Battery capacity tester
You’ll appreciate this circuit if you’ve gathered a large
collection of rechargeable batteries over the years and have no idea of their
condition. This circuit will measure their capacity and display the results on a
digital voltmeter (DVM).
The circuit (Fig.1) calculates ampere-hours (Ah) by measuring
the time that it takes to discharge the battery under test to a preset cutout
voltage, using a constant discharge current. Both the cutout voltage and
discharge current are adjustable over a wide range.
The discharge current is controlled by applying a variable
reference voltage to the non-inverting input (pin 10) of IC1c. This op amp
functions as a voltage follower; it attempts to maintain the voltage at its
inverting input (pin 9) – and hence the voltage across the 1W (or 0.1W) sense
resistor – equal to the reference voltage.
Together with the Mosfet (Q1), this configuration yields an
adjustable constant-current sink. Current ranges of 60-800mA and 0.6-8A are
selectable using toggle switch S1, which simply selects between the 1Ω and 0.1Ω
sense resistors.
Overall, the response of the circuit is non-linear, hence the
need for a pot (VR1) with a log taper as part of the adjustable voltage
reference network. Note also that the pot must be wired in reverse to what you
might expect – so that turning it clockwise decreases the current
setting.
The reset switch (S3) is pressed to initiate a measurement
cycle. This applies a logic high level to the SET input (pin 6) of IC2a, a "D"
type flipflop, driving its output (pin 3) high. This releases the RESET input of
the 555 timer (IC3), which then begins to oscillate at about 0.7Hz, flashing the
LED and clocking a 4020B 14-bit binary counter.
The eight most significant bits of the counter (Q7-Q14) are fed
into IC5, an 8-bit multiplying digital-to-analog converter. The converter
multiplies the counter’s value with the reference current into its
VREF pin. The reference current varies as the discharge current
varies, as both are set by VR2.
It thus follows that the D-A converter’s output current is
proportional to the discharge current multiplied by the digital count. As we’ve
seen, the count is proportional to time, so the converter’s output is
proportional to amperes x time (ie, Ah).
Op amp IC1a is used to convert the output current to a voltage
that directly corresponds to battery capacity, such that 1V = 1Ah. Trimpot VR3
allows the op amp’s gain to be trimmed for calibration purposes.
During the discharge cycle, the battery voltage is monitored by
IC1d, which is wired as a voltage comparator. When the terminal voltage drops
below the cutout voltage (as set by VR2), the IC1d’s output (pin 14) swings
high, resetting the flipflop (IC2a). This terminates the battery discharge by
pulling the gate of Q1 low via D1. It also resets IC3, which stops oscillating,
freezing the count accumulated by IC4. The output of IC1a now sits at the
measured "Ah" value until the reset switch is pressed again.
Clock frequency
The clock frequency is determined by the 1MΩ resistor and 1μF
capacitor connected to pins 2 & 6 of the 555 timer. With the values given,
the period is about 1.4 seconds, which results in a discharge time of about 6
hours (1.4 x 16,384 seconds). To increase the discharge time, increase the value
of the capacitor and/or resistor.
Note that increasing the discharge time may necessitate an
increase in the value of VR3 to allow for a higher Ah reading. Of course, it
will also mean that the LED flashes at a slower rate!
To prepare the system for a discharge test, first connect a DVM
between test point "B" and ground and adjust VR2 to the desired cutoff voltage.
For an NiMh cell, this might be 1.0V, Li-ion 3.0V or SLA 11.0V (refer to the
battery manufacturer’s data for recommended cutoff voltage figures).
Next, set the discharge current to minimum before pressing the
reset switch to start the test. Now connect your DVM to point "C" and adjust VR1
to obtain the desired discharge current. Calibration
To calibrate the circuit, set it going as described above and
note the time or start a stopwatch. Again, measure the voltage at point C to get
the discharge current in amps. After about half an hour, monitor the voltage at
point D. Note the time and the voltage reading when it suddenly jumps upwards.
The counter increments every 64 clock pulses – about a minute and a half – so
you will have to wait for the change.
Now calculate the actual Ah (amps x hours) and adjust VR3 until
the voltage at point D is the same as the calculated value. Repeat the procedure
a couple of times at half hourly intervals to check for linearity.
Note that the Ah reading may be incorrect if the discharge is
not completed within the 6-hour period. The counter will recycle to zero and
continue counting but there will not be any indication that this has occurred.
Always check that the discharge current is set high enough to complete the
discharge within the 6-hour period. Guy Burns,
Ulverstone, Tas.
Fig.1: the circuit calculates ampere-hours (Ah) by measuring the time that it takes to discharge the battery under test to a preset cutout voltage, using constant discharge current.
|
Temporarily silencing a smoke detector
This circuit is an outgrowth from the "Fit a Kill Switch To
Your Smoke Detector" project in the February 1996 issue of SILICON
CHIP. It provides a means of temporarily silencing a battery-powered smoke
detector after you’ve burnt the toast, scorched the baked beans – or
whatever!
Unlike the earlier design, this more sophisticated version does
not cause strange chirps and whistles to emanate from the smoke detector towards
the end of the silenced period. It also flashes a LED and produces a series of
short, unobtrusive tones from its inbuilt buzzer while it is active.
A separate 9V (or 6V) battery is required to power the circuit,
which is mounted remotely from the smoke alarm. Connection to the alarm is made
via a 3-core data cable terminated in a 3.5mm stereo plug, while a matching
switched socket is fitted to the alarm’s casing.
