This is only a preview of the October 1997 issue of Silicon Chip. You can view 28 of the 96 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 A 5-Digit Tachometer":
Articles in this series:
Items relevant to "PC Controlled 6-Channel Voltmeter":
Items relevant to "The Flickering Flame For Stage Work":
Items relevant to "Building The 500W Audio Power Amplifier; Pt.3":
Articles in this series:
Purchase a printed copy of this issue for $10.00. |
By ROSS TESTER
T
This little
lighting gimmick
was used at a recent school
eisteddfod. It uses a 12V 20W
halogen lamp/reflector fitting
mounted in a plastic drink bottle
filled with orange/red cellophane.
It gives a convincing imitation of
fire, hence the name “Flickering
Flame”. Why not set your next
performance on fire?
60 Silicon Chip
HEATRICAL PRODUCTIONS often call for flaming
torches and similar lighting effects. The problem is
that for fairly obvious safety reasons most theatres
and halls have very strict rules regarding the use of naked flames on stage. In fact, in most halls, naked flames
are taboo.
We had just this situation recently when our local high
school needed some props for its Rock Eisteddfod act.
One scene was a castle wall lit by flaming torches – except they couldn’t have flames! Also they (and the Rock
Eisteddfod itself) had a couple of other curly requirements
which made life just that much harder!
For example, mains power was not an option because
the props on which the flaming torches were placed
were to be moved to various positions on the stage. And
the lights themselves needed to be completely portable
because the props were stored offstage and assembled
in a rush!
One other request was that they had to cost as close
to nothing as possible and be reasonably sized – not too
big to get in the way but conversely not too small either.
And they had to look realistic!
For those not familiar with the Rock Eisteddfod, perhaps a word or two of explanation is in order. The challenge is open to all secondary schools and entails an act
of dance and drama set to contemporary recorded music.
The act itself can be no more than eight minutes and
most importantly, any set or props used must be capable of being brought onto stage and set in less than four
minutes (and conversely, removed in the same time).
There are strict limits on the amount of money a school
can spend and there are also limits on the number of
LEFT: it mightn’t look too spectacular up close and as a
still photo but from an audience viewpoint it looks just
like the real thing when it is working.
backstage crew allowed to assist the
“performers”.
Those who have never seen a Rock
Eisteddfod performance before marvel at the spectacle, the professionalism, the choreography, the costumes
and the props.
Ah, the props. This is where we
came in, with a request to help out
with those flaming torches (no pun
intended!). The props people at the
school had come up with the basic
design for the torch (and as you will
see, it’s amazing what you can get
away with from a distance!). What
they wanted was something to make
them flicker.
“Easy”, we thought. First of all we
looked at a real flickering flame (a
candle, to be precise). Effectively, the
“light” was on all the time but every
now and then it dimmed a bit as the
flame was caught by a breeze. All we
needed to do was emulate that. A
microprocessor could easily be programmed to do the job nicely.
“Whooaa! Too expensive”, they
said. OK then, how about a 555 timer
configured as an astable multivibrator
driving a cheap power Mosfet? You
couldn’t get much cheaper than that!
The problem with that idea was
that while it certainly flickered the
lights, it was far too regular: looked
more like a flamin’ lighthouse than a
flaming torch!
What about two 555 timers running
at different, unrelated frequencies?
Would this give the random effect
we wanted? We tried this idea and . .
. sure would! Calculating values gave
us roughly the right oscillation rates,
trial and error gave us the effect we
wanted. Fig.1 shows the final circuit.
The duty cycle (on time to off time)
is set by the ratios of R1 to R2 and the
oscillation rate is set by R1 and R2 in
conjunction with C1. The duty cycles
were set very high – around 10:1. Any
longer than this and the lights actually
went right off – not very realistic at all!
The oscillation rate was set at about
1Hz or longer.
