Close to Your Heart

Monitoring your pulse helps you determine the health of your vascular system. If you are interested in fitness, it is one of the vital metrics in sports monitoring. Whether you are taking measurements out of scientific interest, as a metric to gauge the intensity of your workout, as a measure of fitness, or as a cautionary measure against excessive stress, a DIY pulsemeter provides quick and easy feedback.

The module used in the example here is based on an optical method that measures light absorption of the red blood cells in your veins to determine the rhythm of your heartbeat. Typically, you will want to deploy sensors like this at the wrist, the arch of the foot, temple, or neck, where the veins are just under the skin.

The sensor has an operating voltage of 3-5V, and the output voltage varies depending on how well you position the sensor. Occasionally, it will not output a value at all (e.g., if you press it against the skin so hard that it blocks the blood flow).

While taking a measurement, you should keep the sensor in the same position. The pulse sensor [1] uses green light, which is best suited for measuring at the wrist. Modules with red light, on the other hand, are used for fingertip measurements. The accuracy of the optical sensors is almost equivalent to that of an ECG.

Test Setup

Before setting up the test, you should note that the sensor provides analog values, whereas the Raspberry Pi only processes digital values. Therefore, you need an analog-to-digital converter (ADC). My choice was the MCP3008 [2], because its 10-bit resolution measures with sufficient accuracy.

The schematic (Figure 1) for the test setup is quite clear: It comprises only the MCP3008 ADC, which you connect to the Raspberry Pi through the serial peripheral interface (SPI; specifically, pins 19, 21, 23, and 24), and the sensor, which is connected to the MCP3008 analog input (CH0).

Figure 1: The heart rate monitor requires two components in addition to the Raspberry Pi: an ADC and an optical heart sensor.

In my tests, I used two different oscilloscopes to monitor the analog values provided by the sensor, so I could get a feel for the pressure needed to get correct readings. In one test, I connected a BitScope Micro [3] to my desktop PC via USB, with the BitScope input channel connected to the S output of the pulse sensor.

The output is shown in the BitScope Meter software [4] (Figure 2). The 1.4Hz output frequency corresponds to a pulse of 84 beats per minute. The hertz unit (Hz) indicates the measured oscillations per second; multiply this value by 60 to get the oscillations per minute.

Figure 2: The BitScope Meter software displays the pulse measured by the sensor in hertz. A simple multiplication yields the heart rate.

BitScope is an interesting mini-oscilloscope for your PC. If you don't have a BitScope, though, you can use any oscilloscope (in Figure 3, I use a DSO138 kit [5]) to monitor the signal of the sensor.

Figure 3: The oscilloscope displays pulses, indicating the accuracy of the sensor.

Software

A current Raspbian Stretch Lite [6] provides the underpinnings of this project. After downloading, write it to an SD card with a tool of your choice. After booting, start the raspi-config tool and activate the Rasp Pi's SPI under 5 Interfacing Options | P4 SPI. This setting is required to control the MCP3008.

Additionally, enable the SSH service under 5 Interfacing Options | P2 SSH. Once running, you will be able to manage the Raspberry Pi with PuTTY [7] or from a desktop terminal.

Last but not least, you will want change the password for the pi account, either in raspi-config or with the passwd command. Running the commands in Listing 1 updates the software and installs all the programs you need to control the sensor. To ensure that you have loaded all components correctly, reboot the computer and launch the sample program (Listing 2), which is available on GitHub [8].

sudo apt update
sudo apt upgrade
sudo apt install git python-dev
git clone https://github.com/doceme/py-spidev.git
cd py-spidev/
sudo python setup.py install
cd ..
git clone https://github.com/tutRPi/Raspberry-Pi-Heartbeat-Pulse-Sensor

$ cd Raspberry Pi Heartbeat Pulse Sensor
$ python example.py
No Heartbeat found
BPM: 80
BPM: 80
BPM: 83
BPM: 85

Conclusions

The pulse sensor presented here works quite accurately. Initially, you might need some time to figure out the pressure needed at the sensor to get clean readings. A small oscilloscope helps to get the desired result.