrical connections often used to test circuit boards. The problem with pogo pins is that they don't support very large amounts of current. A 3D print head can draw 3 amps of current, and a milling tool can draw even more. I consulted with a TA at the university electrical engineering shop, and he suggested molex connectors for the higher power connections.
Software for this project is possibly going to be the hardest part. I am somewhat familiar with the open source Marlin 3D printer firmware, and it is compatible with my control board. As I am planning to disconnect the tools, and likely communicate over I2C, I will have to change some significant parts of the firmware. I think that Marlin 2.0, might be best for this, even though it is still in beta. Marlin 2.0 has board abstraction layers, that may allow me to insert serial communications in between the logic and the board layer.In any case, Marlin 2.0 has some other features than could be very useful for tool changing.
The E3D tool changer uses kinematic couplings to attach the tools. Kinematic couplings are meant to exactly constrain the attachment, without either over or under constraint. E3D gets repeatability of 3 microns using kinematic couplings. While they don't seem to be a very popular topic, there are a number of sources on them, including a whole website by MIT on kinematic couplings. They seem simple enough that I plan on using them for attaching tools.
At this point, I want to make the tool changer fully detach the tools when they are not in use. To connect, I am planning on using pogo pins, spring loaded electrical connections often used to test circuit boards. The problem with pogo pins is that they don't support very large amounts of current. A 3D print head can draw 3 amps of current, and a milling tool can draw even more. I consulted with a TA at the university electrical engineering shop, and he suggested molex connectors for the higher power connections.
Software for this project is possibly going to be the hardest part. I am somewhat familiar with the open source Marlin 3D printer firmware, and it is compatible with my control board. As I am planning to disconnect the tools, and likely communicate over I2C, I will have to change some significant parts of the firmware. I think that Marlin 2.0, might be best for this, even though it is still in beta. Marlin 2.0 has board abstraction layers, that may allow me to insert serial communications in between the logic and the board layer.In any case, Marlin 2.0 has some other features than could be very useful for tool changing.
Comments
Post a Comment