User talk:RachelWinsor
(Follow this Git repo for code updates )
Design Consideration (13/6/17):
KROG Robot (sensors/modules)
- Microphone + Gain amplifier => audio detection and preparation for filtering/analysis.
The particular module combines a simple microphone with a gain amplifier to make it easier to perform a frequency analysis on the input data. Our goal with this device is to detect the presence and frequency of an audio input (music) in order to tailor the "dancing" response of KROG.
- Communication chip => communication between subject and robot.
For this communication chip, I propose using the XBee Pro S1 chip. I have more familiarity with ZigBee technology than with other communication technologies. The advantages of XBee include: low power consumption, sufficient range (up to 10m), closed network. I am open to learning other technologies, though I believe Bluetooth and WiFi would be impractical for this project (due to range and security protocols).
The communication module will be used to communicate the accelerometer/gyroscope information to the robot. This is to allow the robot to mimic the movements and speed of the user.
- Custom Shield for Arduino Nano and XBee => Using the open-source Gravitech schematic, I will build the shield necessary to connect the Arduino with the XBee.
The most important aspect of this design is the power regulator to step down from +5V to +3.3V for the XBee input.
- (potentially) Ultrasonic sensor => detect proximity of subject, objects, or Occupational Therapist (OT).
Subject (sensors/modules)
- Communication chip => communication between subject and robot.
- MPU 6050 Acclerometer/Gyroscope => detect movement of subject
This small device can be connected with the Arduino Nano in order to detect movement and rotation of the wearable device. This information will be communicated back to the robot, and then presumably mirrored.
Rachel and Prof Bonarini will meet at 10:00 on 14/6 to discuss this proposal
Design Update (14/6/17):
- Instead of using the XBee module, and building a shield from the open-source schematic (listed above), we have decided to take a simpler approach in using RF communication. I will use one NRF24L01 chip on the subject, and one on the robot. Through configuration, the subject will effectively be the "transmitter", and the robot will be the "receiver" of accelerometer data.
For the microphone/audio input on the robot, I have considered two similar devices available in the lab:
- MAX4466 Electret microphone.
This microphone comes on a breakout board, with an adjustable gain amplifier.
- MAX9814 Electret microphone.
This microphone also comes with adjustable gain, though in this board, it is automatic. This is the device that will be used, for its automatic gain control, availability of online resources/similarity to MAX4466, and its range of noise detection.
- Another board with a LM393/91M/DN1619 was considered, but there was insufficient material online for this device.
This design change was discussed and approved by Prof Bonarini on 14/6
Design Update (3/7/17):
- Problems with joining MPU-6050 and nRF24L01 libraries have been resolved in most recent update of Arduino code. There is now easy and open communication between the two RF modules, including the IMU sensor information. Once the gyro readings are filtered (with the complementary filter), this portion can be considered done.
- Next step: begin working with MAX9814 module. At this moment, I am unable to find a simple connection diagram from the breakout board to the Arduino, but should be able to use the datasheets to create one.
Design Update (13/7/17):
- MAX9814 and nRF24L01 modules working together on MEGA board (receiver). Current challenges are in multitasking on the single core of the Arduino.
- Code has been reformatted for modularity, and moved to GitHub for further development. Archived files are saved in C++ format, and can be found on the repo under archive
- Schematics of circuits can be found in the GitHub repo under fritzing_schematics
- Next step: Filter of audio input & multitasking on MEGA (receiver)
