HOW TO BUILD A WIRELESS HEART RATE RECEIVER.
(c)Tomas Eriksson (marianne_tomas@yahoo.com), 2000-10-01







Abstract
This paper describes a receiver for the signals from a POLAR wireless heart rate sensor. This receiver makes it possible to replace cheap and unreliable photo-coupler "ear clip" heart rate sensors with the well known and very reliable POLAR wireless heart rate sensor in all kinds of exercise equipment. A circuit diagram and a components list is included. Finally, a way of connecting the receiver to a PC, and a Q-Basic heart rate monitoring program is discussed.

Contents

1. Background

2. Design
2.1 Block diagram
2.2 Circuit diagram
2.3 Components list
2.4 Enclosure
2.5 User interface

3. Performance
3.1 Range
3.2 Accuracy

4. Notes from the development work
4.1 The old heart rate sensor
4.2 The bike computers heart rate sensor port
4.3 The new POLAR wireless heart rate sensor
4.4 Inductive pic-ups

5. Notes concerning future development
5.1 Connect a PC to the receiver

1. BACKGROUND

This receiver was born because I bought an exercise bike (Spartacus MAG 6006) which had a very nice computer, but unfortunately also an extremely unreliable heart rate sensor. Since I could not trust the heart rate measurements, or use the computers cool heart rate controlled load regulation, I decided to replace the old heart rate sensor with a new wireless heart rate sensor from POLAR. However, to be able to connect the wireless sensor to the bike computer, I had to design and build a receiver which could receive the heart rate signals from the POLAR wireless sensor, and convert them to signals that the bike computer could understand.

2. DESIGN

2.1 Block diagram



It might look strange that there is no block dedicated to shaping the output signal from the receiver, but it turned out that the rectified output from the detector is a pulse which is accepted by the bike computer as a heart beat.

2.2 Circuit diagram


2.3 Components list



2.4 Enclosure



2.5 User interface

There are no controlls on the front panel, only 2 LED:s. A green LED indicates when power is fed to the receiver, and a red LED flashes for each heart beat. The receiver turns on automatically when an exercise program is started on the bike computer, and it turns off when the program is finished. Two cables goes from the receiver. One cable is the power feed, which I´ve connected to a 9V source in the bike computer, the other cable is the connection between the receiver and the heart rate sensor port of the bike computer.

3. PERFORMANCE

3.1 Range

The receiver pics up the signal from the wireless heart rate sensor at a distance of more than 1 meter. The range can be extended by adjusting the gain of the amplifiers, but the risk of picking up disturbances from other electrical devices like TV-sets is then increased. The range obtained with the components in the component list is enough so that I can't get out of range while I'm sitting on the bike, no matter how much I try to bend away in any direction. And, an important matter is that with the present gain setting, I can have a TV-set turned on at a distance of 2 meters, without any problems with disturbances.

3.2 Accuracy

I´ve tested the accuracy of the system "heart rate receiver & bike computer" by comparing heart rate readings from the bike computer with heart rate readings from a POLAR wrist watch/receiver while exersizing. The diference between the readings never exceeds 3 heart beats, and is mostly 0 or 1 heart beat. The fact that the biggest diferences occurs when the heart rate is increasing or decreasing, indicates that the method of calculating heart rate is not quite the same in the bike computer, and in the POLAR wrist watch/receiver.

4. NOTES FROM THE DEVELOPMENT WORK

4.1 The old heart rate sensor

The old heart rate sensor turned out to be a photo coupler.





It shines infrared light from the IR-diode, through your ear lobe, and to the photo transistor.
The flow of blood thorough the ear lobe varies with each heart beat, and thus the amount of light that reaches the photo transistor. Since the current through the photo transistor is proportional to the light that reaches it, the current will be modulated by the heart beats. The bike computer amplifies these very small modulations and from them calculates the heart rate.

4.2 The bike computers heart rate sensor port

I tested how critical the input wave form from the sensor to the bike computer was, by simulating heart beats by quickly inserting and removing a piece of paper between the IR diode and the photo transistor in the old sensor. The signal from the sensor should look something like a square wave, and nothing like a real heart beat. The result was encouraging. The bike computer had no problem to read this input

Another set of measurements indicated that the sensor input port of the bike computer was equivalent to the following circuit diagram.



Terminal A supplies current to the IR diode, terminal B is input for the current from the photo transistor, and terminal C is ground. The voltage on terminal A and B is 2.6V when an exercise program is running on the computer, but 0V when no program is running. I decided to use this characteristic to controll the power to the receiver, so it would only be on when a program was running. A connection like the one shown in the next figure seemed to be a good idea.



4.3 The new POLAR heart rate sensor

After a bit of market research, I decided to buy a package including a wireless heart rate sensor and wrist watch receiver from POLAR, I selected the "Pacer" modeI since it was relatively cheap, but still advanced enugh to be of use if the receiver project became a failure. When I'd received the package, I started to investigate the wireless heart rate sensor (scroll to bottom of page). I had no idea about which frequency, modulation or encoding the heart rate sensor used. To find this out, I made a set of measurements. Since I don't have an oscilloscope, or the other necessary discrete instruments, I used my PC and Ulrich W Müllers software package called "Audio Tester", which emulates a function generator, a sweep frequency generator, a spectrum analyzer, and an oscilloscope, using a PC and a sound card. My PC is a Compaq Pressario 5670 with a 450MHz processor, and my sound card is a "Soundblaster Live", and they worked splendidly in this measuring task. Using an old coil from a microphone as pic-up, and lots of patience, I caught the output signal wave-form from the heart rate sensor.



At each heart beat, a burst looking like the above figure was emitted from the wireless heart rate sensor. The burst consists of four decaying groups of 5.25 kHz sinewaves, each group is phase shifted relative to the previous group. Oscilloscope settings were 500uV/div and timebase 2ms/div. The measurement was made with the microphone coil pickup at a distance of 10cm from the wireless heart rate sensor.

4.4 Inductive pic-ups



I tried several inductive pic-ups before starting to design the receiver. I wanted to know how much gain I would have to provide in the amplifiers. The first pic-up was a coil from an old microphone, for the second I used a ferite sleeve and wound 50 turns around it, with slightly better results. The third one was an old LW antenna from a radio, and the fourth and final one I made by winding 0.3mm Cu wire 3600 turns around a ferite rod with 8mm diameter and 150mm length. The number of turns was not calculated, I just used all the wire I had. Since I got such a good result with the fourth pic-up, I decided to keep it and design the receiver for the signal strength it provided, that is: about 1mV peak to peak at 1m distance.

5. NOTES CONCERNING FUTURE DEVELOPMENT

5.1 Connect a PC to the receiver.

Information about how to use the printer port can be found here for example: http://home.maine.rr.com/randylinscott/dec98.htm

It should be possible to connect the open collector output of the receiver to an input pin in the printer port, and write a program in Q-Basic that monitored that pin.

The program should measure the time between heart beats, and between each heart beat calculate the pulse (beats/minute), and plot the value in an X-Y chart. As a bonus, it would be nice if the program could regulate the load of the bike, in order to keep the heart rate at the wanted level

If you have any ideas or questions about the wireless heart rate receiver, please send a mail to: marianne_tomas@yahoo.com