drift, and possible battery contact
problems. Any ideas would be greatly
appreciated.
— Dale P. Appleton
AThe IM- 17 draws less than
500 microamps from the
battery, and the circuit is
designed to be somewhat
independent of the battery voltage.
The regulator circuit will draw more
current than the IM- 17, so battery life
will not be improved. I have in mind a
SEPIC converter using the LTC1872B.
The B version has burst mode
inhibited allowing it to operate at low
current output. Because of the low
current output, the efficiency will not
be better than 50%. The SEPIC circuit
will work when the input is less than
the output, possibly as low as three
volts; the battery will crash however,
when the voltage gets below about
seven volts.
Figure 4 is adapted from the
LTC1872B datasheet; I selected parts
from the Mouser catalog, except for
the LTC1872B which is available from
Digi-Key. The LTC1872B is a current
mode PWM regulator operating at
550 kHz. The threshold current sense
between pin 4 (ISN) and pin 5 (VCC)
is normally 400 millivolts. When the
■ FIGURE 5
26
December 2008
drop across R1
exceeds the
threshold, the
pulse to Q1 is
cut off. The
threshold is
variable via the
feedback to
allow regulation.
The capacitor,
C3, has a dual
function: It
■ FIGURE 4 passes power
from the stored
energy in the primary inductance to
the output and acts as a snubber for
the drain of Q1. Note that when Q1
is turned on, the anode of D1 goes
low. It is the stored energy of the
magnetic field that is transferred to
the output.
BLINKING LAMP
SCHEMATIC
QI’m somewhat of a
newcomer when it comes
to electronics. I need a way
to make a 12 volt, 500
milliamp lamp flash on and off
repeatedly every couple of seconds,
like blinking an eye.
— Gary D. Myers
AThe schematic for the blinking lamp is in Figure 5. The
output of the 555 timer is
high, turning on switch Q1 ,
while C1 is charging. When the voltage at pin 6 (threshold) reaches 2/3
of VCC, pin 7 (discharge) and pin 3
(out) go low. The output remains low
while C1 discharges, until the voltage
reaches 1/3 VCC. At that point, pin 7
goes open, allowing C1 to charge
■ FIGURE 6
through R1 and R2; the output goes
high, completing the cycle. Since the
eye is open most of the time and
closes only momentarily, I designed it
for 0.9 seconds on and 0.2 seconds
off. That is quite fast blinking, so you
can increase the resistor values to
make it slower. I suggest using a film
or ceramic capacitor.
PROGRAMMABLE
CURRENT SOURCE
QI want to design a programmable current source for
checking the forward
voltage of unknown LEDs. I
plan on using a small microcontroller
to display the current and voltage at
the test points. What I want to do is
vary the current output using a PWM
signal from the micro. Many would
say this is just a simple op-amp
current source, but when you don’t
know analog electronics, nothing
analog is simple! I want to vary the
current from 0 to 100 mA in 1 mA
steps and run the whole thing from a
nine volt battery. Can you help?
— Tom Blough
AMy solution to your problem is shown in Figure 6.
The input PWM is divided
by 5 via R1 and R2. Five
volts DC input will require one volt
feedback to the – input, which means
there must be 100 mA flowing
through R3 and the diode under test.
You will have to calculate the diode
drop (Vtp-Vfb), but I don’t see a
simple solution to that problem. If
you make the input pulse width
1/100 of the period, the average will
