This is only a preview of the May 2006 issue of Silicon Chip. You can view 39 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 "Lead-Acid Battery Zapper & Condition Checker":
Articles in this series:
Items relevant to "Universal High-Energy LED Lighting System, Pt.2":
Items relevant to "A Passive Direct Injection Box For Musicians":
Items relevant to "Remote Mains Relay Box":
Items relevant to "Vehicle Voltage Monitor":
Articles in this series:
Purchase a printed copy of this issue for $10.00. |
Deluxe Lead-Acid
BATTERY ZAP
& Condition
Checker
This photo shows the same setup as depicted in Fig.6. A battery charger
is needed to provide current for the zapping function.
Here’s an improved design for a lead-acid battery desulphator
or “zapper”, combined with a battery condition checker. It has
output jacks which let you monitor the zapping pulses with an
external multimeter as zapping progresses, while an inbuilt
isolating choke makes it easy to connect a charger to the battery
during zapping.
T
HE SIMPLE LEAD-ACID battery
zapper we described in the July
2005 issue of SILICON CHIP has been
very popular with readers but a few
shortcomings did become apparent
as people started putting it to work
desulphating their batteries.
For a start, when the zapper was
connected to a battery with a high
6 Silicon Chip
level of sulphation, the high voltage
zapping pulses could rise in amplitude above the 100V rating of the
switching MOSFET, causing it to suffer breakdown. A circuit modification
to limit the maximum pulse voltage
was published in the Notes & Errata
section of the September 2005 issue
(page 107).
We also showed how to connect a
switch in series with one of the leads
between the zapper and battery, to
avoid dangerous sparking at the battery terminals when the connection
was made or broken. A number of
readers also enquired if they could fit
an indicator to show when zapping
was taking place, as it wasn’t easy to
siliconchip.com.au
By JIM ROWE
PER
The circuit fits inside a standard UB2 plastic box and has output jacks so
that you can monitor the “zapping” progress using a multimeter.
be sure of this unless you connected
an oscilloscope across the battery
terminals.
Another complication arose regarding the power MOSFET’s over-voltage
protection, because the MOSFET
used in the July 2005 design became
unavailable and the only replacements
we could find were rated at just 60V.
So the over-voltage limiting had to be
changed again.
It also became clear that batteries
needing desulphation must be connected to an external charger at the
same time, because they couldn’t
provide the zapper with sufficient
current. Although we had shown how
this could be done, it did involve the
use of an external “floating” inductor
in series with one of the charger leads.
siliconchip.com.au
Now we have incorporated the extra
inductor inside the box.
Finally, the original design was only
suitable for 12V batteries but many
readers needed to desulphate 6V or
24V batteries as well.
Clearly, the best way of sorting out
these drawbacks was to develop an improved Mk.2 zapper. At the same time,
we decided to incorporate a battery
condition checker, so that users would
be able to check the condition of their
batteries quickly and easily – either to
see if zapping was necessary or after
a zapping session, to see if there had
been an improvement.
So that’s the story behind this new
unit. It’s largely based on the July 2005
design but with a higher zap output
and the ability to be used with 6V,
12V and 24V batteries. It also has the
bonus of a built-in battery condition
checker.
How it works
The zapper section of the new unit
is very similar to the earlier unit. As
we went into a fair amount of detail
explaining how this worked in the
July 2005 article, we won’t repeat it
in the same detail. It is illustrated in
the three diagrams of Fig.1.
The zapping section of the circuit
is shown in the upper part of Fig.2
(from the negative battery terminal
upwards). In this case, it only operates
when switch S1 is in the “Zap” position, connecting this part of the circuit
to the battery and/or charger.
Current from the battery and/or
May 2006 7
Fig.1(a): during the first phase
of the circuit’s operation,
current flows from the battery
(or charger) and charges a
100mF electrolytic capacitor via
inductor L2.
Fig.1(b): next, the switch is
closed for 50ms, and current
flows from the capacitor into L1.
As a result, the energy stored in
the capacitor is transferred to
the inductor’s magnetic field.
charger flows through 1mH inductor
L2 and charges the 470mF capacitor
connected between the inductor’s
lower end and earth (battery negative).
At the same time, current flows
through RF choke RFC1 and its 100W
5W series resistor, applying battery
voltage to IC1, a 555 timer. Zener diode ZD1 is there to limit the supply
voltage for IC1 to 16V when the unit
is used with a 24V battery (and an accompanying charger).
IC1 is configured as an astable oscillator running at 1kHz, with an output
consisting of narrow positive pulses
about 100ms wide and with spaces of
about 900ms between them (ie, 1:10
mark-space ratio).
