This is only a preview of the September 2005 issue of Silicon Chip. You can view 36 of the 112 pages in the full issue, including the advertisments. For full access, purchase the issue for $10.00 or subscribe for access to the latest issues. Items relevant to "Build Your Own Seismograph":
Items relevant to "Bilge Sniffer":
Items relevant to "VoIP Analog Phone Adaptor":
Articles in this series:
Items relevant to "PICAXE In Schools, Pt.4":
Purchase a printed copy of this issue for $10.00. |
Build Your Own
Seismograph
Ever wondered how a seismograph works?
Here’s one that you can build yourself. It uses
a horizontal swinging pendulum to detect
earthquake waves and you can even display
the results on a PC.
By DAVE DOBESON*
M
OST AUSTRALIANS are thankful that we are not seriously
affected by the large earthquakes and
volcanoes that regularly devastate so
many other parts of the world. However, few realise just how close we are
to much of the tectonic action, or how
easy it is to make your own amateur
seismograph.
The design described here can easily detect the half-dozen magnitude
7 quakes that occur around Australia
each year. In fact, the author has observed three major quakes occurring
“live” on the monitor, including one
from El Salvador.
Plate tectonics
Before we take a look at the design
of our seismograph, let’s first find
26 Silicon Chip
out why major earthquakes occur.
In particular, we need to have some
understanding of “plate tectonics”.
The basics are very simple – the
crust of the earth is made up of about
20 major “plates” that “float” on
semi-liquid layers underneath. In our
region, the Australian-Indian plate
(including the ocean floor out to NZ,
Fiji, PNG, Indonesia, most of the Indian Ocean, and also India) is moving
in a north-westerly direction by about
7cm each year. Over millions of years,
India (which is at the leading edge of
the plate) has “crashed” into Asia,
forming the Himalayas.
Earthquakes commonly occur at
the boundaries of the plates, where
they collide and produce stresses in
the Earth’s crust . For example, deep
ocean faults off the coast of Sumatra
produced the magnitude 9.0 “Boxing
Day Tsunami” earthquake last year
and the related Niass 8.7 earthquake
in April, 2005.
Also associated with this plate are a
number of volcanoes, including Krakotoa, which partly circle Australia
from Indonesia, through PNG and
down through NZ. Macquarie Island,
situated half-way between Australia
and Antarctica, had a magnitude 8.1
earthquake on Christmas Eve, the biggest in the world last year until the
Boxing Day earthquake.
Of course, many large earthquakes
go unreported because they occur
under the ocean or in sparsely populated areas and have no impact on
humans.
siliconchip.com.au
This seismograph plot shows a magnitude 6.5 quake that occurred in PNG on
April 11, 2005. A 6.8 quake near Noumea was detected only five hours later.
The detector circuit used was the same as described here but the data logger
was one of the types used in NSW high schools.
The mechanical section of the
seismograph uses parts that are
readily available from a hardware
store. It’s based on a swinging
horizontal pendulum and movement
is detected using a vane and light
sensor circuit mounted at one end.
If you look at the United States Geographical Survey (USGS) home page
and click on “Recent Earthquakes” (to
show the last seven days’ earthquakes
for the US and the world), you will see
that many of the larger earthquakes
occur near the boundaries of our
continental plate – see www.usgs.gov
In addition, Geoscience Australia’s
website (at www.ga.gov.au) has a table
that gives information on recent earthquakes in Australia and significant
worldwide quakes (just click on the
“Recent Earthquakes” link).
Both sites also have detailed information on the tectonic forces causing
earthquakes, the design and operation
of professional seismographs, records
of historically significant quakes and
links to records in other countries.
siliconchip.com.au
Another site that’s worth visiting is
www.geonet.org.nz/drums – it shows
“live” displays from seismographs
around NZ. If your home-made seismograph detects a real earthquake, the
event should also be reported within
minutes by the above three sites.
Designed for schools
This do-it-yourself seismograph
was originally described in “Scientific American” in 1979 and has been
adapted for science teachers to build
and use in the school laboratory – see
http://science.uniserve.edu.au/school/
Seismograph
Movements of the seismograph,
which is basically a horizontal pendulum, are detected using a simple
light sensor circuit. In operation, a
metal vane attached to one end of the
pendulum (or bar) partially blocks
the light between a LED and an LDR
(light-dependant resistor). However,
when the bar moves (ie, during an
earthquake), the amount of light falling on the LDR is modulated by the
metal vane.
