This is only a preview of the March 2006 issue of Silicon Chip. You can view 35 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. Articles in this series:
Items relevant to "PC-Controlled Burglar Alarm System, Pt.2":
Items relevant to "AVR ISP SocketBoard":
Articles in this series:
Items relevant to "Phone/Fax Missed Call Alert":
Purchase a printed copy of this issue for $10.00. |
Salvage It!
BY JULIAN EDGAR
A low-cost large-display
anemometer
Live in a windy area? Like to have a big dial
showing the outside wind speed? Here’s an
anemometer that you can build for next to
nothing.
I
F YOU’RE A SAILOR or kite flyer
it’s a must to know wind speed; and
even if you’re neither of these, it’s fun
to watch the display. The measurement range here will depend on how
you set it up but typically you’ll be
able to read speeds from just a few
km/h upwards. Cost? Well, depending
on how you source the components,
you’re looking at not much at all!
And best of all, this is a project that
will totally stun your friends or spouse
– they will wonder how the hell you
made a working instrument from all
that junk!
The components
Hang onto your seat, folks; this
project is for “Serious Ratters” only.
Why? Well to make this design, you’ll
need a whole bunch of stuff but
most will be able to be picked up for
next-to-nothing at a few garage sales.
Alternatively, it’s a project to keep in
mind as you collect bits and pieces
over a period.
First up, you’ll need the video head
assembly from a VHS video cassette
recorder. The bearings have to be in
good nick, so before removing the head
from the VCR, give it a spin while listening closely. The vast majority will
Fig.1: the anemometer uses the internals of a discarded mouse to generate
a frequency output proportional to wind speed. This signal is fed into
an amplifier (salvaged from a cassette player) which feeds a charge
pump circuit made from a handfull of passive components. The resulting
voltage is displayed on the speedo.
90 Silicon Chip
Fig.2: the mouse plug pin-outs.
You can easily find the pinouts for other mouse plugs by
doing a web search for “mouse
pinouts”. In this application, we
use only +5V and ground (earth)
connections.
siliconchip.com.au
A video head salvaged from an old
VHS VCR provides the bearings,
mounts and precision shaft for the
anemometer. The complete VCR
cost just $1.00.
spin superbly – they have really good
bearings – but occasionally you’ll find
one that’s a bit gritty in its rotation. If
that’s the case, get hold of another! (We
showed you how to scrounge the video
head from a VCR in “Salvage It!” in the
December 2005 issue.) The VCR used
here cost just $1 from a garage sale.
Second, you’ll need an old cassette
player, preferably a battery/mains portable. It doesn’t matter if it’s stereo or
mono but go for a small design that uses
an amplified speaker. We picked one
up for $3 – knocked down from the $5
being requested at a garage sale.
Third, you’ll need an old computer
mouse of the sort that uses a ball.
We already had one stuffed away
in a drawer, so that part didn’t cost
anything.
And last of all, you’ll need an
electronic speedo or tacho from a car.
Alternatively, if you can’t lay your
hands on one of those, you can use a
VU meter from an old cassette deck
(see “Salvage It!” in the June 2005 issue
for more on using salvaged VU meters).
The speedo used here was bought at a
local metal recycler for $3 – in fact, to
be accurate, that price also included
the tacho and the vacuum fluorescent
fuel and temperature gauges!
They’re the major bits but in addition you’ll need some discrete electronic components – some of which
almost certainly can be ratted from
the VCR. You also need three kitchen
measuring spoons, a short length of
90mm plastic pipe and some 90mm
plastic end caps.
The design
So how do we turn all those bits and
siliconchip.com.au
In this view, you can see from top to bottom, the stainless steel measuring
spoons that form the cups, the upper section of the video head, the shaft, the
upper pipe cap, the lower section of the video head containing the bearings, the
slotted optical wheel and sensors, and the mouse circuit board.
The slotted optical sensor wheel is glued to the end of the shaft. The mouse
circuit board is mounted so that the slotted wheel interrupts the light beam
between a LED and its adjoining sensor – just as it did in the mouse. Only one of
the two mouse sensors is used (the unused one can be seen at bottom left).
pieces into an anemometer? In summary, the video head provides the lowfriction ball bearings, hardened steel
shaft and aluminium bearing housing.
