The S×3 is very close to being ready to plug in to the Raspberry Pis and collecting data!
- All have been bootloaded and are programmable via USB on my windows 10 machine.
- A python script has been written to accept data over USB and print to file via stdout.
- Alex has assisted in setting up scripts (python and shell) and my own AWS server to collect the sensor data from the RPis
Calibration of the analogue sensors is needed
- Currently the function (Steinhart-Hart Equation) which transforms the ADC thermistor resistances to temperature is giving ~37°C. I have a feeling this is incorrect…
- This slight discrepancy is possibly due to incorrect coefficients being used in the equation, I assumed the values from the Vishay datasheet would give the correct temperature.
- Potentially it’s a linear offset (due to input impedance/voltage characteristics), but testing is needed (I’ve moved onto sound for now)
- This post in electronics stack exchange explains how to convert from voltage levels (or 10 bit ADC values) to a useful sound pressure level (SPL) in dB.
- Gain will need to be tuned for optimal sensitivity. If gain is too high the Signal to Noise Ratio (SNR) is too low, and no valuable sound info is measured.
- I have removed the resistors for gain and replaced them with a 1 MΩ trim potentiometer (pot) to test the effects of gain.
- Using blink(1)
- blink(1) is a USB controlled RGB LED using the attiny85 and a neopixel. This is very close to what we are doing, and very importantly, the project is open source with a very well documented public github repo.
- The two however are not compatible and there are a lot of things to consider. See this post to get an idea.
- the biggest challenge will be adapting firmware to suit our needs/hardware.
- Preset control via arduino IDE
- I can already control the NP via terminal using preset conditions.
- An odd error occurs
Test server upload
- Alex has adapted code for the badges to work with the S×3