This signal is then fed to a low-cost
op amp circuit which, in turn, feeds
into a data logger. Finally, the output
of the data logger is fed to a computer
to store, display and print the results.
All high schools in NSW have edu-
cational data loggers for use in experiments. Most of these units cost well
over $1000 but a cheap, 4-channel,
10-bit serial data acquisition device
(DI-194RS) from DATAQ in the US is
available from Turnkey Solution for
under $60 plus GST and delivery – see
www.turnkey-solutions.com.au
There’s an even cheaper way
around this problem for the home
enthusiast. A PICAXE-based A/D converter and a freeware graphing program called “StampPlot Lite” can do
the same job for about $10.00 – provided you also have a PC.
Building the seismograph
OK, let’s take a look at the mechanical details of our seismograph and find
out how it’s built.
The seismograph described here is
known as a “Lehman” or “Horizontal
Pendulum” seismograph. It’s also called a “Swinging Gate Seismograph”,
because the bar and its supporting
wire look like an old-fashioned farm
About The Author*
Dave Dobeson is a science teacher
at Turramurra High School and
the University of Sydney Science
Teacher Fellowship holder for 2005.
September 2005 27
TOP PIVOT POINT
(25-35CM ABOVE
LOWER PIVOT POINT)
This labelled photograph clearly shows how the Seismograph is
built. This version uses a magnetic damper but liquid damping
could also be used (see text and photos). Note that the light
sensor and A/D converter unit shown here is an early prototype.
TURNBUCKLE
STEEL WIRE 1-2MM DIA.
2-3KG MASS
DAMPER
METAL VANE
LIMITING
BOLTS
BAR: 5/16-INCH x 800MM
THREADED STEEL ROD
LIGHT-SENSOR & A/D
CONVERTER CIRCUIT
gate. The “hinges” (actually the pivot
points) of the “gate” are not quite vertically aligned, with the top hinge just
forward of the bottom hinge so that
the “gate” will swing shut. In practice,
this means that the horizontal pendulum (or bar) swings slowly back to its
original resting position
The accompanying photos show the
basic set-up. As can be seen, it includes
an 800mm-long 5/16-inch threaded
steel rod that’s fitted with a 2-3kg mass
at one end. The other end of the rod is
ground to an edge and pivots on the
end of a ½-inch bolt – this forms the
lower pivot point.
The supporting wire is attached
to the rod at one end, just before the
weights, and to a turnbuckle at the
other end. This then pivots about 2530cm above the lower pivot.
If we align the seismograph pivots
so that the top pivot is less than 1mm
forward of the bottom pivot, then the
seismograph bar will always swing
back to its central position and will
have a natural period of about 5-10
seconds. However, if the pivots are
exactly vertically aligned, there will
be no restoring force and it will never
swing back. We cannot move the top
pivot too far forward though, otherwise the seismograph will be very
insensitive.
28 Silicon Chip
This unit is very sensitive to the
mostly horizontal motion of earthquake “L-waves” but is insensitive to
“P-waves” which are mostly vertical.
Kiwis, because they are much closer
to the action, might be able to detect
P-waves if they use a spring instead
of the wire. Perth, Tennant Creek and
Yass also have small local quakes
every few months, so you might like
to experiment with a spring support
if you live in these areas.
By t he way, it’s important to remember that although we often talk about
the bar (or pendulum) of the seismograph “swinging”, it’s really the room
that moves during an earthquake. The
bar, because of the inertia of a heavy
mass attached to one end, initially
stays still. In effect, the unit and its associated logger act as a low-pass filter
which renders the unit insensitive to
everyday events (footsteps, doors closing, passing traffic, etc).
The accompanying photos show
most of the construction details. The
only critical dimension is that the top
pivot must be less than 1mm in front of
the lower pivot. As well as the wooden
frame shown, the unit could be built
into any strong cupboard, bookcase,
shelf or even a strong, metal frame.
In that case, the brackets and wooden
frame would not be needed. Any type
BOTTOM
PIVOT POINT
TILT ADJUSTMENT
BOLTS
RIGHT-ANGLE BRACKETS
WITH DIAGONAL STAYS
of metal rod could be used (as long as
it’s strong enough) and the same goes
for the mass at one end.
Note that you will have to “re-zero”
the seismograph for the first few weeks
after building it, as the wire, brackets
and wood flex under the strain. After
that, it will be a matter of making routine adjustments every few months.
