This is only a preview of the May 2001 issue of Silicon Chip. You can view 33 of the 104 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 "Using Linux To Share An Internet Connection; Pt.1":
Articles in this series:
Items relevant to "Powerful 12V Mini Stereo Amplifier":
Items relevant to "Two White-LED Torches To Build":
Items relevant to "PowerPak: A Multi-Voltage Power Supply":
Purchase a printed copy of this issue for $10.00. |
NOT 1
WHITE LED TORC
TOR
Both of these LED torches have considerably more light output
than our original design in the December 2000 issue. They use
more LEDs and they run from a single AA or D cell which will
have a long life. They make very good torches for camping,
walking at night or for emergency work on your car.
Design by JOHN CLARKE
T
hese LED torches produce a
beautiful even spread of white
light which is quite different
from that of conventional torches
using Krypton bulbs.
Conventional torches tend to produce a “hot spot” that can penetrate
the darkness for some distance and
they have a larger cone of much less
intense illumination. Overall, they
tend to have quite a narrow beam and
you have to move the torch around a
lot to adequately light up the area in
front of you.
By contrast, these LED torches have
a much wider diffuse beam, giving a
very even spread of light without a
central hot spot. For most of the time,
this more diffuse beam is much easier
on the eyes and the colour of objects
is much more natural. In fact, it is like
carrying a source of daylight around.
So these LED torches are ideal for
bushwalking (at night!), even in very
58 Silicon Chip
rough terrain, for illumination inside
a tent or over a picnic table and as
noted above, for emergency work on
your car if, perish the thought, you
break down at night.
Constant brightness
Another big advantage of these LED
torches is their constant brightness,
regardless of battery voltage.
Conventional torches start out with
high brightness when the batteries
are fresh but they soon dull down as
the cells discharge. By the time the
cells are down to 1V, the light output
is woeful.
These LED torches have the same
light output even if the cell voltage
goes below 1V. And they can also run
with Nicad and NiMH cells which
give a nominal 1.2V. Conventional
torches are hopeless with 1.2V cells,
unless they have been specifically
designed to run from rechargeables.
Not only that, torch bulbs have a
notoriously short life and they can
fail at the most inopportune moments.
In fact, any time a torch bulb fails is
inconvenient, by definition. After all,
if a torch bulb failed when it was convenient, you probably don’t need it.
Once you change over to a LED
torch, you will never need to change
a LED – they last a life-time (well,
practically).
Two versions
We are describing two versions of
this LED torch, both of which use the
same basic circuit. One version uses
three white LEDs and runs from a
single AA cell in a 2-cell torch. The
second version uses six white LEDs
and runs from one or two D cells and
can fit in a 2-cell or 3-cell torch.
These torches use far less current
than a conventional Krypton bulb
torch. A twin D-cell torch bulb nor-
BUT 2
CHES TO BUILD!
RCHES
Features
mally pulls about 0.8A at 3V, dropping
to around 0.7A at 2.4V. In power
terms, this is 2.4W at 3V, dropping to
1.68W at 2.4V – this is why conventional torches are so dull when the
cells aren’t fresh.
By comparison, our D cell 6-LED
version of the torch pulls only 480mA
at 1.5V, rising to 650mA at 1V. This is
less than one third of the power drain
of the conventional torch.
Again, in a conventional twin AA
cell torch, the Krypton bulb pulls
about 0.47A at 2.2V or just over 1W.
Super soft white light
Constant brightness over cell life
Indefinite lamp life
Extended cell life
Ideal for use with Nicad & NiMH cells
D cell version has brightness control
Our single AA cell 3-LED torch pulls
210mA at 1.5V, rising to about 360mA
at 1V. Again this is one third of the
power drain of the equivalent conventional torch.
Circuit details
As with our original white LED
torch described in the December 2000
issue, both these torches are based on
a DC-DC converter. The DC-DC converter for the AA-cell torch is about
the same size as an AA cell, while the
converter for the D-cell torch is about
the same size as a D cell. The larger
D-cell converter includes a brightness
control and can drive six white LEDs
instead of three.
