Project Based Engineering Instrumentation High Level Coding and Microcontrollers

The light sensor on the CPX is just a simple photocell wired in series with a resistor. A photocell (or Light Dependent Resistor, LDR) is a sensor whose resistance changes significantly based on the intensity of light falling on its sensitive surface. In dimmer light, its resistance is high, and in brighter light, its resistance is low. By integrating it into a simple voltage divider circuit, Adafruit makes it possible to read the voltage across the photocell and correlate that to light level. There is a relevant Adafruit Tutorial on Photocells and the code required to measure the voltage if you’d like to read more about it[32].

The sound sensor uses the audiobusio library and creates a mic object using the (Pulse Density Modulation) PDM library. You have to set the sample rate and the number of bits to use to capture the data. We’re going to set the bits to 16 to utilize the whole spectrum and then set the sample rate to 16 kHz. It’s not quite 44.1 kHz like most modern microphones but it will do. After creating the mic object we have to compute some root mean squared values and thus two functions are defined before the while true loop in the code. The code itself is shown below. The code starts on line 22 because the first 22 lines are copyright from Dan Halbert, Kattni Rembor, and Tony DiCola from Adafruit Industries[32]. I have edited the code a bit to fit my needs and uploaded my version to Github[33]. In the code line 23-27 import standard modules as well as some new ones. The array module is used to create array like matrices. The math module is used to compute functions like cos, sin, and sqrt. Then of course the audiobusio module is used to create the mic object on line 42. Notice the two functions defined on 33 and 39 which create a function for computing the mean and for computing the normalized root mean square value of the data stream. Basically what’s going to happen is we’re going to record 160 samples as defined on line 160. So on line 43 we create a hexadecimal array (hexademical: base 16 hence the num_bits set to 16 on line 31) with 160 zeros. In the while true loop we’re going to sleep for 0.01 seconds and then record some samples. Since we’re sampling at 16 kHz the time it takes to record 160 samples is 160/16000 = 16/1600 = 1/100 = 0.01 seconds. Since we’re taking 160 samples we need to compute some sort of average which is why the normalized root mean square value is computed on line 48.

When I run this code and talk normally into the microphone, I get this output in the Plotter. You’ll notice that the data is pretty noisy in the beginning but then there are noticeable humps in the data. This is me saying something random into the microphone at normal volume. It’s possible we could increase the number of samples we take each loop by editing line 30 but that would slow down our code. So there’s a tradeoff between filtering here and speed. That’s something will investigate in some later labs.

The temperature sensor is actually a thermistor[63]. A thermistor is basically a “thermometer resistor” which means the resistance depends on temperature. Since this thermistor on the CPX/CPB is connected to an analog pin, you can read the analog signal coming from the thermistor just by reading the analog signal from pin A9. If you look for the thermometer symbol on the CPX you’ll see pin A9. Therefore, it is possible to just use the analogio library and just read in the analog voltage but in order to convert to celsius and then fahrenheit you need to use some heat transfer equations to convert the analog signal to celsius.

where \(\beta=3950\) is a heat transfer coefficient specific to the bulk semiconductor material over a given temperature range of interest and \(T_0=298.15K\) is the nominal temperature of the semiconductor in Kelvin. Note that in the equation above, the output is in Kelvin.

If this all seems complex, the folks at Adafruit have done it again with an adafruit_thermistor module. If you head over to their github on this module you’ll see the relevant conversion under the definition temperature which at the time of this writing is on line 127[33]. The Adafruit Learn system also does a bit of work to explain the conversion from voltage to temperature[32] but understand that the tutorial on that page assumes that the thermistor is on the low (grounded) side and thus the voltage divider equation is different. If you look at the adafruit_thermistor module you’ll see that there is a self.high_side boolean that changes between either a thermistor on the high side or the low side. In this case, for the built-in A9 thermistor the thermistor is on the high side. Given all this complexity, we will just appreciate the simplicity of the code below which uses the adafruit_thermistor module. I’ve also created a simple version of the code to read the analog value of the thermistor and posted it on my Github[33]. Note the code below is from a much earlier and simpler version of my code which only reads the temperature in Celsius. The code on my Github has more functionality such as printing the raw ADC value and the voltage.

For the code above, lines 1-3 import the relevant modules and then line 8 create the thermistor object. You’ll notice the input arguments are the pin which is A9 as well as the resistor values which are in series with the thermistor. These resistors are soldered to the PCB so they are fixed at \(10 k\Omega\text{.}\) The 25 is for the nominal resistance temperature in celsius of the thermistor and 3950 is the b coefficient which is a heat transfer property. Running this code and then placing my finger on the A9 symbol causes the temperature to rise just a bit. You’ll notice the temperature rise quite quickly when I place my finger on the sensor but when I remove the sensor it takes some time before the sensor cools off. This has to do with the dynamic response of the sensor. We’ll discuss this in some future labs on dynamic measurements. For now you can move on to the accelerometer.

The accelerometer is a 3-axis sensor. As such, it is going to spit out not just 1 value but 3 values. Accelerations in x,y and z or North, East, Down or Forward, Side to Side, Up and Down. Since it’s reading 3 values we can’t just read 3 analog signals (we can but the accelerometer chip design didn’t want to do that) so instead we’re going to use the I2C (I like "Indigo" and 2C like "Squared C". So "I Squared C". Not "12C" or "one two C". It’s "I squared C") functionality. I2C is a type of serial communication that allows computers to send strings rather than numbers. I2C is beyond the scope of this course but just know it’s a type of serial communication that uses hexadecimal addresses. It’s a much more complex form of communication but since it’s a standard form of communication, we can just use the busio module which contains the I2C function.

In this code we see alot more imports than normal. In addition to the standard time, board and digitalio modules we need the busio module and the adafruit_lis3dh. You might think that LIS3DH is a very weird name for an accelerometer but it’s actually the name of the chip on your CPX. The chip itself is very standard and is well documented on multiple websites. Here’s one from ST[66]. You can also buy the chip on a breakout board from Adafruit and then of course the Adafruit Learn site has plenty of tutorials on reading Accelerometer data in CircuitPython[32][34]. As always I’ve learned what I can from the relevant tutorials and created my own simple version to read the accelerometer data and posted it to Github[33]. I digress, lines 8-11 of the code do alot. It first uses the SCL and SDA pins to set up an I2C object which establishes serial communication to the accelerometer. Line 9 creates an interrupt which is beyond the scope of this course. Finally, line 10 creates the actual accelerometer object by sending it the I2C pins, the hexadecimal address in the I2C protocol and finally the interrupt pin. Line 11 then sets the range. Line 14 in the while loop is where the x,y and z values of the accelerometer are read and then promptly printed to Serial on line 15. If I run this code and shake the sensor a bit I can get all the values to vary. If you put the CPX on a flat surface, the Z axis will measure something close to 9.81. The units of the accelerometer are clearly in \(m/s^2\text{.}\)