Top pivot point
The top “hinge” (or pivot point)
is made by drilling a 5mm diameter
hole about half-way through the outer
section of a large, thick washer – ie,
to make a “dimple”. Smaller washers
and a nut are used to hold the large
washer in position, while a nut and
lockwasher behind the wooden upright panel lock the bolt in place.
As shown in the photos, the hook at
the end of the turnbuckle sits in this
dimple, so that it can freely pivot. In
operation, the turnbuckle adjusts the
tilt of the bar and is set so that the bar
is horizontal. The securing bolt can be
screwed in or out to move the top pivot
point relative to the bottom pivot. This
is important for the overall functioning
of the seismograph because it affects
the natural period of the bar (ie, the
time for one complete swing from the
centre to one side, then back through
the centre to the other side and finally
siliconchip.com.au
back to the centre again).
A period of about five seconds seems
to work best for my seismographs in
Sydney.
The pivot end of the 5/16-inch
threaded rod is ground to a knife-edge
and this sits vertically against the end
of a ½-inch bolt. Wind a nut onto the
rod before you cut and grind it, so that
the thread is restored when the when
the nut is removed. Be sure to use
safety goggles when drilling, cutting
or grinding metals – you only have
one pair of eyes.
Note that the lower mounting
point must be directly below the upper mounting point. The best way
to ensure this is to use a plum-bob
made from fine fishing line and a
lead sinker.
The two rear-most vertical bolts
that go through the support brackets
are used for tilt adjustment – see
photo. These both screw into threads
that are tapped through the wooden
base and the brackets (nuts under the
wooden base will do) and each has
a screwdriver slot cut into its end.
This allows you to use a screwdriver
to tilt the seismograph sideways and
forwards or backwards, to alter the
position of the bar and thus its period
and sensitivity.
The far end of the seismograph
wooden frame has a single central
support. A sheet of plywood or particleboard underneath will stop the
three supports from sinking into the
carpet when the unit is positioned on
the floor.
Swinging the weight
Just about any mass of 2-3kg will
provide sufficient inertia to initially
keep the bar still during an earthquake,
provided it doesn’t hang too far below the bar. A pair of 1.25kg barbell
weights are ideal for the job. They cost
less than $3 each from a sports store
and come with a ready-made hole
through the middle. This means they
can be simply slipped over the end of
the bar and clamped in position using
nuts and washers on either side.
Damping
Once earthquake waves set the bar
swinging, it will keep swinging for
hours unless it is damped. Perfect
damping would stop the bar with a
few swings but in practice, under 2-3
minutes is OK.
You can use either liquid or magnetic damping. For liquid damping,
a 40 x 50mm plastic paddle dipped
into a rectangular container of water
will do the job. You can use a small
bulldog clip to attach the paddle to
the bar. The water will need topping
up each week or so.
Magnetic damping involves attaching one or two super magnets to the end
of the bar using a U-shaped bracket.
A thick sheet of aluminium or a coil
of wire with the ends joined is then
placed in the magnetic field.
When the bar moves (ie, during an
earthquake), current is induced into
the aluminium or wire coil. This in
turn produces a magnetic field that
counters the magnets and so damps
the motion of the bar.
Discarded computer hard disks are
a good source for super magnets but
be careful – supermagnets are dangerous and the author has been badly cut
when a pair decided to play “north
attracts south with my hand in be-
The hook at the end of the turnbuckle
sits in a 5mm dimple that’s drilled
into a large washer to form the top
pivot point.
The lower pivot point is formed by
first grinding the end of the bar to a
sharp edge. This sharp edge then rests
vertically against the end of a 1/2-inch
x 40mm-long bolt.
tween”. They can also be a disaster if
they get too close to your credit cards
or a computer monitor!
On the other hand, the good thing
about magnetic damping is that once
Above & right: these two views show the alternative
damping methods for the swinging bar. Magnetic
damping (above) uses a couple of super magnets
and a coil of wire, while liquid damping (right)
uses a 40 x 50mm plastic paddle dipped into a
rectangular container of water.
siliconchip.com.au
September 2005 29
This side-on view clearly shows
the tilt adjustment bolts. These
are set so that the base is perfectly
horizontal (both east-west and
north-south), so that the pivot
points are in the same vertical
plane. The turnbuckle is then
adjusted so that the bar is also
horizontal.
TILT ADJUSTMENT
BOLTS
you get it right, it stays right.