The narrow pulses are used to turn
on switching MOSFET Q2, with diode
D2 and transistor Q1 used to ensure
that Q2 is switched on and off as rapidly as possible. So Q2 is turned on for
100ms, off for 900ms and so on.
During the 900ms “off” periods, the
470mF capacitor is able to charge up
to the battery voltage via inductor L2.
When Q2 turns on, it connects the lower end of 220mH inductor L1 to ground,
allowing some of the energy stored in
the capacitor to be transferred into the
magnetic field around L1. Then when
Q2 turns off 100ms later, the magnetic
field in L1 collapses again, delivering
the stored energy back into the circuit
in the form of a high voltage pulse
(positive at the drain of Q2). Most of
the energy in the high voltage pulses
is fed to the battery via fast switching
diode D3.
A number of small changes to the
original zapper circuit have substantially increased the pulse energy.
Over-voltage protection
Fig.1(c): finally, the switch
opens again, interrupting the
inductor current and causing
a high-voltage pulse across
the inductor with the polarity
shown. The green arrow shows
the discharge current path.
8 Silicon Chip
Diode D4 and zener diodes ZD2
and ZD3 form the over-voltage
protection circuit for Q2, limiting the
maximum pulse voltage at its drain to
about 60V.
At the same time, diode D4 also
functions as a half-wave rectifier and
feeds a low-pass filter comprising a
47kW resistor and 100nF capacitor.
This provides a DC voltage proportional to the maximum pulse amplitude
to the “Meter” terminals. This allows
monitoring of the pulse level with a
standard multimeter.
As zapping progresses, the pulses
will initially be quite high in amplitude. But if the zapping is working to
Fig.2 (right): IC1 and MOSFET Q2
provide the zapper function while the
lower section of the circuit involving
IC2-IC5 and MOSFETs Q3-Q6 provide
a battery condition checker.
successfully desulphate the battery, its
internal impedance should drop and
so the zapping pulses will be reduced
in amplitude. So if you are monitoring
the progress with a multimeter, the
voltage should gradually reduce. If it
doesn’t, you know that the battery is
effectively beyond redemption.
Visual indication
LED1 is provided to show when the
zapper is generating pulses and also to
give a rough idea of their amplitude.
Because the pulses are quite narrow,
diode D13 is used to charge the 22nF
capacitor to their full voltage (less
the battery voltage across the 470mF
capacitor) and LED1 is able to draw a
steady current from the capacitor via
the 6.8kW resistor. Incidentally, the
22nF capacitor, in conjunction with
diode D13, also functions as a snubber
circuit to provide further damping of
the high-voltage pulses produced at
the drain of Q2.
The circuitry at upper right in
Fig.2 is to allow safe connection of
a standard battery charger to the battery at any time (ie, during zapping,
condition checking or when neither
is being carried out). Inductor L3 acts
as a blocking choke for the zapping
pulses, preventing the charger from
possibly being damaged, while switch
S3 with its 10nF spark suppressor allows the charger to be safely connected
or disconnected, without producing
any sparks.
The 10W 5W resistor in series with
the negative charger lead is to limit
the current that can be drawn from
the charger, preventing damage when
heavy current pulses are drawn from
the battery during condition checking.
It also reduces the likelihood of overcharging the battery if it is connected
to the Zapper for a period of days.
Condition checking
The condition checking circuit is
broken into two distinct parts: the
centre section of Fig.2 incorporating
IC2, IC3 and transistors Q3-Q8 and
the lowest section involving IC4, IC5
and LEDs 2-9. Essentially, the centre
section is a pulsed current load which
draws a sequence of three very short
siliconchip.com.au
siliconchip.com.au
May 2006 9
Par t s Lis t
1 PC board, code 14105061, 101
x 185mm
1 UB2 size plastic box, (197 x
113 x 63mm)
1 3-pole 3/4-position rotary
switch (S2)
1 DPDT centre-off mini toggle
switch (S1)
1 SPDT mini toggle switch (S3)
1 SPST momentary contact
pushbutton switch (S4)
1 1mH RF choke (RFC1)
1 220mH air cored inductor (L1)
