Original content in french. Translation in progress.
The idea is to transform a classic sewing machine into an embroiderer.
Mechanical modifications must be simples and reversibles (nondestructive).
The system is designed to be accessible and easily reproductible with common tools (no heavy machining like Electrolab's ones).
The system is quite simple:
- a position sensor to know if the needle is up or down, and move only if it's outside the fabric.
- a XY table with low precision requirements (0.1mm is already pretty good). In fact there is real constraints on the frame that hold the fabric, and the precision is hard to keep a good rigidity on every mechanical parts. It impacts maximum frame size and configuration of thread strain.
- motor control: 2 viable solutions: reprap drivers, which are easy to acquire and works out of the box but are expensive. Or a fully diy driver, with a custom driver chip and pcb, which is cheaper but harder to acquire. We have chosen this last solution, because the Electrolab may sell mechanical and electronic kits, which can give an alternative solution to those who can't order parts or create/solder pcb.
Not translated yet, sorry we are french, original page here
In progress, its strengths:
- structure rigidity
- a cleaner mechanic, with standard parts (aluminum angle, reprap belts and pulleys, ...) or 3D printed in PLA. Its a step further in accessibility :)
- the belt/pulley XY table rather than ballscrew, provides increased speed but lower precision (we had too much of it anyway). Acceleration is quite the same because we have less torque but lighter moving mass.
- a compact motorization, which can be closed for a complete integration.
- a generic shape for an easier cross-vendor compatibility.
- really more beautiful.
For this version we are working with a Singer Curvy (a midrange sewing machine that we already have).
Work has to be done on needle sensor and presser foot (which highly influence thread strain.
First tests with our diy motor drivers, based on the toshiba stk682 chip, 36V, 3A continuous, large enough to attach big power dissipator. The results are an encouraging speed/torque behavior, further acceleration tests must be done but going at a 10 stitches per second speed seems achievable.
The tested motor heats quite a lot and has no acceleration ramp, we must work on that.
To show a working machine on the MakerFaire Paris we gave an extra boost on the project:
- First version of the presser foot and needle sensor part. A 2-in-1 part to reduce global complexity.
- Pcb and soldering of enough motor drivers version 2
- Firmware with acceleration ramp
- upgrade and 3D printed version of a lot of parts (frame, pressure foot, fixation clips, ...)
Finally, yesterday (between midnight and 5pm), version 2 starts its first points :D
- New firmware is out, the communication library is working, the machine is now remote controlled and stitches are streamed.
- Improvements in software, new stitches, remote control, but nothing visible since interface is not yet ready, you have to trust us :)
Update from the CCC 2015
- during the CCC we're worked on a new foot (derived from a standard embroidery one) compatible with all machine, cheap, easily reproducible and with a pressing function that missed to ours precedent model. Results are much better with this improvement
- Now we're working a lot on the computer software (we gonna push it on a git repository as soon as possible)
A video of the machine at work (at maximum speed :p) is visible on our blog: http://blog.electrolab.fr/2015/08/20/electrolabccc2015/
All the code is in an early stage, so it's quite messy but already accessible.