Old aquarium air pumps have coils
of fine wire, which can be used for
magnetic damping if the ends of the
wires are joined together. A 400g coil of
0.7mm enamelled wire with the ends
joined together and a super magnet
that moves inside the coil gives almost
perfect damping.
Use your multimeter to check that
the winding hasn’t burnt out before
using the coil.
The perfect location for your seismograph is on a concrete block that’s set
into bedrock at the bottom of a sealed
mine shaft! If you don’t have access
to a mine shaft(!), the seismograph
should be installed in a closed room
or cupboard, or in a strong bookcase
surrounded by a Perspex cover (to
prevent air movement over the unit).
Circuit details
Many different seismograph detector and A/D (analog-to-digital)
converter circuits are available on
the net. The best-known site is called
the Public Seismic Network at www.
psn.quake.net (in California). It has
designs that go from pens writing on
rolls of paper to very complex circuits
with low-noise op amps, 16-bit A/D
converters and damping using negative feedback.
By contrast, the circuit used here
is quite simple – see Fig.1. As previously stated, it’s based on a light sensor
circuit that’s interrupted by a metal
vane attached to the end of the bar. In
practice, the unit is set up so that the
vane normally blocks about half the
light from the LED to its LDR.
The light detector and its associated
op amp circuit is exactly the same as
the one designed for use with school
data loggers. The logger output is
simply taken from the output of IC1,
as shown. Alternatively, you can add
your own data logger, based on A/D
converter stage IC2 (a PICAXE-08M).
In greater detail, power for the circuit comes from a 9V DC plugpack supply. Diode D1 provides reverse polarity
protection, while the associated 100W
resistor and 470mF capacitor provide
decoupling and ripple filtering.
The filtered DC rail is used to power
LED1 via a 1kW current limiting resistor. The LDR and its associated 10kW
resistor effectively form a voltage
divider across this supply rail, the voltage at their junction varying according
to the resistance of the LDR. This in
turn depends on the amount of light
reaching it from the LED.
The output from the LDR is fed to
the inverting (pin 2) input of op amp
IC1 (741) via two back-to-back 470mF
capacitors. These capacitors block the
DC component at the output of the
LDR while allowing signal fluctuations to be fed to the op amp. They
also block any slow variations in the
LDR signal due to thermal variations
in the room.
IC1 functions as an inverting amplifier stage. Its non-inverting input (pin
3) is biased to half-supply using two
In the prototype, the LED & the LDR were brought out through
holes in the case, with the vane sitting between them – see above.
By contrast, in the final version, the LED & LDR are inside the
case and the vane rides in a slot. The vane is positioned so that
it normally “shadows” about half the LED body.
30 Silicon Chip
siliconchip.com.au
10kW resistors, while its gain can be
varied from 0-10 using potentiometer
VR1, which is in the feedback loop.
Note that although the circuit shows
a 741 op amp, you could also use an
OP27 device for improved accuracy.
IC1’s output appears at pin 6 and
is fed to a voltage divider consisting
of two 3.3kW resistors. The top of this
divider (ie, at pin 6) can be used to
directly drive an external data logger.
Alternatively, the divider output (at
the junction of the resistors) can be
used to provide a nominal 0-5V signal, which may be required by some
loggers.
Pin 6 of IC1 also drives trimpot VR2
and this is used to set the maximum
signal level into pin 3 of IC2 (to about
4V). IC2 is programmed to function as
an A/D converter, using the simple
program shown in the accompanying
panel (more on this later). Its output
is taken from pin 7 (P0) and fed to pin
2 of DBF9 socket CON2.
This socket is in turn connected to
the serial port of a PC, to provide the
alternative data logger.
The PICAXE-08M is programmed
via pin 3 of the DBF9 socket. The
incoming data signal is fed to pin 2
(SER IN) of the IC via a voltage divider
consisting of 22kW and 10kW resistors.
Power for IC2 is supplied via 3-terminal regulator REG1. This provides
a regulated +5V rail to pin 1.
Building the circuit
Building the circuit is easy since all
the parts are mounted on a small PC
board coded 04105091. Fig.2 shows
the assembly details.
Note that REG1 and the PICAXE
(IC2) are required only if you don’t
already have a data logger. If
you do have a logger, these
parts can simply be left
out, along with the DB9F
socket, trimpot VR2, the
100nF capacitor and the 22kW
and 10kW voltage divider resistors from pin 2 of IC2.