2 1mH air cored inductors (L2,
L3)
1 20mm knob
1 130mm length of 0.5mm tinned
copper wire (PC board links)
1 150mm length of 2.5mm heatshrink sleeving
2 dual red/black binding posts,
19mm spacing
1 pair of 4mm panel-mount banana jack sockets (red/black)
1 M205 LV panel-mounting
fuseholder
1 3A slow blow M205 fuse
cartridge (F1)
4 15mm long M3 tapped metal
spacers
4 6mm long M3 machine screws,
countersink head
4 6mm long M3 machine screws,
round head
3 200mm long x 2.5mm cable ties
1 1.5m length of light duty figure8 flex (for LED connections)
1 600mm length of 13 x 0.12mm
wire, red PVC insulation
1 200mm length of 13 x 0.12mm
wire, black PVC insulation
1 300mm length of 24 x 0.2mm
wire, green PVC insulation
1 100mm length of 41 x 0.3mm
wire, red PVC insulation
1 100mm length of 41 x 0.3mm
wire, black PVC insulation
1 200mm length 13 x 0.12mm
wire, blue PVC insulation
4 QC “eye” connector lugs,
5.3mm ID/9.5mm OD
Semiconductors
1 555 timer IC (IC1)
1 4093B quad Schmitt NAND
gate (IC2)
1 4017B decade counter (IC3)
1 4066B quad bilateral switch
(IC4)
10 Silicon Chip
1 LM3914 LED display driver (IC5)
2 BC327 PNP transistors (Q1,Q7)
5 STP60NF06 N-channel
MOSFETs (Q2-Q6)
1 BC338 NPN transistor (Q8)
5 5mm red LEDs (LED1, LED2,
LED7-9)
2 5mm green LEDs (LED5, LED6)
1 5mm yellow LED (LED4)
1 5mm orange LED (LED3)
1 16V 1W zener diode (ZD1)
1 27V 1W zener diode (ZD2)
1 30V 1W zener diode (ZD3)
1 12V 1W zener diode (ZD4)
1 10V 1W zener diode (ZD5)
9 1N4148 diodes (D1, D2, D6D12)
1 1N4004 1A diode (D5)
1 BY229-200 fast recovery diode
(D3)
2 UF4003 fast power diodes
(D4,D13)
Capacitors
1 2200mF 16V RB electrolytic
1 470mF 63V low ESR RB
electrolytic
2 470mF 25V RB electrolytic
1 10mF 16V tantalum
3 100nF 100V MKT metallised
polyester
3 100nF 50V monolithic
2 22nF 100V MKT metallised
polyester
2 10nF 100V MKT metallised
polyester
1 4.7nF 100V MKT metallised
polyester
Resistors (0.25W, 1%)
1 4.7MW
1 6.8kW
1 270kW
3 4.7kW
3 100kW
1 2.2kW
1 82kW
1 1.2kW
1 47kW
2 1kW
1 27kW
1 470W
1 22kW
1 270W
2 15kW
4 220W
2 10kW
1 100W
2 100W 5W wirewound
1 10W 5W wirewound
3 0.22W 5W wirewound
Where To Buy A Kit
This project was sponsored by
Jaycar Electronics and they own
the design copyright. A kit of parts
is available from Jaycar for $A99.00
– Cat. KC-5427.
high-current pulses from the battery,
shortly after you press the CHECK
pushbutton S4.
The lowest section of the circuit is
basically a sample-and-hold voltmeter,
which samples the battery voltage only
during the last of the three current
pulses and compares it with the battery’s no-load voltage. This indicates
the battery’s condition by showing
how much its terminal voltage droops
under load. In effect, the heavy current
pulses across the battery enable us to
measure its output impedance.
If the battery voltage doesn’t droop
much at all, green LED6 (GOOD) will
light; if it droops by only a small
amount, green LED5 (OK) lights up;
if it droops more but not too much,
yellow LED4 (FAIR) lights up. And if
it droops even more than this, either
orange LED3 (POOR) or red LED2
(FAIL) will light, giving you an idea
of how urgently the battery should be
replaced. This assumes that you have
just charged the battery, of course.
If none of the LEDs light, your battery is dead or flat. If charging and
zapping does not fix it, it is beyond
redemption.
In more detail, the heart of the
pulsed current load section is IC3, a
4017B decade counter. This can count
clock pulses from gate IC2c, which is
configured as a relaxation oscillator
running at about 66Hz. Switch S2a
increases the feedback resistance
when the circuit is connected to a 6V
battery, to maintain about the same
clock frequency.
The oscillator only runs when the
level on pin 9 of IC2c is high and this is
controlled by the “run flipflop” made
up of gates IC2a and IC2b. When power
is first applied to the circuit (ie, when
S1 is switched to the CHECK position),
the flipflop immediately switches to
its “stopped” state, with pins 3 & 5
low and pins 2 & 4 high. So IC2a is
prevented from oscillating and at the
same time, IC3 is held in its reset state
by the logic high applied to its MR pin
(15). The only output of IC3 at logic
high level is O0, pin 3.