The values of R1 & R2 are significantly higher than one might “normally” expect to be used in 555 circuits. The reason for this is that high
values of resistance allow low values
of capacitance. A high value resistor
costs the same as a low value resistor
while a high value capacitor costs
significantly more than a low value
unit. We used the same values for R1
The 20W halogen lamp can be directly soldered to the PC board as shown here,
or connected via flying leads. The component at top right of the PC board is a
pair of header pins with a shorting link – this formed our on/off switch.
Fig.1: the circuit employs a 556 dual timer to drive a Mosfet. The two oscillators
run at different rates to give a random flickering effect from the halogen lamp.
& R2 in the two oscillator circuits but
picked different capacitor values to
ensure that the operating frequencies
were not too close together.
While on the subject of costs, we
looked at the lolly shop catalogs and
found that a 556 (two 555s in one
package) was a few cents cheaper than
a pair of 555s, so we went this route.
The outputs of both astable oscillators are fed to a diode “OR” gate and
these feed the gate of the Mosfet. In
effect, the difference between the two
oscillators is fed to the Mosfet gate.
What happens is that when either of
the oscillator outputs goes low, one of
the diodes is forward biased, taking
the Mosfet’s gate low.
When the gate is taken high, which
will be most of the time, the Mosfet
is turned hard on and is a very low
resistance. Therefore, the lamp lights
at full brilliance.
When the Mosfet gate is pulled low
by either of the oscillator outputs, the
Mosfet turns off, turning off the lamp.
But because of the very short “off”
time and the thermal inertia of the
lamp filament, it doesn’t actually turn
off but flickers. So we have random
flickering of the lamp, which is just
what we want.
Viewed up close, it doesn’t look all
that impressive. From more than a few
metres away though, the effect is quite
convincing – there’s fire in that thing!
In the end, we made quite of few of
these torches, varying the timing capacitors in each to ensure they never
flickered “in sync”.
Making the torch
You would be surprised at how
much you can get away with in
designing props! Looking at the
photograph, you’ll see that our torch
appears exactly what it is: crumpled
cellophane in a plastic drink bottle,
fitted to a length of cardboard tube.
But to the audience, it looks just like
October 1997 61
Fig.2: the halogen lamp was soldered directly to the PC board
but it would be easier if you used a standard halogen lamp
socket base.
a flaming torch!
We cut the bottom off a PET 1.25l
Coke bottle with a sharp serrated
knife. Did we forget to mention that
we drank the contents first? Next, we
cut some thin strips, about 15-20mm
wide, of red cellophane and laid these
down the inside of the bottle. About
five or six strips seemed to work best
but you can experiment for the desired
effect. Then we crumpled a sheet of
orange cellophane and placed this
inside the red strips. Presto, a torch!
By the way, PET stands for Polyethylene Terephthalate, which is the
long-winded moniker for polyester.
Now you know.
The other end of the bottle was
removed to suit the lamps used. We
mentioned safety before as this was
a major concern. The lamp we used
was especially chosen for the job: a
12V 50mm dia
meter halogen type
with dichroic reflector, normally used
in low voltage downlights or shop
display lights.
However, we used a specific type.
Most 12V downlights have a 50 watt
rating; we used a 20 watt type to keep
heat to a minimum. And to ensure
that the hot halogen lamp would
not be able to ignite the cellophane
(we’ve seen that happen before!) we
chose a lamp with an integral clear
glass cover.
Should these lamps not be available
at your normal shop Jaycar stores have
them (Cat No SL-2732). They are coded “BAB” on the reflector, indicating
that they have a beam width of 38
degrees which is pretty well optimum
for this job.
Their price is right, too at $4.95
each (we were quoted $15 each at a
62 Silicon Chip
Fig.4: this is the full-size etching pattern
for the PC board.
lighting store!) Incidentally, Jaycar
have a similar 12V/20W lamp just
35mm in diameter if space is a problem. They also have bases to suit these
lamps but at $2.95 each we decided
to forego these to keep cost to a minimum and solder directly to the lamp
pins. In retrospect, that may not have
been such a brilliant idea. The reasons
will follow shortly . . .