The measuring spoons – they’re usually stainless steel – are used to form
the anemometer cups (they mount on
one end of the shaft). The computer
mouse donates the LED/phototransistor pair and also the finely slotted
wheel that interrupts the light beam as
it spins. These components are used
March 2006 91
Fig.3: this diagram shows how the mouse, cassette player and charge pump are interconnected. Note that only the
existing external connections to the cassette player PC board are used – you don’t need to probe into its internal
circuitry. The regulated power supply is optional – in most cases, the original cassette player power supply can be
used without modification.
to generate a frequency that varies in
proportion to wind speed.
The signal from the mouse is then
amplified by the cassette player and
fed into a charge pump circuit that
comprises just a handful of passive
components. This circuit converts the
frequency into a voltage which is then
read on the car speedo (or VU meter).
If you use a speedo, you’ll be able to
read the wind speed directly in km/h
from the dial.
By altering the charge pump capacitors, a variety of meters can be
catered for.
To make it all happen you don’t
need to get deeply into the intricacies
of the circuits of the mouse, cassette
deck or speedo – provided you have a
frequency reading multimeter, it’s all
pretty straightforward.
Main Features
•
•
Large analog display
•
Works down to very low wind
speeds
•
•
•
Linear or non-linear scales
Span can be set to suit local wind
conditions
Makes use of junked equipment
Very cheap to make
92 Silicon Chip
Fig.1 shows a block diagram of how
the anemometer works.
Building it
The key to making the anemometer
is to build it in the right sequence of
steps – that way, you can test each part
of the system as you go along.
THE OPTICAL SENSOR: the mouse is used
to provide the optical sensor of the anemometer. The PC board in the mouse
remains intact – we just tap into it to
extract the signal. The first step is to
power up the mouse and then find the
signal output, which is taken directly
from the photosensor.
Fig.2 shows the pin-outs of the plugs
used on PS2 mice. In this application,
we need to use only the power supply
and ground connections.
Open up the mouse, cut off the cable and then use Fig.2 to identify the
power and earth leads. Apply 5V to
these leads (the voltage doesn’t have to
be absolutely precise, so four partially
flat 1.5V cells are fine, or you can use
an adjustable bench power supply)
and then use a frequency measuring
multimeter to probe the pins of one of
the two internal light receptors (positive probe of the meter to the device
and negative to the ground wire). Alternatively, you can probe the pins of
the IC to find the same signal.
Now spin the small slotted wheels
by hand and keep probing until you
find a pin that has an output frequency
that increases with the speed of one of
the wheels. In the prototype, this varied from about 40-2000Hz. Of course,
if you have one, a scope is ideal for this
sort of pin finding. Carefully solder a
wire to this signal pin.
The output of the sensor is likely
to be a varying DC signal. In fact, you
don’t even need a frequency measuring multimeter to check this – just
use your trusty old analog multimeter
switched to a low DC voltage range. At
low frequency outputs, the needle will
flicker faster or slower, depending on
the speed of the wheel.
To block this DC component of the
signal, wire a 470nF (0.47mF) capacitor
in series with the output – this converts the signal to an AC waveform.
You now have an optical sensor with
a high-resolution frequency output!
THE AMPLIFIER: the cassette deck is used
to amplify the small signals coming
from the optical sensor. To achieve
this, the signal output from the optical
sensor is connected to the tape head
input of the cassette player.
Access the cable that goes to the tape
head. In most cheap cassette players,
this will comprise just a single signal
wire inside the shield. Connect this
signal wire to the signal output of the
optical sensor, then connect the shield
of head input wire to the ground wire
of the mouse.
siliconchip.com.au
The mouse circuit board
is held in position by
a bracket formed from
scrap aluminium sheet.
Note that heatshrink has
been used as an insulator
between the board tracks
and the bracket.
Now power up both the cassette
player and the mouse, set the cassette
player volume to full and press the
“play” button. When you spin the optical wheel in the mouse, you should
hear a noise from the cassette player’s
speaker that changes in pitch with
wheel speed (in fact, if all you want is
an audible wind speed indicator, you
can pretty well stop right now – the
wiring part of the project finished!).
If you have difficulty finding the
right wires from the head (perhaps
because there are four wires or multiple heads), touch the different head
connections with a finger while the
tape player is running. Touching the
correct signal wire will result in a loud
hum in the speaker. (If you are using
a mains-powered cassette player, you
should take care that you cannot come
into contact with high voltages. In this
case, it is best to extend the head wires
outside of the case and then temporarily close it up again.)
Because the amplifier has very high
gain, it is susceptible to picking up
noise. To reduce this, a 1kW pot is
wired across the mouse output, with
the wiper connecting to the amplifier.