No further action takes place until
you press the CHECK pushbutton
(S4), whereupon one side of the 22nF
capacitor connected to pin 1 of IC2a
is pulled down to ground, forcing it
to charge via the 10kW resistor. Until
it charges, pin 1 of IC2a is pulled low,
causing pins 3 & 5 to swing high and
pins 2 & 4 to swing low. Thus, clock
siliconchip.com.au
Fig.3: the scope waveforms at left were measured using a 12V battery with a series resistor of 2.7W to simulate a
sulphated battery. The lowest trace (yellow) is the pulse train fed to the gate of Q2 while the top trace (purple) is the
resultant high-voltage pulse developed at the drain of Q2. The blue trace shows the accompanying ripple voltage
across the 470mF low-ESR capacitor. At right is the sequence of three current pulses used by the condition checker
(measured across the paralleled 0.22W source resistors).
oscillator IC2c is enabled and at the
same time the reset is removed from
pin 15 of IC3.
The counter begins to count the
pulses from IC2c and its outputs then
switch high in sequence: first O1, then
O2, O3 and so on up to O9. Each counter output switches high for around
15ms (milliseconds), so the complete
sequence takes 9 x 15 = 135ms.
When output O9 finally drops low
again at the end of the ninth clock
period, the 100nF capacitor connected
between this output and pin 6 of IC2b
feeds a negative-going pulse back to
IC2b, which resets the flipflop. This
stops the clock and activity again
ceases until S4 is pressed again.
So IC2a, IC2b, IC2c & IC3 form a
digital sequencer which generates nine
15ms long pulses when pushbutton
S4 is pressed.
Diodes D9, D8 & D7 are connected
to the O1, O5 & O9 outputs of IC3.
These diodes form an OR gate to feed
the inputs of IC2d which are normally
pulled down to 0V via a 22kW resistor. But when the sequencer runs and
outputs O1, O5 & O9 switch high in
turn (with 45ms gaps between them),
the inputs of IC2d also switch high for
15ms each time.
As a result, IC2d’s output (pin 11)
switches low during these three 15ms
periods, providing pulses of base current to turn on transistor Q7 for the
same periods. And when Q7 conducts,
it turns on MOSFETs Q3-Q6, to draw
pulses of current from the battery.
siliconchip.com.au
Q3-Q6 are enhancement-mode MOS
FETs connected in parallel, with their
drains connected to battery positive
and sources connected to battery negative via a parallel combination of three
0.22W resistors, giving an effective
common source resistance of 0.073W.
The MOSFET gates are pulled down
to 0V via a 4.7kW resistor, so normally
they are switched off and not conducting. But when the sequencer turns on
Q7 for three 15ms pulses, this also
turns on the MOSFETs and they draw
pulses of current from the battery.
Pulse current limiting
The battery current pulses are
limited by transistor Q8 and the two
diodes connected in series with its
emitter, in conjunction with the three
0.22W resistors in the source circuit of
the MOSFETs. The base of Q8 is connected directly to the top of the source
resistors, so that when the MOSFETs
conduct, the resulting voltage across
the source resistors provides forward
bias for Q8.
Q8 doesn’t conduct to any significant extent until the voltage drop
across the MOSFET source resistors
rises above 1.95V, where it matches
the forward voltage drop of D11, D10
and Q8’s own base-emitter junction.
When that voltage level is reached,
Q8 begins to conduct, draining away
some of the MOSFET forward bias
reaching their gates via the 470W and
100W resistors. As a result, the MOSFET current is automatically limited
to a value which produces about 2V
across the source resistors; ie, around
2V/0.073W, or 28A.
So when you press pushbutton S4, a
sequence of three pulses of around 28A
is drawn from the battery, each around
15ms in duration and 45ms apart.
Checking the droop
As explained earlier, the circuitry
around IC4 and IC5 forms a sampleand-hold voltmeter. It compares the
battery voltage during the last of the
Warning!
This circuit generates high voltage
pulses which could easily damage
the electronics in a vehicle. Do not
connect it to a car battery installed
in a vehicle.
Disclaimer!
Not all batteries can be rejuvenated by zapping. They may be too
heavily sulphated or may have an
open-circuit cell connection. Nor
can the zapper restore a battery
which is worn out; ie, one in which
the active material on the plates has
been severely degraded.
Depending on the battery, it is also
possible that any rejuvenation effect
may only be temporary.
May 2006 11
Fig.4: follow this parts layout diagram to assemble the PC board and complete the external wiring. Make sure
that all polarised parts are installed with the correct orientation.
three 15ms pulses against the voltage when no current is being drawn,
because this “droop” is a fairly good
indicator of the battery’s condition.
The heart of the voltmeter is IC5, an
LM3914 LED driver IC. The LM3914 is
basically a set of 10 voltage comparators, with the reference inputs of the
comparators connected to taps on an
12 Silicon Chip
internal voltage divider. The top of
the divider connects to pin 6, while
the bottom connects to pin 4.
