January
2005, Issue 174
Light-to-Frequency
Conversion (Part 2)
Pulse
and Oxygen Content
PULSE
Figure
1 is similar to the graph I presented in Part 1. You
can see the systolic and diastolic phases of each heartbeat.
The former is the contraction, or working, phase of
the heart when the pressure is highest. As the pressure
increases, more blood squeezes into the arterial system.
More blood absorbs more light. Less light getting through
reduces the frequency output of the TSL230R. Lower frequencies
require longer period counts. Thus, the data has higher
counts during the systolic phase. During the diastolic
phase, the heart rests and pressure drops as blood draws
back to the heart in preparation for the next heartbeat.
|
(Click
here to enlarge)
|
Figure
1—The TSL230R sensor’s sample period counts were
taken every 31.25 ms. Code algorithms pick out the
maximum and minimum peaks used to determine a heart
rate. Peak values are allowed to leak off after
a change in slope has been detected when a sample
exceeds the opposing peak. |
You’ve
had your blood pressure checked. Usually, a cuff is
placed on your upper arm and pumped until it cuts off
the flow of blood. A stethoscope is used to listen for
returned blood flow as the pressure in the cuff is released.
The numbers you receive from this test like 120/80 are
actually your systolic pressure over your diastolic
pressure. It’s the systolic reading that identifies
potential high blood pressure.
What
you need to extract from the TSL230R’s data is the amount
of time for a complete systolic-diastolic cycle. In
this case I’ll determine the systolic (maximum) peaks
and calculate the heart rate based on the number of
samples between consecutive peaks. Although the output
data in Figure 1 might be typical, the actual wave shape
might be quite different from person to person. (I’m
sure a specialist could glean a lot of information about
your body’s performance from this wave shape. But I’m
not a doctor, and I don’t play one on TV.)
One
of the biggest problems when looking at the sampled
data is the level of constant absorption. It isn’t really
constant. This level varies because of source lighting
changes or light path changes. You can control light
source changes by creating a stable current supply for
the LED used as a light source, but beyond creating
a snug yet comfortable sensor, even the slightest twitch
or movement of the measured appendage will change the
path of light and look to your sensor like an AC component
where you’re considering it a (constant) DC level. This
creates havoc when you’re trying to measure the minimum/maximum
excursions because of systolic and diastolic pressures.
