Spending consecutive nights trying to make the PostureAroma Necklace work, being awake for long hours, being charged up by constant supply of caffeine, nicotine and indian food, Shubho and I put some serious effort to make this thing functional. We hunted alternate ways to make the sensor work accurately, viewed various forums, youtube videos, blogs etc. to find methods which could give us accurate values in the scenario in which we wanted it to work, i.e. the sensor being placed on the neck (inside the 3d printed case of the necklace). But we never came across any solution because we were making one fundamental mistake of “not keeping the sensor in a perfectly horizontal position”. Inside the 3D printed case of the necklace, the sensor was not always in a static horizontal position which made it impossible to generate a threshold value from it, which could be used to trigger the aroma mechanism. Some things which have to be taken care of while working with prototyping materials like Arduino, Sparkfun Sensors etc. is that they never function in the exact way we want it to work. There are limitations and errors which occur due to certain miscalculations we make during the process which go unnoticed. This makes it even more hard to debug and rectify the error we are making. For example, while working with the sensor initially we were using pin headers with jumper wires attached to the sensor. The pin headers were not perfectly attached with the sensor which in turn was giving us false values on our preview monitor. This error was easily rectified by soldering the pin header to the sensor, a valuable advice given to us by Corne (iFabrica guy).
Once we soldered the sensor and the pin header, we could reduce the noise from the sensor by a large amount but still it was not enough to get an accurate threshold value using which we could trigger the aroma mechanism. Taking Corne’s help, we were able to figure out the circuit for the aroma, but it became very difficult to connect the aroma circuit and the sensor circuit together. The current required or the power required to make both the things work simultaneously was not being achieved without a 9Volt battery. Even with a 9Volt battery we were not sure, that if it would work perfectly or not since, excess power or excess current could also damage the circuit/sensor/filament/arduino. Going through this long and frustrating process without any positive outcomes, Shubho and i decided to start all over again, with an Arduino Uno and test out the sensor with LED’s. In order for the necklace to work, the Arduino and the aroma mechanism should be hooked up with a battery source which is adequate enough to power both, the arduino+sensor and the filament.
Current status:
The Arduino Uno was used with the ADXL 335 Accelerometer, and was tested out using LED’s along different axis. The code was written in a way in which the LED would light up when the sensor was tilted along a particular axis. So currently the LED is working with all the 3 axes. Depending upon the movement of the sensor, it is giving constant readings but at the same time it also gives a lot of garbage values which are of no use. By modifying the code, the Z-axis was figured out and was giving the desired values but the other 2 axes were still giving out random values. Therefor with the LED, the response from the sensor is positive and gives consistent readings as per the tilt of the sensor. In trying to make this process work on the correct power source-all LED’s got burnt. In order for the necklace to work, instead of the LED, the aroma mechanism should be triggered with a tilt in the sensor. But that is still not working because there is not enough current to power the aroma circuit. Therefor in theory, the principle of the project is working, but the actual aroma mechanism is not working due to inadequate power/current supply.