In use, this pot is adjusted so that
adequate signal is provided without
there being too much noise present
(indicated by lots of noise in the
speaker even with no rotation of the
wheel). This wiring – and in fact the
complete circuit of the anemometer –
is shown in Fig.3.
You now have a high gain amplifier
suitable for amplifying the output of
the optical sensor!
siliconchip.com.au
THE SPINNING ASSEMBLY: disassemble
the video head, gutting it of any electronics that you see. Pulling the head
apart usually requires a Phillips head
screwdriver and a small metric Allen
key. Some brass collars are also a light
press-fit on the shaft – these can be
removed by gently using a hammer and
a punch. Prise out the black magnetic
material that is within the head. It easily shatters, so be careful when doing
this – it’s best to wear safety glasses
when performing this operation.
Once you’ve got the head bare, you
can build the impeller. We used three
small (1 teaspoon or 20ml) measuring
spoons from a supermarket. These
particular ones were made of stainless steel with a non-slip (and noncorrosive!) coating.
The spoons were bolted together,
using the existing holes located at
one end of the handles. The spoons
were then spread evenly (ie, with a
120° spacing) and matching holes
were drilled through the spinning
aluminium housing and the handles
of the spoons. The spoons were then
bolted in place using short screws and
nuts and once this was done, the heads
of the spoons were carefully twisted
through 90° to form the anemometer
cup assembly.
The completed assembly should
spin freely in even the lightest puff
of wind. If the assembly is out of
balance, hold the shaft horizontally
and see which cups always points
downwards. Place a small weight on
the side opposite. Getting the assembly well-balanced yields dividends
The spinning disc has lots of slots
in it – we counted 40 but that might
not be right! In any case, the output
resolution of the sensor is very good
– if you wish, you can calibrate the
scale to read wind speeds of just a few
kilometres per hour.
Stainless steel measuring spoons were
used to form the anemometer cups.
These were bought (gasp!) new for the
project.
in longevity – an out-of-balance shaft
puts a greater load on the bearings.
The half of the video head that
contains the bearings is bolted to the
inside of a 90mm PVC pipe cap. As
with the rotating part of the head,
some new holes will probably need
to be drilled through the aluminium
for the mounting bolts.
The next step is to fit the slotted
mouse wheel to the opposite end of
the shaft to the cups. Cut the slotted
encoder wheel off its plastic shaft and
then use a fine flat file to smooth each
side, being careful not to burr the tiny
March 2006 93
The rotating assembly
can be balanced by
adding weights – here a
bolt and some extra nuts
(arrowed) have been
placed on one side of the
assembly.
The speedo was mounted in a small picture frame. Note
that it is easy to backlight the dial – all car speedos have
this facility and in some, even the needle is illuminated!
slots. Then, using instant adhesive,
very carefully glue the slotted wheel
to the end of the anemometer shaft. It
needs to be perfectly concentric; ie,
when the shaft is turning there is no
run-out.
The mouse PC board is mounted so
that the slotted wheel spins between
the LED and its adjacent photosensor.
We used a small piece of scrap aluminium to make the locating bracket.
Note that if the shaft has a tendency
to slide downwards through the bearings, so causing clearance problems
between the slotted wheel and its
sensor, place a drop of instant glue on
the shaft right next to a bearing before
A Fun Instrument
If you want a fun instrument rather than
a calibrated km/h design, simply pick
capacitors in the charge pump that
give full-scale deflection of the speedo
when the cups are quickly flicked. Then
use a computer, scanner and printer to
make a scale that shows wind speeds
like “Boring”, “Some Excitement”,
“Hell It’s Blowing”, “Where’s The Cat
Gone?”, “Take Shelter!” and “Are We
Still Alive?”.
sliding it through the bearing to the
correct position.
You now have a very sensitive and
durable spinning anemometer head
with a variable frequency output!
THE DISPLAY: the display can comprise
an electronic car speedo or tachometer,
or a cassette deck VU meter.
The car instruments make for a
much more impressive readout, so
we’ve used one of those. In any case,
we don’t need the frequency-to-voltage converter that’s used within these
car instruments; instead, as mentioned
above, we make our own charge pump
system. Doing this means that we can
match the amplified output of the
optical sensor to a very wide range
of meters, as well as easily changing
characteristics like smoothing and
range.
Remove the speedo or tacho and
strip it down until just the meter and
its electric movement remain. When
a low voltage (eg, 2V) is applied, the
meter should swing full scale. Take
note of the positive and negative leads,
as revealed by this test.