The second input of all 10 comparators is fed with the input voltage
from pin 5, via an internal buffer
amplifier.
The outputs of the comparators are
used to drive current sinks, one for
each LED driver output pin. Only five
LEDs are used, with each LED connected to an adjacent pair of outputs
so they provide a resolution of only
five voltage ratio levels.
Although the LM3914 has an internal voltage reference, it’s not used
here. The reference pin (pin 7) is simply connected to ground via a 1.2kW
siliconchip.com.au
Our prototype has the
LEDs mounted on sleeved
standoffs, for clarity. In
practice, the LEDs are
wired with flying leads
and fitted into bezels in
the lid.
Table 1: Resistor Colour Codes
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
No.
1
1
3
1
1
1
1
2
2
1
3
1
1
2
1
1
4
1
Value
4.7MW
270kW
100kW
82kW
47kW
27kW
22kW
15kW
10kW
6.8kW
4.7kW
2.2kW
1.2kW
1kW
470W
270W
220W
100W
resistor, to set the LED current levels
correctly.
So that we can use the circuit to
compare the on-load battery voltage
with its off-load value, we use the offload battery voltage as the voltmeter’s
reference. Actually, we use a proportion of the battery voltage, as selected
by switch S2b, because the LM3914’s
siliconchip.com.au
4-Band Code (1%)
yellow violet green brown
red violet yellow brown
brown black yellow brown
grey red orange brown
yellow violet orange brown
red violet orange brown
red red orange brown
brown green orange brown
brown black orange brown
blue grey red brown
yellow violet red brown
red red red brown
brown red red brown
brown black red brown
yellow violet brown brown
red violet brown brown
red red brown brown
brown black brown brown
input voltage range must be limited
for linear operation. So S2b selects
a suitable proportion of the battery
voltage, depending on whether a 6V,
12V or 24V battery is being tested. This
voltage is fed through a 1kW resistor
and diode D12 to charge the 470mF capacitor and this provides our ‘no load”
voltage reference for the LM3914.
5-Band Code (1%)
yellow violet black yellow brown
red violet black orange brown
brown black black orange brown
grey red black red brown
yellow violet black red brown
red violet black red brown
red red black red brown
brown green black red brown
brown black black red brown
blue grey black brown brown
yellow violet black brown brown
red red black brown brown
brown red black brown brown
brown black black brown brown
yellow violet black black brown
red violet black black brown
red red black black brown
brown black black black brown
Table 2: Capacitor Codes
Value
100nF
22nF
10nF
4.7nF
μF Code
0.1µF
.022µF
.01µF
.0047µF
EIA Code
104
223
103
472
IEC Code
100n
22n
10n
4n7
May 2006 13
Fig.5: this cross-sectional diagram shows the mounting details for
the LEDs and the rotary switch.
The top of the LM3914’s internal
voltage divider is connected to the
top of the capacitor, while the bottom of the divider is connected to
ground/battery negative via a 15kW
resistor. This expands the range of the
LM3914’s comparator voltage divider
to the upper 40% of the total reference
voltage.
Voltage sampling
Sampling of the on-load voltage
is performed by IC4, a 4066B quad
bilateral switch with all four switches
connected in parallel to minimise
on-resistance. The control inputs of
the switches are connected to the
O9 output of IC3, so the switches are
normally “off” and are only turned
on during the third pulse of each load
pulse sequence. When this occurs,
the switches allow the 10mF capacitor
connected to pin 5 of IC5 to charge up
to the proportion of battery voltage
selected via S2b – the same voltage
proportion used to charge the 470mF
capacitor but in this case it samples
what happens to it when the battery
is attempting to provide 30A pulses
of current.
The LM3914 therefore compares the
selected proportion of the battery’s
no-load voltage (pin 6) with the same
proportion of its on-load voltage (pin
5). If the voltage droops very little,
LED6 will light; if it droops a little
14 Silicon Chip
more, LED5 will light and so on.
Note that if the on-load battery voltage drops below 60% of its no-load
value, none of the LEDs will light –
that’s why a “no glow” indicates that
the battery is either flat or completely
dead.
Note too that regardless of which
LED lights during the test to indicate
battery condition, after a few seconds
the glow will transfer down through
the lower LEDs and then finally they’ll
all go dark again. That’s because the
sampled on-load voltage held by the
10mF capacitor is gradually leaked
away by the parallel 4.7MW resistor,
to ready the circuit for another test.
The 10V 1W zener diode (ZD5)
connected to the wiper of switch S2b
is there to protect the inputs of IC4 &
IC5, in case the 6V battery position is
selected while a 24V battery is connected. Without ZD5, both IC4 & IC5
could be destroyed by this mistake.