Fig.1 shows the D-cell torch while
Fig.2 shows the AA-cell version.
Both use a Maxim MAX1676 high
efficiency step-up DC-DC converter
and an inductor to provide the power
conversion. The Maxim MAX1676
was originally intended for use in
mobile phones, as a single cell voltage
booster, so it is ideal for this torch
application.
An “exploded” view of the “D” torch which has a DC-DC converter capable of driving six ultrabright white LEDs from a
single C or D cell. The white cylinder insulates the PC board assembly from any metal fittings in the torch.
MAY 2001 59
Top trace is the inductor waveform at pin 9 of IC1 for the A-cell version. Its
frequency is 176kHz. The period when the voltage is low charges the inductor
and the high level is when the charge is transferred to the output. The lower
trace is the output voltage. It is 3.96V and has a 160mV of ripple.
The block diagram of Fig.3 shows
the internal schematic of the MAX1676 and the external components
needed, including the key component
– L1, a 22µH inductor. The internal
Mosfets, Q1 & Q2, do all the high
speed switching work.
Circuit operation is as follows:
current flows through inductor L1 and
Mosfet Q1. When the current builds
up to 1A, Q1 turns off and Q2 turns
on. The charge in inductor L1 is then
transferred via Q2 to capacitor C1 and
the load.
The voltage at Vout is fed back to
the MAX1676 via a resistive divider
comprising R5 & R6.
The internal control circuit derives
its power from the Vout terminal and
so when power is first applied to the
circuit, current flows through L1 and
Q2 to power the control circuit.
Q2 is a P-channel Mosfet which
requires at least 1V at the power
source in order to be closed and
pass the voltage back to the control
circuit.
For lower voltages it is necessary to
include an external Schottky diode in
parallel with Q2 to allow current to
flow to the control circuit.
Once the circuit starts up, it is
powered from the Vout supply and
Q2 then performs its task of switching
the charge from L1 to the load with
minimal voltage loss and the diode is
effectively out of circuit.
60 Silicon Chip
As shown in Fig.1 & Fig.2, the two
circuits are very similar.
Let’s have a look at Fig.2, the AA
cell version. It has a fixed resistive
divider for the voltage feedback at pin
1. Inductor L1 must have very low DC
resistance to ensure high efficiency of
the circuit.
As mentioned above, the inductor
is charged until the current through
it reaches 1A. The inductor must not
saturate at 1A and also it must have a
low enough resistance to ensure that
the current actually rises to 1A. The
step-up circuit will not operate if the
1A limit is not reached.
Thus we have used an inductor
which has a DC resistance of 0.2Ω.
Standard commercially wound inductors with wire resistances of more than
0.5Ω will not let the circuit operate.
The output supply rail is close to
4V, as set by the 100kΩ and 47kΩ divider resistors and it is bypassed with
a 47µF tantalum capacitor. Each LED
is powered separately using a 27Ω
current limiting resistor to ensure
equal current sharing.
The nominal LED forward voltage
is about 3.5V and so the calculated
current through each LED is (4V 3.5V)/27Ω = 18.5mA. In practice, the
LED current is slightly higher than
this.
By the way, the AA-cell version
could be powered with a C cell, if
built into a C-cell torch.
D-cell version
The D-cell version uses an inductor
which has a lower resistance again
than in the AA-cell version and it uses
a larger core. The value of inductance
is the same at 22µH but the lower resistance ensures higher efficiency for
step-up conversion. This circuit can
Top trace is the inductor waveform at pin 9 of IC1 for the D-cell version. The
glitches are a reset that automatically occurs within the IC to ensure operation
at low loads. Frequency of operation is 133kHz and the low output is when the
inductor is charging. The energy is transferred to the load when the waveform is
high. Lower trace is the output voltage at 4.07V with a 350mV ripple.
Fig.1 (above left) and Fig.2 (above right) show the “D” cell and “AA” cell variants
respectively. Both are based on the MAX1676 IC high-efficiency DC-DC converter,
a chip originally designed for use in mobile phones.
MAY 2001 61
Fig.3: inside the MAX1676 DC-DC converter. Its operation is fully described in the text.
drive up to six white LEDs.