If you’re using a speedo, you should
be able to retain the standard km/h
scale. Alternatively, if you use a tacho
or you want the scaling to be different
Fig.4: this charge
pump circuit is
used to convert
the amplified
frequency signal
from the mouse
to a DC signal
proportional to
the wind speed.
94 Silicon Chip
to the original on the speedo, a new
scale will need to be made using a
scanner, PC and printer (see “Salvage
It!” in the March 2005 issue for more
on rescaling car tachos). In this case,
the positions of the increments will
be found during the calibration procedure (see below).
You now have a large analog anemometer readout!
POWER SUPPLIES: two voltages need to
be provided: 5V to the mouse circuit
and (usually) 6V to the cassette player
(we now know these components as
the optical sensor and amplifier, respectively!).
If absolute accuracy in the wind
speed readout isn’t required, the amplifier can be powered directly by the
mains, batteries or a plugpack – whatever was originally used by the cassette
player. The down-side of this approach
is that the displayed wind speed will
vary with supply voltage fluctuations.
This is because the square-wave amplified output is driven from rail to
rail – the cassette player is no longer
acting as a feedback amplifier.
The alternative is to use a voltage
regulator, which is what we chose
to do. As well as providing better
instrument accuracy, this also allows
easy calibration in a car as the system
can be powered from the car supply.
We powered the regulator from a
spare plugpack we had previously
salvaged.
The supply for the optical sensor is
obtained by simply using a 10kW pot
across the power feed that originally
went to the cassette player motor, adjusted to provide 5V when loaded by
the optical sensor.
Fig.3 shows the power supply wirsiliconchip.com.au
➊
➎
➋
➌
ing, both for the amplifier and the
optical sensor.
FREQUENCY-TO-VOLTAGE CONVERTER: the
frequency-to-voltage converter (charge
pump) is the final stage in the build
and is best optimised on the bench
with the whole system working.
Fig.4 shows the way in which the
charge pump works. For the moment,
disregard the variable resistor VR1.
Initially, C1 and C2 are discharged.
When the input voltage goes high, C1
starts to charge through D1 and C2.
Because C1 is much smaller than C2,
C1 fully charges earlier than C2 and
when this occurs, current stops flowing. However, during this process, C2
has received a small charge increase.
When the input voltage goes low,
C1 discharges through D2, but C2
does not discharge because D1 blocks
the discharge path. The result is that
each time the input voltage goes high,
a small amount of charge is added to
C2, resulting in C2’s voltage rising in
proportion with the input frequency.
C2 powers the meter; ie, C2 is being
constantly discharged by the meter’s
load. VR1 allows adjustment of the
meter’s deflection for a given voltage
level across C2.
C2 should be kept as low as possible
but must be sufficient to provide a
damped meter movement at the lowest
frequency output at which the amplifier will work. If the speedo needle
flickers when the cups are turned at
the slowest speed at which you will be
making measurements (this value desiliconchip.com.au
➍
pends on the scale you have chosen),
then C2 needs to be increased until
the needle moves smoothly.
C1 needs to be small enough to allow
it to fully discharge during the time
that the input signal is low. In the prototype, where the car speedometer has
a 100W resistance, C1 comprises two
10mF electrolytic capacitors (wired
negative to negative to make the pair
non-polarised), while C2 has a value of
220mF. The 16W resistor in series with
C1 reduces the peak current through
the amplifier.
Note that if you are using a VU meter
instead of a car speedo, C1, C2 and the
resistor in series with C1 will all be
much lower in value.
It all starts to sound a bit compli-
The basic layout: (1)
cassette player circuit
board, being used as
an amplifier; (2) pot
that provides the 5V
supply to the mouse
board; (3) voltage
regulator and associated
capacitors powering
the amplifier (only
required if there will
be major mains supply
variations); (4) amplifier
input attenuating pot
and capacitor; (5)
charge pump circuit.
Incidentally, the
expensive looking NEC
pots were bought very
cheaply on eBay.
cated but when you realise that the
frequency-to-voltage charge pump
circuit uses only six low-cost components, you can breathe easily again!
To find the best values for C1 and
C2, initially lash up the anemometer
circuit on the bench – power supplies
and all. Start with the capacitor values
cited above and set VR1 so that its resistance is as low as possible. Spin the
anemometer cups by hand – rotating
them fairly slowly – and check that the
speedo (or VU meter needle) smoothly
deflects a little.
Now spin the cups faster and check
that the deflection is greater. Adjust
VR1 and check that the deflection for
a given cup speed is reduced.