The lamps were fitted to the neck
end of the bottle, shining upwards
through the cellophane. To do this, we
cut the neck off, leaving an opening
about 25mm in diameter. What you
want to achieve is an opening not
too large for the lamp to slip through.
We used contact adhesive to hold
the lamp in place. Yes, the lamp does
get rather hot; in fact, enough to distort the PET bottle but we were able
to operate our torches for half an hour
or more without any problems.
Allow the adhesive to dry completely. While this is happening, you
can fashion the mounting hardware.
We used 100mm cardboard tube long
enough to take the battery pack (see
below) and the lamp itself. Eight slots
were cut in the tube, about 40mm
down from the top, which allowed the
lamp to be simply pushed into place.
It is quite important that the tube be
made deep enough to ensure that no
white light can be seen by the audience – this ruins the effect completely.
We said before that our idea was to
solder the pins of the lamp directly
to the PC board. This is not quite as
simple as it seems the pins simply
did not want to solder! They’re probably nickel plated or similar which
is certainly not designed for ease of
soldering.
Eventually, after much scraping of
the pins and with a very hot iron we
were able to make a soldered joint but
this was definitely the weakest link in
the chain. If you can afford to invest
another couple of dollars in the bases,
they would make life much easier.
The circuit is designed to operate
from 12V DC. The electronics drain
is negligible but the lamp itself draws
the best part of 2A (ie, 20W/12V). We
used a battery pack made of six 2.5A
SLA rechargeable cells (mainly ‘cos
the school had these on hand), giving
an operating time of more than an
hour. As the whole Rock Eisteddfod
performance was over in eight minutes, this was more than enough. If
you need longer times, you will need
a suitably larger battery.
Assembling the PC board
This view shows the 20W halogen
lamp and its integral reflector.
We designed a PC board to suit
the project. It is coded 11410971
and measures 58 x 38mm. This is
quite straightforward to assemble as
the only polarised parts are the IC,
Parts List
1 PC board, code 11410971, 58
x 38mm
2 PC Stakes
1 SPST switch (optional)
1 12V 20W sealed halogen
reflector lamp (see text)
12V battery capable of supplying
2A
Semiconductors
1 556 dual timer
1 MTP3055E power Mosfet
2 1N914 diodes
Fig.3: this diagram shows the overall scheme. The battery and PC board/lamp
assembly is mounted in a cardboard tube with the lamp illuminating a PET soft
drink bottle filled with crumpled cellophane.
Mosfet and diodes. Make sure you
get these right, along with the battery
connections
Therefore, construction is very simple – apart from soldering the lamp, as
mentioned above. We made provision
for an on/off switch on the board or
you could simply connect and disconnect the battery as required.
Before you solder in the Mosfet
and the lamp, you can check the operation of your circuit with an analog
multi-meter. Apply 12V and check the
voltage at pins 5 & 9 of the IC using a
multimeter (set to measure 12V DC).
At both pins, you should see the meter
dipping regularly – but not down to
zero volts, unless you have chosen
a much higher value resistor for R2.
When you measure at the junction
of the two diode anodes, you should
see the meter dipping more or less
randomly, indicat
ing that the two
outputs are being added, or rather
Capacitors
1 0.1µF MKT polyester or ceramic
1 .039µF MKT polyester or
ceramic
2 .01µF MKT polyester or ceramic
Resistors (0.25W, 1%)
2 10MΩ
1 100kΩ
2 1MΩ
subtracted, through the diodes.
With the 20W lamp suggested, a
heatsink for the FET is not really
necessary. It does get reasonably warm
to touch but should be quite happy at
this level.
Do not use a 50W lamp. There is
no doubt that it will melt the PET
bottle and it could well set fire to the
SC
cellophane.
October 1997 63
|