The third pole of switch S2 (S2c) is
used to indicate which battery voltage
has been selected, via LED7-LED9.
Construction
To make the new Battery Zapper
& Checker reasonably easy to build,
almost all of the components used
are mounted directly on a PC board
coded 14105061 and measuring 101 x
185mm. This has rounded cutouts in
each corner so it will fit snugly inside
a standard UB2-size plastic utility
(Jiffy) box.
The only components which don’t
mount on the PC board are the LEDs,
switches S1, S3 & S4, the fuseholder
for fuse F1 and the various input terminals and banana sockets. The three
switches mount on the lid of the box,
while the fuseholder and terminals
mount on the sides of the box. All of
these off-board components connect to
the board via short lengths of insulated
wire – see Fig.4.
Begin the board assembly by fitting
the seven wire links. Don’t forget the
short link between diodes D7 and D9,
just to the right of rotary switch S2,
or the longer link just to the left of the
same switch.
Next, fit the smaller resistors and
the small RF choke (RFC1), followed
by the 5W wirewound resistors. Take
care to fit the three 0.22W resistors in
their correct positions just below the
indicated position for inductor L3.
Next, fit the capacitors, starting with
the smaller non-polarised multilayer
monolithic and MKT parts and then
progressing through to the polarised
tantalum and electrolytic types. There
are not many of these but take care to fit
them with the correct orientation.
Now you can fit the semiconductors,
starting with the various diodes and
then the bipolar transistors (Q1, Q7 &
Q8), the ICs (or sockets for them if you
wish) and the power MOSFETs. The
semiconductors are all polarised, so be
sure to install them correctly.
When fitting MOSFETs Q3-Q6, leave
about 5mm of their leads above the
board (ie, the wider 4mm long sections
plus a further 1mm). This is necessary
because they need to be bent over at
about 45° later, so that their top tabs
clear the contacts of switch S1 when
everything is assembled. Although not
shown in the photos, the two lower
MOSFETs must be bent downwards
towards D11, while the upper MOSFETs are bent upwards towards L3.
Mounting the LEDs
The LEDs are all connected to the
PC board using 150mm lengths of
light-duty figure-8 flex and the LEDs
themselves fitted into bezels on the
front panel. Each LED is fitted with
its connecting lead first. Do this
by separating the two lead wires at
one end for about 20mm and then
removing about 6mm of insulation
from each. Then slip a 15mm length
siliconchip.com.au
of 2.5mm heatshrink sleeving down
over each wire, before soldering the
two wires to the LED leads (which
have been previously cut short, to
about 12mm long).
When you solder the wires, make
sure you solder the wire with the
black stripe to the LED’s cathode
lead. After both joints are made,
slide the heatshrink sleeves up
and over the solder joints, so
they are fully covered, and heat
them with a hot-air gun or by
rubbing them with the barrel of
your soldering iron, so they shrink
into place.
Once the leads have been fitted,
the LEDs can all be attached to the
PC board. Be sure to fit them in the
correct positions and with the correct
polarity.
Special note: our photograph of
the prototype shows all LEDs except
LED1 mounted on sleeved standoffs
about 40mm high, just high enough to
let the LEDs protrude through the lid.
This has the advantage of showing an
uncluttered board in our photographs
and allowing more easy comparison
with the wiring diagram of Fig.4.
That done, it’s time to fit the largest
components to the board – ie, rotary
switch S2 and the three air-cored inductors. There’s no need to cut S2’s
control shaft before it’s fitted to the
board. Instead, it’s left at full length so
that it will later protrude far enough
through the box lid to accept the
control knob. However, you do need
to make sure that the switch is set for
only three positions.
This is done by first turning the
control shaft as far as it will go in
the anticlockwise direction and then
unscrewing the mounting nut and
removing this from the threaded ferrule, along with the star lockwasher.
That done, use a small screwdriver to
prise up the indexing pin washer from
its position under the star lockwasher
and then carefully replace it so that
its indexing pin slips down into the
rectangular hole between the numerals ‘3’ and ‘4’ which are moulded into
the plastic. Make sure the washer is
sitting down flat before replacing the
lockwasher and mounting nut.
If you now try turning the control
shaft by hand, it should have three
only possible positions.
You can now fit the switch in position, making sure that all its connection pins pass through the board holes
siliconchip.com.au
This is the view inside the completed prototype. Note that in the kit
version, the LEDs are connected to flying leads and clipped into bezels
mounted on the front panel.
and that the bottom of the switch sits
flush against the board. Note that in
this project, the switch orientation is
NOT with the locating spigot pin at 12
o’clock but at the 5 o’clock position.
This is shown clearly on the overlay
diagram (Fig.4).