Begin by installing the re-
Par t s Lis t
1 PC board, code 04105091,
123 x 57mm
1 9V DC plugpack
1 2.1mm DC power socket, to
suit plugpack (CON1)
1 DB9F connector, PC mount
1 plastic utility box, 130 x 67 x
44mm (UB-3 size)
4 9mm-long untapped spacers
4 M3 x 15mm machine screws
4 M3 nuts
3 PC stakes
1 serial computer cable (see text)
2 8-pin IC sockets
1 100kW linear potentiometer
(VR1); Jaycar Cat. RP-8518
1 5kW horizontal trimpot (VR2)
1 Light Dependent Resistor (LDR1)
1 3-way pin header
4 10kW
1 100W
1 3.3kW
Plus 1 x 10kW or 1 x 3.3kW or 1
x 1kW to match LDR resistance
– see text
Mechanical Parts
1 800mm-long x 5/16-inch
threaded steel rod
5 5/16-inch nuts and washers to
suit rod
1 50mm-long x 1/4-inch bolt
3 1/4-inch nuts and washers
1 40mm-long x 1/2-inch bolt
1 1/2-inch nut and washers
1 3/8-inch washer
1 1-metre length 1-2mm diameter
steel wire
2 bull-dog clips to suit
1 2-2.5kg mass (eg, 2 x 1.25kg
barbell weights)
1 piece of thin aluminium sheet
(to interrupt light beam)
1 50 x 50mm piece of aluminium
or rigid plastic for paddle (see
text)
2 small bolts & nuts to fasten
paddle to bulldog clips
2 braced right-angle brackets,
250 x 250mm
8 1/4-inch x 40mm bolts, nuts &
washers
3 5/16-inch x 100mm roundhead bolts, nuts & washers
1 wooden base, 900 x 250 x 20mm
1 wooden back, 400 x 250 x 20mm
Semiconductors
1 741 or OP27 op amp (IC1)
1 PICAXE-08M microcontroller
(IC2)
1 7805 3-terminal regulator
(REG1)
1 1N4004 diode (D1)
1 red or white high-brightness
LED (LED1)
Capacitors
3 470mF 25V electrolytic
1 100nF MKT (code 104 or 100n)
Resistors (0.25W, 1%)
1 22kW
2 1kW
sistors and capacitors. Table 1 shows
the resistor colour codes but it’s also a
good idea to check each resistor using
a digital multimeter before soldering
them into circuit, just to make sure.
Follow these parts with diode D1,
the two IC sockets (don’t install the
ICs yet) and trimpot VR2. Take care
to ensure that D1 and the electrolytic
capacitors go in the right way around.
LED1 can go in next. Bend its lead
down through 90° close to its body
before installing it at full lead length
on the PC board – ie, the centre of the
LED should be about 22mm above the
PC board (see photo). Again, take care
to ensure that it’s oriented correctly.
That done, you can install the LDR
but there’s just one wrinkle here. The
10kW resistor shown in series with
the LDR on Fig.1 is correct for most
LDRs. However, some LDRs have a
Table 1: Resistor Colour Codes
o
o
o
o
o
siliconchip.com.au
No.
1
4
2
1
Value
22kW
10kW
1kW
100W
4-Band Code (1%)
red red orange brown
brown black orange brown
brown black red brown
brown black brown brown
5-Band Code (1%)
red red black red brown
brown black black red brown
brown black black brown brown
brown black black black brown
September 2005 31
REG1 7805
100
LED1
K
OUT
100nF
10k
LDR1
K
470F
9V
DC IN
1
6
470F
10k
VANE
ON
SEISMIC
MASS
CON1
SERIAL
OUTPUT
CON2
DB9F
SENSITIVITY
VR1 100k
A
470F
25V
7805
GND
GND
A
D1
1N4004
IN
+5V
OUT
IN
2
7
3
IC1
741
4
6
VR2
5k
4
3.3k
10k*
(SEE TEXT)
1k
3
Vdd
P0
P1
P3
2
3
IC2
5
PICAXE P2
-08M
SER 2
IN
P4
Vss
5
22k
8
H
L
E
10k
7
3.3k
10k
LED
SC
2005
LOGGER
OUTPUT
SIMPLE SEISMOGRAPH
1N4004
A
K
K
A
22k
IC2
PICAXE
3.3k
IC1
741
5
470F
470F
E L H
5k
10k
VR2
3.3k
470F
2
3
100nF
10k
K
CON2
LDR1
10k
A
LED1
(SLOT IN BOX ABOVE)
CON1
REG1
7805
10k
100
5002 ©
1k
9V DC IN
D1
19050140
(BEND LEADS
SO LED FACES
LDR1)
10k
1N4004
Fig.1: the circuit uses a light detector based on LED1 & LDR1 to detect movement of an interrupter vane placed
between them. The resulting signal is then amplified by IC1 and fed to the logger output. IC1 also drives IC2, a
PICAXE-08M chip programmed to function as an A/D converter. Its output can then be fed to the serial input of a
PC, to provide an alternative data logger.
