how to improve the reliability of an heart rate monitor

snipped-for-privacy@gmail.com hath wroth:

Try not moving. Whenever I get my BP/pulse measured, I'm required to keep rather still.

If you're trying to measure pulse rate while exercising, you'll probably need to do it accoustically or by directly measuring pressure, rather than optically. The change in IR density through the finger is rather small and movement will create substantial "noise". You can't tightly clamp the device to the finger as you'll restrict blood flow.

However, there is one place where you can clamp it down and not wreck the waveform with movement. The ear lobes. This is where some bicycle like exercise machines do their sensing. A contivance similar to a clothes pin is used. IR emitter on one side of the ear lobe. IR detector on the other.

The trick is to get the clamping pressure just right. Too much and you restrict circulation. Too little and it falls off. However, since you're now measuring the expansion of a thin capillary instead of a large valve, the amplitude of the pulse will be MUCH less. You'll probably find it difficult to even see a resting pulse, but it works well once the heart starts pumping.

I strongly suggest you concentrate on sensor construction and optimization rather than trying to clean up the waveforms (and mess) produced by a non-optimized sensor construction. That turns this from an electronics design exercise into a mechanical design exercise.

For example, you'll need to make sure your optics doesn't pickup stray light from the side as the ear lobes are quite translucent to IR. Burying both the IR emitter and IR detector down a small hole inside the clamp arms, is usually sufficient. See various patent drawings at:

This is one I've worked with:

I have some photos of the above (somewhere) but can't find them.

More, thanks to Google:

Thank you very much for you suggestions Jeff. I am actually using a TCRT1010 which has led and sensor on the same component: do you think that applying this to my ear lobe could be as good as applying a sensor on one a side and a led on the opposite? Thank you

snipped-for-privacy@gmail.com hath wroth:

Sure, it can be made to work. You can put a mirror on the opposite side of the ear lobe. It will work if you can keep it aligned. As slight angular misalignment of the mirror can be fatal. Also, please note that the ear lobe is fairly transparent to IR so you'll get quite a bit of leakage and light ingress. The mechanical requirements don't appear to be too formidable, but the direct path, with emitter and sensor on opposite sides, seem much easier to construct. Excessive size may be a problem depending on sensor selection. You'll also have to keep the beam with somewhat narrower as going through the ear lobe twice will create double the path length. Given the choice for a workable product, I would prefer opposing sensors, rather than the reflector contrivance. Also, the emitter and sensor are cheap, and should not be used as a driving design requirement. In other words, select your components AFTER you have selected your methodology and initial design.