Pulse Sensor Amped Heart Rate Monitor

I am looking to use the PulseSensor Amped heart rate monitor for a project (pulsesensor.com). Arduino code is available, but I haven’t found anything for the electric imp. Does anyone know if there is programming available for the electric imp? If not, I would be willing to hire someone to do the programming and share it on the forum. I don’t need to display a real-time pulse, just take periodic measurements and post to Xively.

What a great sensor a quick look at the Arduino code and there doesn’t seem to be too much to do and some very able folk on the forum do paid stuff.

One question will the imp/device be permanently powered or battery powered?
As battery power conservation will a tad more complexity.

@joeolives Just got a chance to look at this, and it looks like a perfect match for the Electric Imp and small boards like Aria as a wearable. I’m going to order one and see if I can write up some code. I’d like to resell on MakeDeck if it works out well. I wonder if there is a less conspicuous place to wear it that works well?

With my rudimentary testing (hooked to oscilloscope) extremities seem work best like earlobe or fingertips, but if placed carefully, other areas seem possible. There seems to be a delicate balance between placement and pressure. Increased pressure seems to push the blood out of the capillaries and reduce signal return. The skin around the neck works well, so a wearable type necklace could work. The inside of the arm seems less ideal, but without processing the signal it is difficult to tell. So an armband may work. The ankle is good spot but the sensor needs to have direct contact with the skin, so that rules out sock wearers:).

My device will be battery powered so power consumption is important. I only need periodic measurements to be made and passed to Xively. I think the sensor draws too much current to run directly off one of the pins so some kind of circuit will be needed to turn it on and off.

I don’t need real-time measurement so it may be possible to use a fast Fourier transform (FFT) on a sample of the analog signal to determine the pulse. I don’t know if that is easier than the method used in the Arduino programming.

How about inside wrist?

One reliable way is to use a chest heart rate strap. Look up


It makes a big difference what you want as the scope for your project. Is it for education or are you more interested in the end. Heart rate is really difficult and is also becoming pervasive. it is even on the new Galaxy phone.

To read from the skin you need surface contact and muscle activity will affect it. That is why the chest strap works the best - it is not affected by electrical activity of your muscles.

be safe.

The chest type sensors use two electrodes to contact the flesh so they need some moisture for conductivity, and they also greatly increase cost as you need the chest sensor and receiver unit. Also, the way the Polar sensors are designed is that they are always transmitting while in contact with your skin. So for a persistent sensor this would be impractical as the batteries would wear down. My project requires long term monitoring powered by 2 AAAs or small LiPo rechargeable.

The inside of the wrist appears to be a relatively poor spot but it does seem possible. I cannot say for sure as I am still just using the oscilloscope return to judge. The pulsing of the signal seems to be relatively weak at that spot.

Shouldn’t be too hard to build an enable circuit for the sensor so that the Imp could control powering the sensor on and off, no? Then all you would need to do is wake the Imp on a schedule, power up the sensors, read, broadcast, and go back to sleep. I’ve powered an Imp module on a LiPo as small as 150mAh. It should last for a while if the Imp is properly sleeping most of the time.

That’s exactly what I had in mind. I am already measuring and passing temperature data in this manner. I was hoping to be able to add pulse data as well.

According to UPS, I have one of these sensors arriving tomorrow. I emailed the designer and they said they were willing to help port code to ImpOS.

Running the Imp on 2AA might be tricky without a boost convertor. The C3V0 design that is coming out was built around a 3V step-down convertor to help maximize LiPo life. I just ordered several LiPo’s from Adafruit and one of them has the same package size as a pair of AA cells.

MakerDeck I run all my devices of 2xAA lithium using Aprils. I bypass the regulator’s as they take too much current in deep sleep. Any data on what Aria typically draws in deep sleep? Same goes for C3V0.

I’m very interested in the pulse sensor it’s just the thing for some education work I want to do with the imp & there’s a good example of it at TedX Kids http://pulsesensor.myshopify.com/blogs/news/7328752-pulse-sensor-at-tedx-kids imagine this on an imp and graphing on an smartphone :slight_smile:

Interesting… where do you tie in? I assume the ATSHA is ok with that? Sounds like you need an Aria board with just the capacitors installed since you don’t want to use the reg. I’ll get some deep sleep numbers for you.

Don’t laugh but in on the 3.3v out
I tried, as others have, to remove unwanted components just ended up wrecking a board. Have 5 devices running a year plus in this way!

No laughing here. The easiest way to get components off is probably a heat gun and a pair of tweezers, but going in through the 3.3V output should work just fine. The caps are still in line and the ATSHA will run on your battery. Maybe a simple breakout should be made for that purpose.

Don’t laugh it’s a something us Brits say when when think we’ve been a bit of a twat :slight_smile:

Battery brakeout is neat idea. Hugo posted something a long time about have two regs one for max power and one for sleep or maybe better to use the Nora approach “uses a Torex XC9128 which is enabled by a line on the imp; this allows a true 6uA sleep”

Just in case it sparks some lateral thinking… I’ve always been impressed with the Heart Rate iPhone app (by Azumio). You place your finger on the rear of the phone, which uses a bright light (led camera flash) and a sensor (the camera) to pick up pulse. Wondered if a two LED configuration could be used, one as the light source, and one as the sensor to replicate something similar on the Imp. Perhaps place the LEDs each side of an ear lobe clip… or… get something like this:


Note to self. Read whole article next time :slight_smile: I see that the Pulse Sensor seems to use the same iPhone principle…

Just scoured the pulsesensor forum for current draw data (they don’t seem to have a spec sheet for the device). They state that current draw at 3.3V is only 3.25mA, so it looks like it can safely be connected directly and switched using one of the imp’s pins.

That is convenient… one GPIO to toggle it on and off. I think I’ll start out testing it on my earlobe, and run wires down to an Aria and LiPo clipped to the back of my collar so I can wear it around.

Scratch that, I’m going to make a heart monitor earring.

You might want to cut up a decent pair of earbuds to replace the wires on the sensor. The ones that come with it seem like they might be too stiff for that type of application. You could even preserve the phono plug for ease of connecting to the Aria.