DB9F
VR1
100k
LOGGER
OUT
Fig.2: install the parts on the PC board as shown here, making sure that all polarised parts are correctly oriented.
IC2, REG1, VR2 and CON2 can be left out if you already have an external data logger.
04105091
© 2005
Fig.3: this is the full-size etching pattern for the PC board.
32 Silicon Chip
lower resistance than others in the
presence of light and you may have to
adjust the value of the series resistor
accordingly.
That’s easy to do – just measure the
resistance of the LDR in a brightly lit
room and use a series resistor that’s
about the same value. The value isn’t
all that critical. In practice, you can
buy 1kW, 3.3kW and 10kW resistors
and use the one that’s closest to the
measured LDR value.
The LDR is mounted in similar
fashion to the LED – ie, bend its leads
down through 90° before installing it.
It should be mounted so that its face is
siliconchip.com.au
This view shows the fully
assembled PC board. Note the
arrangement for the LED & the LDR.
directly aligned with the LED.
Regulator REG1 is mounted with
its metal tab flat against the PC board.
To so this, bend its leads downwards
by 90° about 5mm from its body, then
secure it to the board using a 3M x
6mm machine screw and nut before
soldering its leads. There’s no need
for a heatsink, as it supplies just a few
milliamps to IC2.
The board assembly can now be
completed by fitting CON1, CON2,
potentiometer VR1 and a 3-pin header
for the external logger interface.
Serial cable options
A standard serial cable is used to
connect the PC board to the computer
(if you’re using a PC as the data logger).
There are several options here.
First, you could go out and buy a
serial cable but that’s the expensive
way of doing things. It’s far better to
scrounge a cable instead. For example,
if you have an old modem (left over
from your dial-up days), you can use
its serial cable (you did keep it, didn’t
you?) to connect to the PC.
Another possibility is to use a serial
cable from a discarded mouse. Just cut
the cable off close to the mouse, then
strip the wires back and use a multimeter to identify which lead goes to
which pin on the socket – you need
to use the leads that go to pins 2, 3 &
5 (the rest can be trimmed off).
These leads can then be soldered
directly to three PC stakes mounted
at the appropriate points on the PC
board. As a bonus, you don’t need the
siliconchip.com.au
Above: a slot is cut into
one end of the case to provide
access for the metal vane that’s
attached to the seisomograph bar.
on-board DBF9 socket, which means
you can save even more money.
Checks & adjustments
Before fitting the two ICs, it’s necessary to make several voltage checks.
First, connect a 9V DC plugpack supply and switch on. The LED should
immediately come on. If necessary,
adjust it so that it shines directly on
the LDR.
Next, use a digital multimeter to
check the voltages on IC1’s socket
pins. Pin 7 should be at the supply
voltage (about 9V, depending on the
plugpack), pin 2 should change when
the light to LDR is suddenly inter-
rupted and pin 3 should be at half
supply voltage. That done, check for
+5V on pin 1 of IC2’s socket and for
0V on pins 2, 3, 7 & 8.
If it all checks out so far, disconnect
the plugpack and install IC1 (but not
IC2). You now have to adjust trimpot
VR2 so that the voltage on pin 3 of IC2
can never exceed 5V. This is done as
follows:
(1) Connect a clip lead across the two
back-to-back 470mF capacitors (ie,
short them out);
(2) Set both VR1 and VR2 to their midrange positions;
(3) Place a piece of thick cardboard (or
other opaque object) between the LED
September 2005 33
Tectonic Plates, Earthquake Waves & The Richter Scale
”An earthquake is the way the Earth
relieves its stress by transferring it to
the people who live on it.” – Dr Lucy
Jones, USGS.
E
ARTHQUAKES occur when adjacent blocks of the Earth’s crust
slide past each other along a fracture
we call a fault line.