When you are happy that the switch
is orientated correctly and is sitting
flat on the board, turn the board over
and solder all of the pins to the pads
underneath.
Air-cored inductors
The air-cored inductors are also
mounted directly on the PC board.
It’s important to dress each inductor’s
leads carefully so they’re each straight
and at close to 90° to the side cheeks of
the inductor bobbin, to prevent strain
as the inductor is lowered against the
board.
Make sure also that you orientate
each inductor so that its “start” lead
(nearer the centre of the bobbin) passes
through its matching “S” hole on the
board. The “finish” lead (further out)
goes through the hole marked “F”.
When each inductor is sitting flat
against the top of the board, you can
solder its leads to the pads underneath
and trim off any excess. That done,
use a 200mm-long cable tie to hold
the inductor in place, passing the tie
down through one of the edge holes
provided in the board and up through
the other.
All that remains now is to plug the
ICs into their sockets (taking care to
fit them with the correct orientation)
and then prepare and fit the various
short lengths of wire for the off-board
connections.
There are 14 of these connection
wires to be prepared: two each for the
charger and battery terminal connections; two for the meter jacks; two for
the charger on/off switch (S3); two for
the main function switch (S1); two for
pushbutton switch S4; one for the end
terminal of the fuseholder; and finally,
one for the connection between the
May 2006 15
fuseholder side lug and the centre lugs
of switch S1.
To make it easier to prepare all these
wires, their details are shown in Table
1. Note that the wires for the meter
terminals are of light-duty hookup
wire and this also applies to the wires
for S4, S1 and the fuseholder. On
the other hand, the wires for charger
switch S3 and especially the charger
and battery terminals should be made
from heavier wire, because they carry
higher currents.
Warning!
Hydrogen gas (which is explosive) is generated by lead-acid
batteries during charging. For this
reason, be sure to always charge
batteries in a well-ventilated area.
Never connect high-current loads
directly to a battery’s terminals.This
can lead to arcing at the battery
terminals and could even cause the
battery to explode! Note too that the
electrolyte inside lead-acid batteries is corrosive, so wearing safety
glasses is always a good idea.
16 Silicon Chip
Note also that the wires for the meter
jacks have matching large solder lugs
fitted to their far ends, while the wires
for the charger and battery terminals
are fitted with suitable “QC eye” connector lugs (see parts list) for easy attachment to the rear of the terminals
using the nuts provided.
Once all of these wires are prepared,
you can pass the “board end” of each
wire through its corresponding hole on
the board and solder it to the pad underneath. Your board assembly should
then be complete and ready to be fitted into the box, although you should
first give it a thorough inspection, to
make sure there are no dry solder
joints, joints that have been forgotten
altogether or accidental solder bridges
between pads or tracks.
Final assembly
Before lowering the board assembly
into the box, secure the four 15mmlong tapped spacers inside the bottom
of the box using countersink-head M3
x 6mm machine screws. That done,
lower the board onto the spacers and
secure it in place using four round-
head M3 x 6mm machine screws.
Next, fit the meter connection jack
sockets, the charger and battery connection terminals and the fuseholder
to the sides of the box. With both the
meter jacks and the charger/battery
terminals, you have to disassemble
them first before you can fit them to
the box and then reassemble them with
a single nut inside. When you have
tightened these nuts, slip the solder
lugs or QC connectors over the ends of
the threaded sleeves or shafts and then
add a second nut to each connector to
fasten them in place.
The fuseholder is pushed through
its mounting hole and the washer
and nut refitted. Don’t use excessive
force to tighten the nut though, as this
may strip the plastic thread. Once the
fuseholder is in place, you can solder
the end of the wire from the PC board
to its end connection lug.
Next, fit toggle switches S1 & S3 to
the box lid. S3 is a single-pole switch
which mounts in the central hole of
the lid, while S1 is a double-pole
centre-off switch which mounts in the
righthand hole. After these, fit pushbutton switch S4 in the centre hole at
the bottom of the lid.
You should now be ready to make
the last off-board connections, so turn
over the box lid and bring it close
alongside the box itself. First of all, use
the remaining loose length of prepared
wire (80mm of 13 x 0.12mm, red PVC
insulation) to connect the side lug of
the fuseholder to the two centre terminal lugs of switch S1 (note: the two
sections of S1 are connected in parallel, to give greater current handling
capacity). That done, solder the free
ends of the remaining red wires from
the board (“S1a” and “S1b”) to the lugs
at each end of switch S1 – see (Fig.2).
The “S1a” wire goes to the two lower
lugs of S1, while the “S1b” wire goes
to the two upper lugs.
