I’ve been working with a PCB (Printed Circuit Board) designer for the last two months to put together the board for the Sleeptrack device. The Grind Control device was built on top of breadboard-ready through-hole components which could all be soldered easily. In this new iteration, one of the objectives is to keep the size small. This means using SMD components, and I don’t have the skills or equipment to solder those. So PCB design and assembly is contracted out and a few days ago, we received the first revision of the boards.
The board is a nice looking 2”x2” (5 centimers square) board. Unfortunately, the cutting wasn’t done properly by the manufacturer (our fault, we forgot to include a design file). The above photo is taken after we sanded the corners to round them.
The same board after soldering the Photon P1. The P1 is an SMD version of the Particle Photon, which includes a built-in wifi antenna. This was soldered using a reflow oven. While this step went well, it turns out there were a number of problems with this board.
- Silkscreen missing: most of the markings on the board are missing, making it difficult to assemble components.
- Some footprints are incorrect: the speaker footprint is too large. The push button footprints are in an incorrect orientation.
- RGB LED too close to P1
Despite these issues, we went ahead and soldered on more components in an attempt to identify more issues:
On the picture above, you can see:
- Photon P1 left side. This is the heart of the device, with microcontroller and Wifi chip.
- Speaker, lower left: this small buzzer is used for feedback to the user when turning on or activating certain functions. Feedback is important as the user cannot see the device while wearing it (it will be on the side of their head).
- 3.5mm audio jack, bottom: one of the core functionalities is to have a bruxism alarm which wakes up the user after a predefined threshold of muscle activity has been breached. There is a volume-control circuit to start the alarm at a low volume and gradually increase if the bruxism activity doesn’t stop.
- On/Off switch, right side, red/black: opens/close the battery circuit.
- EMG connector, lower right: this will be used to connect to the Myoware EMG sensor. The connector is wrong, it should be a right-angled connector.
- JST Battery connnector, upper right: connects to a LIPO battery. Connector is also wrong.
- Micro-USB connector, top right: used to charge the device (charging only, as firmware flashing is over the air)
- Two push-buttons, top left: will be used to control the device, for example, starting data capture after the device has been configured.
A few more pictures with the same components:
While this board has issues, overall it feels like progress is being made, and we’re getting ready to submit revision 2. We will also try to have the assembly done by an assembly house as we feel they’ll be able to assemble boards faster, even at this stage of the project.Do you have questions about this project ?
Get in touch: firstname.lastname@example.org @SleepTrackIO sleeptrackio
Project GoalsTo build an open-source wearable device and associated cloud services designed to analyze sleep bruxism and condition the wearer out of bruxism using biofeedback techniques.
Current Status 2016/12/18
- Changed approach to accelerometer/IMU only solution
- Built Phase 3 prototype with battery charging and IMU.
- Using a chin sensor to detect jaw movement.
- Device firmware working stably.
- Backend servers (collecting data from device) stable, deployed in a docker instance.
- InfluxDB and Grafana instances deployed.
- Angular front-end allows user to interact with device, start realtime mode, or night mode and view nightly stats.