Most active faults are located
near the boundaries of the Earth’s
tectonic plates. These plates move
in several ways: (1) they can slide
past each other; (2) they can move
away from each other (diverge); or
(3) they can move towards each
other (converge).
For example, the west coast of New
Zealand’s South Island – which is at
the eastern edge of the AustralianIndian plate – moves north along
the Alpine Fault. This movement is
relative to the eastern side of the
island, which is part of the Pacific
plate. This area experiences several
magnitude five quakes every year, as
well as much larger but less frequent
earthquakes.
Plate divergence generally occurs
at mid-ocean ridges such as the
Atlantic’s, which rises above sea-level
to form Iceland’s central rift valley.
Convergence occurs at “subduction zones” like the one that caused
Aceh’s Boxing Day earthquake. Here,
the northern edge of the AustralianIndian plate is descending under
Indonesia, which is part of the Eurasian Plate.
While most active faults are located
near plate margins, about 10% of
active faults occur well away from
the plate margins. The earthquakes
generated in these locations are
known as intra-plate earthquakes
and are mostly thought to occur either
as a response to stress transmitted
through the plate from its interaction
with adjacent plates or from thermal
equilibration, which can cause con-
and the LDR (to block the light);
(4) Reapply power and check the voltage at pin 6 of IC1. It should be about
1V less than the supply rail;
(5) Monitor the voltage at pin 3 of IC2’s
socket and adjust VR2 for a reading of
4V (or slightly less).
34 Silicon Chip
traction as the plate cools down or
expansion as the plate warms up. The
Northern Territory’s Tennant Creek
fault is a world-famous example of
one of these intra-plate structures
and generates a number of generally
small earthquakes each year.
Several types of vibrations are
generated as blocks of rock grind
past each other during an earthquake
and these propagate around and
through the planet as different types
of earthquake waves. The fastest
(and the first to arrive) are “Primary”
or P-waves, which are longitudinal
compressional waves that propagate
at speeds of 1.5-8km/s (depending on
rock density). The next fastest are the
“Secondary” or S-waves which are
shear waves (or transverse waves)
and these propagate at speeds of
about 3.2-4.8km/s.
Both P and S-waves move through
the body of the planet and are
refracted and reflected as they encounter rock density and composition
changes. However, S-waves cannot
propagate through the liquid part of
the Earth’s core. In fact, it was by
examining the geographic pattern of
P-waves and S-waves that led to the
formulation of the core-mantle-crust
model of the Earth.
The slowest waves are surface
waves, which propagate at speeds of
about 2-5km/s. There are two types
of surface waves: Rayleigh and Love
(L) waves. It’s the shear and surface
waves that generally cause the damage associated with earthquakes.
By measuring the time gap between the arrival of the P and S
waves, it’s possible to calculate how
far away the earthquake was from the
seismograph. This is roughly 500km
for every minute between their arrival.
The location of the epicentre is determined by a form of “triangulation”.
To do this, a circle corresponding
to the calculated distance is drawn
Once that’s done, disconnect the
plugpack and install the PICACE-08M,
with its notch facing to the left – see
Fig.2.
Final assembly
The PC board is designed to fit
around at least three different seismograph locations on a map of the
region. Where the circles intersect
is the likely epicentre. Most earthquakes occur at depths of less than
100km.
P waves have higher frequencies and are best detected with a
“Short Period (one second or less)
Vertical Seismograph”, while S, L
and Rayleigh waves have lower frequencies and are best detected by a
“Long Period (10 seconds or longer)
Seismograph”, such as the design
described here. Professional seismic
stations have short, long and wideband seismographs mounted northsouth, east-west and also vertically,
with both low and high-sensitivity
detectors.
Analysis and filtering of the seismic patterns allows the arrival of
each type of wave to be determined
from the mixture of P, S, L and Rayleigh waves, reflections (PP and
SS waves), refracted waves and
alternative path surface waves. Our
seismograph with a 1-second (or 10
second) sample rate, will probably
only detect S waves and the much
larger displacement L waves and
Rayleigh waves. If you live very close
to the action, such as in NZ or PNG,
you might also detect P waves.
The Richter value, devised by
Charles Richter in 1935, is basically
a logarithmic measuring scale. It’s
calculated according to the largest ground motion waves that are
detected 100km from the epicentre
of the earthquake. Because the
scale is logarithmic, a magnitude 7
earthquake has 10 times the ground
motion (and more than 30 times the
energy) of a magnitude 6 quake.