There is also a trimpot, VR1, to
adjust the output voltage so that the
LED brightness can be varied from
almost zero to maximum brilliance. A
200Ω resistor at pin 7, in conjunction
with internal Mosfet Q3, provides
damping for the inductor when it is
released from charging. This damps
oscillations and ringing which can
otherwise cause electromagnetic interference (EMI).
The Schottky diode D1 is not required if the circuit is powered with
two D cells.
Both circuits include reverse polarity protection, by virtue of diode D2,
which conducts if the battery is inserted incorrectly. Diode D2 provides
only short-term protection since the
current flow will be high.
You should check the battery polar-
ity immediately if the torch is found
not to work.
Construction
Construction of these LED torches
will require patience, good eyesight, a
magnifying glass and some experience
with soldering. Why? Because we are
using a surface-mount IC for IC1.
The IC is soldered onto a u10MAX
carrier PC board for the D-cell version
(Fig.4). but solders directly to the PC
board of the AA-cell version (Fig.5).
Regardless of which version you
build, soldering this IC in place will
require a modified soldering bit which
has been filed to a narrow screwdriver
shape. The idea is to solder all five
pins on each side of the IC at the one
time.
Before soldering in the IC, check the
PC boards for any shorts or breaks in
the tracks. Any problems in the surface mount area probably cannot be
fixed unless there is only a small short
between tracks which can be cleared
with a sharp knife. The PC boards
must be tinned (solder-plated) before
use so that the IC can be soldered in
without damaging the fine tracks. This
should have already been done by the
PC board manufacturer.
One method of soldering in the
IC by hand is to initially cover the
underside of the IC pins with solder
by wiping over them with a standard
chisel-shaped soldering bit which
is lightly coated with solder. Make
sure that the solder does not bridge
between the IC pins. If it does, clean
the soldering tip and wipe the excess
solder off the IC pins with the now
cleaned tip.
Check the IC with a magnifying
Fig.4: the component overlay of the “D” cell version. Note the position of the “daughter board” containing the MAX1676
SMD (surface mount device) IC. These devices can be a little tricky to solder – the text of this article should help! The
photo at right shows the complete board but it is rotated through 180° compared to the component overlay.
62 Silicon Chip
glass to be sure the IC pins are all
tinned, without any shorts between
the pins.
Then place the IC onto the PC board
and align the pin 1 indicator on the
IC (a small dot on the body) with the
pin 1 pad on the PC board. Straighten
up the IC so it sits correctly on the
IC pads. Now heat up the modified
soldering iron tip (sharp screwdriver
shape) which is untinned or cleaned
of solder with a wet sponge. Apply
the tip to the leads on one side of the
IC to solder it in place.
Check that it is still aligned onto
the IC pads correctly. If not reheat the
pins and align correctly. When one
side has been soldered in place heat
the remaining pins on the other side
of the IC to the PC board.
Now you will need to carefully
inspect the IC soldering using a magnifying glass. Check for lifted pins on
the IC and shorts between pins.
Finally, use a multimeter to check
that each pin is indeed connected to
its respective track on the PC board.
D-Cell version
The D-cell version of the LED Torch
can be assembled as shown in Fig.4.
Insert the PC stakes with the long end
going down into the PC board to give
a similar pin height above and below
the PC board.
Install the u10MAX PC board onto
the main board using short lengths of
tinned copper wire passing through
each PC board. Make sure that the
u10MAX board is oriented correctly,
with pin 1 lined up on both boards.
Insert and solder all the resistors and
capacitors, taking care with the tantalum and electrolytic types which must
be oriented with the polarity shown.
Now solder in the diodes and trim-
Here’s the 6-LED array for the “D”
cell version. The five 27Ω LED current
limiting resistors all solder to a spacer.
