Pulse Speed Timer
the pulse had propagated to the other
microphone. Naturally, that causes problems when the circuit triggers on that.
There is also the issue of triggering
on the same point of each signal. Since
the signals are over 100 ms in length
and the delay to be measured is less
than that, the trigger points must be
right at the beginning of the signal. This
works well for the pulse signal, which is
very large. However, it is more problematic with the heartbeat signal, which
is much smaller. There is a much greater
likelihood that the heartbeat signal will
be delayed until the signal reaches a
sufficient level to trigger the circuit.
It is certainly possible to measure
the signals directly with an oscilloscope
without using the signal conditioner
section. This is shown in Photo 5.
(Note: The figure is retouched to show
the beginning of the heartbeat trace. It
was difficult to get a good picture of
a single-shot event with a nonsynchro-nized camera. The zero-voltage, flat
part of the heartbeat trace is presumed
to be a negative signal that was clipped
because the op-amp is using a single
supply and can’t handle negative input
voltages.) The proper delay is from the
start of the heartbeat and the start of
the pulse. The second time-bar in not
placed properly. The proper delay is
approximately 60 ms (not 83 ms).
This procedure can provide a fairly
accurate measure of the delay. In this
case, the delay is about 60 ms from
heartbeat to wrist. The straight-line
distance is 27 inches. This corresponds
to about 2. 2 mS per inch or 27 mS per
foot. Converting to speed: 454 inches
per second or 37 feet per second.
the same size and ■ PHOTO 5. Actual delay
triggered at the same measurement taken at TPP
point. One way to and TPH (see Figure 1).
improve matters Actual delay is about 60
significantly is to cou- ms (not 83 as shown). The
ple the microphone beginning of the heartbeat
signal was retouched for
to an inexpensive clarity. (It’s very difficult to
stethoscope. This was synchronize a camera to
tested and a very nice an oscilloscope!)
signal of over 100
mV before the amplifier was obtained.
This is very similar to the pulse voltage.
These signals should be passed
through an AGC (automatic gain control) to get them to match well. Then a
comparator could be used to create the
digital signal. (This would make the project too complex for most readers, however.) The polarity of the signals should
also be the same. The pulse moves the
skin outward to start. Depending on
where you place the sensor, the heartbeat may or may not do the same.
It should also be noted that the
reference article refers to aortic pulse
velocity, not peripheral pulse velocity.
So it is not proper to directly compare
the numbers presented there with the
delays measured here. (Additionally,
some of the statistics appear suspect.)
And always remember to leave the
medical diagnoses to the professionals.
It is possible to measure the propagation speed of the blood pulse from
the heart to the forearm with a simple
circuit. In the process of building this,
practical experience in high-gain
amplifiers and mixed-signal designs
can be obtained. NV
In order to reduce the variability of
the digital conditioning circuit, the
pulse and heartbeat signals need to be
Aortic Pulse Wave Velocity:
An Independent Marker of
Cardiovascular Risk, Michel Safer
MD, Oliver Henry MD, Sylvie
Meaume MD, 2002 Medscape. www.
April 2007 43