The Aceh Earthquake measured
9.0 on the Richter scale and released many thousands of times more
energy than the 5.6 Newcastle earthquake of 1989.
inside a standard UB3 utility case.
It’s mounted on the lid on four 9mm
untapped spacers and secured using
M3 x 15mm long screws and nuts.
That done, you have to make a cutout in one end of the case to provide
clearance for the DBF9 socket (CON2)
siliconchip.com.au
and the pot shaft. This cutout measures
45mm long x 12mm high and is about
12mm from the lip of the base.
Alternatively, if you’re not using
CON2, the serial cable can be run
through a small hole in the case and
secured using a small cable tie. The
same applies if you are connecting an
external logger to the 3-pin header.
You also need a hole directly in-line
with the DC power socket (CON1).
This is horizontally centred 17mm
from the lip of the case and should be
drilled and reamed to 8mm.
Finally, a slot must be cut in the
case in line with the light sensor to
provide access for the vane that’s attached to the bar. This slot should be
positioned 37mm from the end of the
case and can be about 4mm wide. The
unit can then be assembled into the
case and attached to the base of the
seismograph.
Position the vane so that it normally
blocks about half the light between the
LED and the LDR.
Programming the PICAXE
To program the PICAXE, you first
have to download the free “Programming Editor” from www.rev-ed.co.uk/
picaxe
That done, connect the board to
your computer via the serial cable
(this should be done with the computer off) and download the simple
program shown in Listing 1 into the
PICAXE chip.
If you increase the logging interval
to 10 seconds by changing line 5 to
“wait 10”, you can keep a continuous
seismograph record for up to a week.
You could also hang a piezo transducer
off the PICAXE and add an “Alarm”
loop to the program to warn you if b1
exceeds a certain value.
Once the program is loaded and
running in the PICAXE-08M (check
by looking at the “debug” screen), you
must close down the PICAXE Programming Editor to free the COM Port, so
that the StampPlot Lite program can
use it. StampPlot Lite is available free
from www.selmaware.com
Fig.3: this simulated plot of an earthquake was produced during final sensitivity
tests of the seismograph. A gentle puff of air aimed towards the seismograph
masses from two metres away produced the first “earthquake” waves, while
similar puffs from one metre gave the full scale deflection.
(3). Click on “Connect” and “Plot
Data” – the program should immediately begin to graph the values sent by
the PICAXE-08M. You can test this by
blowing on the bar from a distance of
about one metre. Adjust the sensitivity
control (VR1) for full-scale deflection.
The “action” near the bottom of the
screen indicates that data is being
collected.
(4). Set the maximum number of points
to 200,000 or higher.
(5). A “Time Span” of 400 seconds will
show each swing of the bar during testing but increasing this to 25,600 will
let you see most of a night’s recording. Australia is normally a long way
from the action and different types
of earthquake waves will continue to
arrive for more than an hour after a
distant quake.
(6). Click on “Save data to file” so the
program saves the data as a .txt file.
siliconchip.com.au
Acknowledgement: thanks to Dr Tom
Hubble of the University of Sydney for
his geological knowledge and neighbours Jo and Manfred for computing
and design assistance.
Program Listing 1
Using StampPlot Lite
StampPlot Lite is the logging program. Once it’s installed, you need to
carry out the following steps:
(1). Set the COM port so that it’s the
same as the port that connects to the
PICAXE.
(2). Change the Baud rate to 4800.
(7). Click on “Clear min/max on reset” and you will be able to see if any
values have been detected that are
significantly above the background
line (ie, an earthquake) and when
this occurred (approximately). If you
deselect “Connect” and “Plot Data” to
stop the recording, you can look back
at stored parts of the graph by moving
the bar next to “Enable Shift”. The running graph can be seen on the screen
and “.txt” values can be exported to
Excel and graphed.
(8). Click on “Time Stamp” so that
Excel will show “Time” on the graphs.
Good luck and I hope that the Earth
SC
moves for you.
main:
readadc10 4, b1
debug b1
sertxd (#b1,cr,lf)
wait 1
goto main
'makes an A-D conversion of the value at input 4 and sends to b1
'allows you to see the value at b1 on the Picaxe debug screen
'sends the value of b1 out to the StampPlot Lite program
'sets the time gap in seconds between readings
'makes the program loop back to the start
September 2005 35
|