Parts List – D Cell Version
1 2 x D-Cell torch (Eveready E250K or similar) or a 3 x D-cell torch
1 PC board coded 11105011, 59 x 33mm (46 holes)
1 micro-DIP x 10-pin PC board coded u10MAX, 13 x 12mm (10 holes)
(must be solder plated)
1 ferrite toroid, 19 x 10 x 5mm (L1) (Jaycar LO-1230)
1 200mm length of 1mm enamelled copper wire
1 60mm length of 0.8mm tinned copper wire
1 50mm length of red hookup wire
1 50mm length of green hookup wire
1 12mm OD steel or brass washer
1 16mm OD x 10mm ID steel or brass washer
2 3mm x 70mm steel or brass threaded rod
1 M3 tapped metal spacer
1 M3 crimp solder lug
1 M3 x 10mm screw
1 M3 star washer
1 100mm long cable tie
9 PC stakes
1 plastic translucent diffuser (cylinder 23mm ID x 17mm long)
(ours was cut from a cover cap supplied with a “FRUITY FLAVORITS”
250mm drink container)
1 72 x 115mm piece of thin cardboard
Semiconductors
6 5mm 5600mcd white LEDs (LED1-6)
1 MAX1676EUB step-up DC-DC converter (IC1)
1 BYV10-20 Schottky diode (D1)
1 IN5404 3A diode (D2)
Capacitors
2 47µF tantalum capacitors
1 1µF PC electrolytic capacitor
3 0.1µF monolithic ceramic capacitors (code 104 or 100n)
Resistors
1 100kΩ
(brown black black orange brown or brown black yellow brown)
1 43kΩ
(yellow orange black red brown or yellow orange orange brown)
1 200Ω
(red black black black brown or red black brown brown)
6 27Ω
(red violet black gold brown or red violet black brown)
1 50kΩ horizontal trimpot (VR1)
pot VR1. Inductor L1 is wound using
4 turns of 1mm enamelled copper
wire around the ferrite toroidal core.
Bare the ends of the wire with some
fine emery paper or a sharp knife, to
remove the enamel insulation before
soldering to the PC stakes. Inductor
L1 is secured to the PC board using
short lengths of tinned copper wire
which wrap over the toroid in the two
positions shown.
Solder a 12mm washer to the PC
stakes at the positive end of the PC
board (lefthand side of Fig.4).
A crimp-type solder lug is attached
to the other end of the PC board. You
need to pry open the crimp end with
pliers and flatten it and then solder
the flattened section to the PC pins on
the top side of the board; the circular
lug section then hangs beneath the PC
board. Solder a short length of hookup
wire between the “A” PC board pin
and one of the eyelet PC stakes.
LED Array
All of the steps for assembling the
LED array for the D-cell torch are
shown in Fig.5. First, we have to make
the LED array.
The 6-LED array for this torch is
made using a 16mm OD (outside
dia-meter) washer which has five
1mm holes drilled evenly around it.
Insert the K (cathode) lead, which is
the shorter lead, of each LED into a
hole and solder in place. Do this for
five LEDs and each should have about
4mm lead length above the washer.
Also the anode lead should be orientMAY 2001 63
Fig.5: step-by-step assembly of the D-cell version of the torch. Naturally,
this assumes you have already completed the PC board!
64 Silicon Chip
These three photos give a good idea of
the mounting “hardware” associated
with the D-cell PC board. In particular, note the opened-out crimp eyelet
in the shots above and the washer in
the shot at right; also the soldered joint
between the threaded rod and D2.
ed toward the centre of the washer.
The sixth LED is placed in the centre
of the washer with its cathode lead
bent over to be soldered to the washer.
Each anode lead is cut to about 5mm
long and a 27Ω resistor soldered to
it. The other ends of the resistors are
soldered to a tapped spacer so that
there is 25mm between the end of the
spacer and the lower lip of the washer.
The spacer should be mount-ed along
the centre axis of the washer.
The torch bulb holder is unscrewed
from the reflector and the bulb, spring
and contactor plate are removed. Drill
a 3mm hole in the end for the screw.
Remove the reflector cap and glass by
squeezing the cap to an oval shape
and then prising it off. Insert the LED
assembly from the reflector end. Screw
on the bulb holder and secure the LED
assembly with an M3 x 10mm screw
and star washer through the crimp lug
on the PC board. Solder a wire from
the GND PC terminal to the reflector
switch flange.