In addition to the socket, only three other components are
installed inside the smoke alarm. These are Mosfet Q3, its 100W gate resistor
and 15V zener
diode ZD1. These parts can all be mounted on a small section
of prototyping board or soldered point to point from the socket terminals.
The Mosfet is wired in series with the smoke alarm’s negative
battery lead and acts as a switch. As shown, the contacts of the socket must be
wired so that the Mosfet drain-source connections are shorted out when the plug
is removed, thus allowing immediate restoration of the smoke alarm to normal
operation.
When the silencer circuit is inactive, the reed relay (RLY1) is
off, so battery power is disconnected from the circuit. An exception to this is
Q3’s 4.7kΩ gate pull-up resistor, which is powered directly from the battery.
This holds the Mosfet switch on, powering the smoke alarm from its on-board 9V
battery.
Now consider what happens when the "silence" switch (S1) is
pressed. This action applies battery power to the entire circuit through the
switch contacts. At the same time, IC1 (which is wired as a monostable) is
triggered by a brief pulse on its reset input (pin 2). This initiates the 555’s
timing sequence, so its output (pin 3) immediately swings high, switching on Q1
and activating the relay.
A second transistor (Q2) wired to IC1’s output also conducts,
pulling Q3’s gate low and switching it off. As a result, the smoke alarm is
disconnected from its 9V battery and all of the noise ceases instantly!
When the relay is closed, an additional path exists from
battery positive to the circuit’s power rail – so that when the switch is
released, the circuit keeps running. The circuit then continues to run for the
duration of IC1’s timing period (over 8 minutes).
The remaining two 555 timers (IC2 & IC3) are configured as
astable multivibrators. IC2 is used exclusively to flash an indicator LED at a
rate of about once per second. IC3 has a longer period, sounding a piezo buzzer
briefly about once every 10.5 seconds.
Use a 5V reed relay when the circuit is powered from a 6V
battery and a 12V version when powered from 9V. Because of the high impedance
and low leakage of the Mosfet’s gate, the silencer’s battery can be expected to
last almost its shelf life – assuming that you don’t burn the toast too often!
Warning: (1) this circuit must not be used with mains-connected
smoke detectors; (2) test your smoke detector and this silencer circuit
regularly.
W. A. Fitzsimons,
Mount Eliza, Vic. ($40)
Cheapskate's headset adapter
Here’s a cheaper and easier method of making a telephone
headset adapter than that described in the July 2002 edition of SILICON
CHIP. All that’s required is a cheap headset ($5 at Harvey Norman), a DPDT
switch and a few connectors.
Each transducer in the headset measures about 40Ω. There is
also an inline volume control measuring about 500Ω per leg, across which each
earphone is connected. This means that each earpiece has a minimum resistance
(at maximum volume) of about 37Ω.
>As described in the SILICON CHIP
project, 128Ω is the desirable impedance. This can be achieved by wiring the
earphones in series and adding a 56Ω resistor. Although there is a reduction in
maximum volume due to the resistor, this was easily accommodated by the author’s
telephone, which has an amplified audio output.
The headset does not have a connection between the earphones
and the microphone, so no other modifications were required. Series connection
of the earphones is achieved by not picking up the sleeve connection at the
socket and connecting only across tip and ring.
As the author’s telephone uses an electret microphone in its
hand-piece, no additional biasing circuitry is needed for the headset’s
microphone. Brian Critchley,
Elanora Heights, NSW. ($30)
|
Reservoir pump controller
This circuit operates an automotive windscreen washer pump to
fill a 20-litre drum from a 205-litre water reservoir. The drum is suspended
above a drip line, which irrigates a vegetable garden.
Two stainless steel probes mounted in the drum act as sensors
for the system. One probe is positioned at the high water mark, the other at
about half-full. The pump power is switched by a 12V automotive relay
(RLY1).
Two op amps (IC1a & IC1b) connected as voltage comparators
form the basis of the circuit. Initially, assume a falling water level with the
pump switched off.
When the water level exposes the lower probe, the non-inverting
input (pin 5) of IC1b rises to about 7.4V. With trimpot VR2 correctly adjusted,
this will be higher than the voltage on pin 6. The output (pin 7) therefore
swings high, biasing Q1 into conduction. This in turn causes Q4 to conduct,
switching on the relay and starting the pump.
In addition, when Q4 switches on it supplies base current to Q3
via a 6.8kΩ resistor. Initially, this current flows through the 47μF capacitor,
but once its base-emitter voltage reaches about 0.6V, Q3 conducts. This action
latches Q4 in the "on" state, as its base current can flow to ground via Q3 when
Q1 stops conducting – which will occur when the rising water level reaches the
low probe.
When the water level reaches the high probe, the voltage on the
non-inverting input (pin 2) of IC1a decreases markedly due to the conductivity
of the water. If trimpot VR1 is correctly adjusted, the output (pin 1) swings
high, switching on Q2. This discharges the 47μF capacitor and robs Q3 of its
base current, switching this transistor off. This in turn switches off Q4 and
the relay.
The zener diodes and 1kΩ series resistors at the probe inputs
protect the op amp’s high impedance inputs from the effects of static discharge.
The 47μF capacitor in parallel with the base-emitter junction of Q1 prevents the
latching function from being activated when power is applied to the circuit.
The author’s setup is powered from an old car battery charged
from a 12V solar panel.
Peter Howarth,
Gunnedah, NSW. ($35)