Next, solder the leads to pushbutton
switch S4. The switch wiring can then
be completed by soldering the free
ends of the two green wires coming
from centre left of the PC board to the
centre and uppermost lugs of S3 (the
charger on/off switch).
Panel-mounting the LEDs
You can now fit the plastic bezels for
the nine LEDs into their holes in the
lid. When each bezel is in place, push
its LED up from below until it clicks
into place. Just make sure you fit each
siliconchip.com.au
LED into its correct position or you’ll
get some strange results later!
That done, you can lower the lid
down onto the box, with the rotary
switch spindle passing through its
clearance hole. Fasten it with the selftapping screws provided, fit the small
plastic bungs over each screw recess
and fit the control knob on the rotary
switch spindle.
Using it
Now for the smoke test. First, make
sure that the Zapper’s switches are
set as follows: S1 in its centre-Off
position, S2 for the correct nominal
battery voltage and S3 in its upperOff position. That done, connect it as
shown in Fig.6. The Zapper’s battery
terminals are connected directly to the
battery using heavy-gauge cables. Just
make sure you connect the positive
terminal to battery positive and the
negative terminal to battery negative,
or very nasty things can happen.
If you are going to zap the battery,
you’ll also have to connect your
charger to the Zapper’s charger terminals: again, positive to positive and
negative to negative. This is because
a sulphated battery cannot deliver the
200mA or so of current required by
the Zapper.
Once the charger is connected,
switch S3 on the Zapper to “On” (assuming you’ve already connected the
Zapper to the battery). Note that if you
are using a multimeter to monitor the
zapping pulses, it should be set for a
DC voltage range of 20V or 50V.
To begin zapping the battery, switch
S1 to its “Zap” position. The Zapping LED should immediately light,
showing that the high-voltage zapping
pulses are being applied to the battery.
If you have a multimeter connected,
it should be giving a reading of about
30V DC or thereabouts; this is not the
actual peak-to-peak pulse voltage but
an average value proportional to it.
As zapping progresses, this voltage
reading should slowly drop, as the
lead sulphate crystals in the battery
are gradually dissolved. So let’s say
you’ve been zapping the battery for a
day or two and also charging it at the
same time. Now you want to check
the battery’s condition. This is done
as follows:
First, turn the Zapper’s Charger
switch S3 to the Off position, so you’ll
be checking the battery by itself and
not the charger as well. Then, after
siliconchip.com.au
Fig.6: this diagram shows how the Zapper
is connected to a battery and charger. The
multimeter monitors the zapping pulses.
making sure S2 is set for the battery’s
nominal voltage (6V/12V/24V), move
function switch S1 down to its lower
Check position. One of the LEDs above
the knob for S2 should light, confirming the battery voltage setting. The
Good Condition LED (LED6) will
also light briefly, then the OK LED,
the Fair LED and so on, down to the
Fail LED. This “ripple down” effect
is caused by the time taken for the
LM3914 reference voltage to stabilise
after switch-on.
Once the Condition LEDs have all
gone dark again, simply press the
Check pushbutton (S4) briefly. Now
one of the Condition LEDs should
light again, to show the battery’s actual
condition – hopefully it will be the
“Good” or “OK” LED, if the battery
has responded to the zapping.
After a few seconds, the lit LED will
fade out and the LED next down from
it will light instead. Then the next LED
to its left will light and so on, until all
Machine screws can be fitted to the
Zapper’s charger terminals to provide
handy contact points for the battery
charger’s alligator clip leads.
five LEDs are dark again.
When they are all dark it’s a good
idea to press S4 again for a second
check, because a single check may
give a reading that’s lower than the
battery’s actual condition. So if you do
press S4 again, you’ll very likely get
a higher reading than the first time if
the battery really is in “Good” or “OK”
condition.
If you only get a reading of “Fair”,
“Poor” or “Fail”, even on the second
check, your battery isn’t in good shape
and needs more zapping. And if further zapping doesn’t give better readings, your battery is essentially dead
and ready for replacement.
By the way, you can check the battery
condition any time you wish. Because
each check only draws three very short
pulses of current from the battery, it
draws a negligible amount of charge –
about 1.35 coulombs or 0.000375Ah.
Your charger can probably replace this
in a couple of seconds.
You’ll also notice that when you
exit the battery checking function by
switching S1 back to its centre-off position, the Condition LEDs again light
briefly, this time from the lowest to the
highest. This occurs as the LM3914’s
reference voltage decays and is nothing to worry about.
By the way, note that regardless of
the battery charger you use, the charge
current is limited by the circuit to less
than 1A. We did this because we did
not want the risk of severely overcharging a battery during a period of
zapping over several days. So after
zapping successfully, the battery may
SC
still need further charging.
May 2006 17
|