Two 70mm-long threaded rods are
attached by soldering to the PC stakes
on the positive end of the PC board
and secured to the bulb holder with
a plastic cable tie. This will provide
a stiff mechanical assembly. Solder
diode D2 between the GND PC stake
under the PC board and the threaded
rod as shown.
The inside of the torch includes
a spring as the negative contract for
the cell. This spring is too stiff and
may distort the PC board when it is
assembled inside the torch.
We recommend removing the spring
and squashing it down so that the
overall height is about half of its original. Squash the spring by bending the
smaller diameter loops closer together
with pliers.
The PC board assembly will require
a cardboard tube around it to prevent
it from being caught within the torch
as it is turned while the cap is screwed
on. We made our tube with a piece of
cardboard measuring 72 x 115mm. It
was wrapped around to make a 30mm
ID (inside diameter) cylinder x 72mm
long. We glued the ends with PVA adhesive and used pegs to hold the joint
in place while the glue dried.
The LED array is surrounded with a
cylinder of translucent plastic 23mm
in diameter by 17mm long and it is retained between the reflector and front
glass. This prevents star effects caused
by the reflector focussing the light
emitting from the sides of the LEDs.
The plastic cylinder diffuses this light
to substantially reduce the effect.
Our cylinder was obtained from
the cap cover of a “Fruity Flavorits”
250mm drink container.
The whole assembly can now be
inserted into the torch with the D cell
inserted first, negative end down. Then
place in the diffuser, the reflector glass
and then press on the screw cap. Now
screw the assembly in place.The torch
should operate when switched on. You
can remove the assembly to adjust VR1
for the brightness required. In most
cases this would be at maximum (fully
clockwise) but for some uses it may be
helpful to turn it down.
Testing
If your torch does not work, firstly
check that the cell has voltage across
it. It should be at least 1.0V when
measured with a multimeter. Clean the
cell terminals to ensure good contact
and check that the torch switch is
operating correctly.
Sometimes the switch contact is
bent incorrectly so it does not make
contact with the reflector switch
flange. You can check that the washer
for the LED array makes contact with
the inside of the reflector.
Other problems could be that the
LEDs have been installed with reverse
polarity or the components on the PC
board have been incorrectly oriented
or placed. Check that the leads on IC1
make contact with the PC board tracks.
You can operate the torch using a
power supply which produces about
1.2-1.5V, but make sure the polarity
And here’s the final
assembly, ready
to be placed into
the torch barrel –
naked (left) and
clothed (right)!
MAY 2001 65
Fig.6 (top right) and the above photographs show the “AA” version PC
board from both sides. Inset at right is an enlarged view of the MAX1676
IC – in this version it is soldered direct to the PC board.
is correct. Check that the converter
produces voltage at the “A” terminal.
It should be adjustable from below
3V up to about 4.2V by varying VR1.
AA-cell version
First solder the surface-mount
IC direct to the PC board (see “D”
version for the method used). Next,
insert the PC stakes with the long end
going down into the PC board to give
a similar pin height above and below
the PC board.
Insert and solder all the resistors.
They are shown mounted vertically
in the diagram but should sit parallel
with the PC board. The capacitors go
in next, taking care with the tantalum
types which must be oriented with
the correct polarity.
Now solder in the two diodes and
wire link. Inductor L1 is wound on
a Xenon trigger transformer former.
The original windings are removed
from the trigger transformer; unwind
the primary winding and then cut the
Parts List – AA Cell Version
1 2-AA cell torch (Dorcy FrostBrite or equivalent)
1 PC board coded 11205011, 49 x 13mm (must be solder plated)
1 Xenon tube trigger transformer (L1)
1 900mm length of 0.4mm enamelled copper wire
1 10mm OD steel or brass washer
1 15mm OD x 10mm ID Neoprene “O” ring
1 50mm length of red hookup wire
1 50mm length of green hookup wire
6 PC stakes
Semiconductors
3 5mm white LEDs (LED1-3)
1 MAX1676EUB step-up DC-DC converter (IC1)
1 BYV10-20 Schottky diode (D1)
1 IN4002 1A diode (D2)
Capacitors
2 47µF tantalum capacitors
3 0.1µF monolithic ceramic capacitors (code 104 or 100n)
Resistors (0.25W, 1%)
1 100kΩ
(brown black black orange brown or brown black yellow brown)
1 47kΩ
(yellow violet black red brown or yellow violet orange brown)
3 27Ω
(red violet black black brown or red violet black brown)
66 Silicon Chip
finer secondary wires with a knife.