If the deflection is too small, in-
This is the cassette player that donated its amplifier. In many cases it will
be easiest to use the original cassette player power supply and mount the
new components inside. The garage sale purchase price was knocked
down form the marked $5 to $3.
March 2006 95
player PC board to mount it in a new
box, keep in mind that you must bridge
the switch that is normally activated
when the “Play” button is pressed –
otherwise, the amplifier won’t work.
The display is easily mounted remote
to the main box, so if retaining the cassette player housing, it’s easy to tuck
it out of sight.
While it might appear that the distance between the head and amplifier
should be kept very short, we had no
difficulties in stretching this distance
to 25 metres, using salvaged multi-core
alarm cable.
Calibration
The working anemometer, seen positioned high on the roof. The cup covering
the centre section of the rotating assembly was made from an aerosol cap. For
improved durability, everything you see here should be painted.
crease the value of C1. If the needle
deflection becomes non-linear at high
speed (ie, its deflection is much less
than expected), reduce the value of C1
and then reduce C2 proportionately.
In short, just play around with the
capacitor values (always keeping C1
much lower than C2) until the needle
behaves as wanted over a variety of
cup speeds.
Note that as a set-up guide, a fast
flick of the anemometer cups will
spin them to a wind speed of about
40km/h. If you only want to measure
wind speeds up to 50 km/h, size the
capacitors so that you get nearly full
scale deflection with a fast whiz of
the cups.
Final assembly
The rotating assembly is completed
by adding the short section of 90mm
plastic pipe and the second end-cap.
Use PVC pipe adhesive to glue these
parts together.
Alternatively, if you want to be able
to easily disassemble the container,
use self-tapping screws to hold one of
the end-caps in place. Make sure you
seal the hole where the cable exits.
Note that the anemometer is orientated
so that its rotating cups are below the
plastic housing – this helps prevent
the ingress of water. The prototype
was mounted using square aluminium
tube. This tube was bolted to the upper end cap.
We mounted the electronics in a
new box. The cassette player PC board
was removed from its original case.
However, especially if you are going to
use the cassette player’s power supply,
we suggest that you leave everything
inside the cassette player, placing the
charge pump and other minor components inside. If you remove the cassette
Rat It Before You Chuck It!
Whenever you throw away an old TV (or
VCR or washing machine or dishwasher
or printer) do you always think that surely
there must be some good salvageable
components inside? Well, this column is
for you! (And it’s also for people without a
lot of dough.) Each month we’ll use bits
and pieces sourced from discards, sometimes in mini-projects and other times as
an ideas smorgasbord.
And you can contribute as well. If you
have a use for specific parts which can
96 Silicon Chip
easily be salvaged from goods commonly
being thrown away, we’d love to hear from
you. Perhaps you use the pressure switch
from a washing machine to control a pump.
Or maybe you have a use for the highquality bearings from VCR heads. Or
perhaps you’ve found how the guts of a
cassette player can be easily turned into
a metal detector. (Well, we made the last
one up but you get the idea . . .)
If you have some practical ideas, write
in and tell us!
Calibration is easily achieved by
placing the whole device in a moving
car, locating the rotating assembly
outside, and then calibrating against
the speedo reading. Just make sure
that you do the calibration on a still
day! The device can be powered by
the car supply or the cassette player’s
internal batteries.
If you are using a preformed, linear
scale, setting the correct needle position with VR1 should be done at a
couple of speeds. Note that because
of non-linearities in the anemometer
aerodynamics, amplifier and meter,
you won’t get a perfectly accurate
readout at all wind speeds – but you
should be within 10% everywhere.
If you are devising your own scale,
start with one with linear markings
(eg, 1-10) on the scale. Write down the
wind speed at each of the markings
are then print out a revised scale with
these speeds in the correct positions.
Incidentally, if you want to decrease
the sensitivity to high wind speeds (ie,
expand the lower wind speed scale),
tweaking the value of C1 upwards will
do this for you!
Conclusion
This is a fun and engrossing project
to make – from disassembling the
mouse and video head, to trying different charge pump capacitor values
to give you the scale and sensitivity
that you want. The anemometer is
sufficiently sensitive to spin with
wind speeds of just 2-3km/h (and has
an output resolution to measure those
speeds too!) and if well balanced, is
still rugged enough to cope with high
speeds and full weather exposure.
Best of all, it makes use of a heap
of stuff you’d otherwise just throw
SC
away!
siliconchip.com.au
|