What should concern me when testing my heart-rate monitor prototype?
Recently I came across a humorous video on YouTube. I can't believe we made it showed many activities that were common when I was a kid, but are considered dangerous by today’s standards. While it may seem downright irresponsible to let children get involved in most of these activities (some would actually get us in trouble with the law), it was not a big deal back then. But because some people got hurt, sick, or worse, we have learned to be more careful. The video made me reflect on activities we do today, and how we will someday look back and think, “How did we ever do that?” Fortunately, electrical engineering has become less dangerous than it used to be (±300-V supply voltages for op amps like the K2-W are no longer necessary). Yet, we still find ourselves in situations where we are walking into unsafe territory.
During my days as an electrical engineering student in the late 90s, we had to choose a capstone project. Interest in biomedical instrumentation was growing, so my team and I decided to design a portable ECG. The goal was to diagnose difficult to detect arrhythmias by monitoring heart rate. Back then, we understood that isolation was necessary for a final product, but never really stopped to think too much about the development stages. Neither did we understand that test equipment may or may not be isolated, or even the type of isolation that may have been used.
We quickly identified that acquiring the signal was the most important first step, so we got an AD620 instrumentation amplifier and a few op amps for filtering and right-leg drive. To achieve isolation, we applied power through a 9-V battery and used a dc-to-dc converter to generate the ±15-V supplies. We purchased some silver/silver chloride electrodes, and twisted the wires between the electrodes and the breadboard to avoid noise pickup. So far, so good. Now, a test subject had to open his shirt, stick the electrodes on, and… Oh yes, we needed to observe what came out the other end, so we probed the output with an oscilloscope.
The thing about oscilloscopes is that their ground is supposed to be connected to earth ground. So, they ground the system, thereby violating the isolation requirement. We were now exposing ourselves to become conductors of leakage currents. What was worse, these could flow right across our chest. And, because nothing ever works the first time, we were probing with the scope and a bench DMM at the same time, all while having someone connected to the electrodes. If you know a thing or two about leakage currents on isolated supplies, you might be wondering, “How is this guy still alive?”
Fast-forward 15 years. Today, thanks to increased health awareness and wearable computing devices, heart rate monitoring is going mainstream. This has increased the number of people experimenting with heart rate monitors, such as the AD8232, or looking into newer alternatives for high-quality ECG recording with AD620 successors such as the AD8421 and AD8422. This has also increased the number of unsuspecting engineers that may be putting themselves at risk. I like to remind my fellow EEs to be careful, and to make sure that they understand, and follow, safety guidelines before testing prototypes on live subjects. For this purpose, several resources are available in print and on the Web. If you have any doubts, commercial ECG signal generators can be purchased at relatively low prices, increasing your chances of surviving the 2010s!
Tuite, Don. Simple Heart Monitor AFE Multiplies Opportunities for Medical OEMs. Electronic Design, November 2012.Zarola, Tony.
Webster, John. Medical Instrumentation Application and Design, 4th Edition. Wiley, 2009.