Unsolder the wires from the end
leads and attach one end of the 0.4mm
enamel copper wire to one end of
the former, making sure the end is
stripped of insulation before soldering. Wind on 45 turns and terminate
the wire to the other end of the former.
The inductor is mounted from the
underside of the PC board.
Solder a 10mm washer to the PC
stakes at the positive end of the PC
board. Solder a short length of hookup
wire between the “A” PC board pin
and one of the end PC stakes.
LED array
The 3-LED array is made within
a torch bulb socket. The details are
shown in Fig.7.
First, remove the glass and filament
from inside it. Wear goggles when
doing this; crack the glass with pliers and scrape out the inside with a
screwdriver. The solder at the end can
be removed with some solder braid or
by using a solder sucker.
Cut the LED anode leads to 5mm in
length and solder each one of these
leads close to the bodies of a 27Ω
resistor. The other end of the resistor
is passed through the solder hole
at the end of the bulb. The K (cathode) leads need to be cut to 5mm in
length and soldered to the rim of the
The reflector must be slightly modified to fit the three LEDs through, as
shown here.
bulb. The metal switch flange is also
tack-soldered to the bulb. Now solder
the resistor leads to the solder end of
the bulb and cut the lead ends flush.
The reflector will need to have cutouts made so that the LED array can
be inserted into the reflector area. You
can do this with a small round file.
The PC board pins at the end of the
board solder directly to the brass end
cap on the bulb holder. This must be
done quickly to avoid melting the
plastic.
We found that the internal spring
contact did not give a reliable connection so we drilled a small hole in
the side of the bulb holder just at the
base of the spring and passed a wire
through this and soldered it directly
to the solder end of the bulb. The
other end of the wire connects to the
“A” PC stake.
Insert the LED assembly into the
reflector and secure the bulb holder
in place with the wire soldered to the
end of the bulb. Now attach a ground
wire to the switch flange.
The positive end of the PC board
requires a 15mm diameter locator so
that it will be centrally positioned
inside the torch. We used a 15mm
outside diameter “O” ring which was
secured with some hot glue adhesive.
Fig.7: here's how to assemble the
“AA” version of the LED torch.
These details suit the Dorcy
FrostBrite torch but should be
adaptable to most similarly
switched and similar size torches.
This circuit can also be powered
by a single “C” cell in installed in a
C-cell torch.
This close-up of the “AA” torch reflector assembly shows the three-LED
array and the way it pokes through
the reflector. In this case, the LEDs are
soldered into the old (filament) globe
base.
The corners of this end of the PC board
will require filing down a little so
that the “O” ring is not distorted out
of shape when attached to the end of
the PC board.
Insert an AA cell (negative end first)
and place the PC board and reflector
assembly into the torch body. Secure
with the reflector cap. The torch
should now work.
If it does not work, check that the
cell has voltage across it. Again, it
should be at least 1.0V when measured with a multimeter. Clean the cell
terminals to ensure good contact and
check that the torch switch is operating correctly. Sometimes the switch
contact may be bent incorrectly so it
does not make contact with the switch
flange on the reflector.
Other problems could be that the
LEDs have been installed with reverse
polarity or the components on the PC
board have been incorrectly oriented
or placed. Check that the leads on
IC1 make contact with the PC board
tracks.
You can operate the torch using a
power supply which produces about
1.2-1.5V, making sure the polarity
is correct. Check that the converter
produces voltage at the “A” terminal.
SC
It should be about 3.9V.
Fig.8: PC board patterns for
the “AA” version (above)
and the “D” version (far
right), with its SMD IC
daughter board at
immediate right.
MAY 2001 67
|