■ FIGURE 2. Three-axis
to sensor port.
Signal conditioning circuitry is therefore limited to the simple RC low-pass
filters that are included as standard on
the ADC inputs of the sensor port.
Figure 2 shows the connections
required between the breakout board
and the sensor port. Although it is possible to operate with different sensitivity on each axis, by changing the state
of the control pin ranges between successive ADC readings, the associated
settling time would adversely affect
the sampling rate. The range of all
three channels was therefore fixed at
the maximum specified range of ±6g.
(However, the ADC accepts signal
levels that exceed this range.)
The accelerometer is only available in a difficult-to-hand-solder surface
mount package, but this limitation was
overcome by using a small breakout
board from Spark Fun Electronics
www.sparkfun.com) that has the
MMA7260 connections brought out to
0.1 inch pads. Silicone
glue was used to mount
the breakout board
onto the flight recorder
that seemed to provide
while offering the
protection against excessive g-forces.
Figure 3 is an example plot of
three acceleration channels measured
during a model rocket flight. The vertical acceleration trace is labelled ‘X’ and
the two horizontal traces are labelled
‘Y’ and ‘Z.’ All three accelerometer
channels were sampled at 100 Hz for
27 seconds, which only used 25% of
the available storage capacity.
If we start by looking at the vertical channel, the plot shows that it was
briefly overloaded at engine ignition
but could easily accommodate the
following level of sustained thrust. At
approximately 1.5 seconds into the
flight, acceleration dropped abruptly
with propellant burnout. The rocket
then coasted upwards for about 3. 5
seconds before the vertical channel
was overloaded by the nose cone
ejection and chute deployment.
Things then became a little wild until
the nose-cone dangled, inverted at an
angle below the chute, and then
began a relatively sedate descent.
Touchdown at 18 seconds appears to
have been surprisingly gentle. The
nose cone eventually came to rest on
its side, with the vertical accelerometer measuring near zero g.
If we now turn our attention to
the Y and Z traces, we see that they
understandably show lower levels of
acceleration. They do, however,
indicate that the rocket performed a
definite ‘wiggle’ after leaving the
launch rail that decayed in amplitude
over the next three seconds.
Following touchdown, the horizontal
traces seemed to show the nose cone
rolling after touchdown. It is possible
that the initial wiggle and ground roll
were both caused by the effects (upon
the fins and chute) of a moderate
side-wind at the launch site.
Figure 4 lists the PICAXE firmware
that was used to record three channels
of acceleration during flight and export
the data to Microsoft Excel. The function of the trigger, record, and playback
routines were explained in last month’s
article. Note that the setfreq m8 command doubles the default PICAXE
clock rate to 8 MHz, which means that
data is sent at 9. 6 kbaud (not 4. 8 kbaud
as listed). The SelmaDAQ macro is
used in Excel to accept data and place
it in the spreadsheet for subsequent
processing and display.
Martin Hebel, of SelmaWare
Solutions and a professor in the
Electronic Systems Technologies program at Southern Illinois University
Carbondale, wrote SelmaDAQ. This
application only uses a fraction of the
capabilities of this extremely flexible
software tool, which may be downloaded from
■ FIGURE 3. Acceleration plot of model