Introduction

This semester I'm taking the junior-level electronics course (PHY335) and this soon made me realize that I am very interested in electronics. Therefore it was natural to look for a project that involves both optics and electronics. Such a project would also help prepare me for the work I'll do this summer in Prof. Eden Figueroa's lab. The project we found uses reflected near infrared (IR) light to observe pulse strength and rate. We started out using a commercial bandpass amplifier but I later built my own amplifier with two op-amp stages, that flashes an LED on each pulse. This experience was a nice complement to the theory of filter circuits that was covered in my course.

The method I used is called photoplethysmography or PPG for short. It's based on the fact that light reflected from skin, or transmitted through a finger tip or earlobe, changes very slightly with each heart beat, as fresh blood is pushed into the tissue. For an adult over 18 years old a normal pulse rate is 60-80 beats/min or 1.0 - 1.33 Hz.

PPG method has been used in many biomedical applications and has many advantages as it uses simple and inexpensive sensors which can easily be placed in a small portable device. There are two types of sensor: transmitting and reflecting sensor. Transmission in PPG is commonly used when measuring the oxygen concentration in blood rather than heart rate measurement. Such pulse oximeter devices are more complicated than simple PPG and use two or more wavelength of light.

PPG Sensor

The sensor for our device is made from a IR LED and a photodiode mounted side by side. The matched pair of devices was purchased from Radio Shack (model 276-0142) for just about $3. The emission peak is at 850 nm; the detector responds to 750-980 nm and and is most sensitive at 900 nm. The 99.5 ohm resistor in series with the LED limits the current to half its maximum value 100 mA. We put a 10k resistor across the photodiode to convert its current into a voltage.

The emitter and detector are set up next to each other, in a small container made from black foam and black tape, as shown in the picture. To observe pulse we only need to rest our finger gently on the sensor covering both emitter and detector. Pressing too hard migh block circulation causing noise in signal or no signal at all.

 

First Experiments

We looked at Dr. Noé's pulse first, and got a strong signal right away. The amplifer gain was 100, the oscilloscope was set at 1.0 Volt and 1.0 second per division. According to the calculator in the scope the pulse rate was 78 bpm. The signal displayed was in sync with pulses felt on the wrist.

We next tested Yuning, a graduate student. We had to use a higher gain (500) and more 'scope sensitivity (500 mV/div) to get a good signal but it wasn't difficult to find one. The measured rate was 61 bpm.

Finally we tested me, the student. Even with the sensitivity turned up even more, to 200 mV/div, it was VERY difficult to get a stable signal, and when we did get a signal it didn't look like a "pulse," but more like a sawtooth. Getting a signal was very dependent on having just the right position and pressure, and keeping the finger (I used my fore-finger) very still. The difficulty could also be due to the fact that I have poor blood circulation in general.

Building an optimized band-pass amplifier

The circuit I build was described on a web site I found [1]. It includes 2 stages of active band-pass filter, and a non-inverting buffering stage. I calculated the gain to be 104 per stage or 10,816 in all. The gain is so high that the output clips at the +5 Volt rail on each pulse. The narrow band-width of 0.7 - 2.4 Hz is centered around the expected rate for an adult's pulse.

The amplified signal drives an LED light and can be observed on an oscilloscope. Some results for the signal obtained are shown below:

Dr.Noe's pulse, 1.4Hz (84bpm), 2V/div

My pulse, 1.2Hz (72bpm), 2V/div

Unanswered questions and future work

The circuit itself is proved to work well. However, we need more work on the sensor part since it couldn't detect signals from a few individuals. Even in my case, I had to be very still to get the signal detected on the scope. It would be nice to talk to the person who designed this circuit to see if his experience with it is similar to ours.

Acknowledgement

I want to thank Dr. Noe for suggesting this project to me as well as wholeheartedly guiding me throughout this semester. His encouragement and time dedication helped me so much in completing this project in a short period of time.

I would also like to thank Professor Tom Allison for the use of oscilloscpe to capture the images of the detected pulses. Finally, I would like to thank Marty for giving me some insights of what might have been the problem with the origional sensor; and many other friends who were willing to try out this device.

References

[1]Introducing Easy Pulse: A DIY photoplethysmographic sensor for measuring heart rate