voltages, but that’s optional. Also, I
used pins and jumper blocks for J1,
J2, and J3. You may want to just
solder in jumper wires. Note the
stand-offs mounted at each corner.
The first board I assembled
started to smoke when I applied the
input voltage. After checking the
board several times, I decided it
really should have worked. As it
turned out, several of the tantalum
capacitors I had were marked
incorrectly; the polarity was
reversed. That’s what I get for
buying them at a flea-market. Once
the caps were replaced, it worked
fine. With no load, VOUT was - 9.01V
when Vin was + 9.02V. Using a 500Ω
load and measuring the drop in
output voltage, I calculated that the
circuit had a source resistance of
35 ohms. That was consistent with
the TC962 datasheet value. The
maximum output current of a
TC962 is 80 mA.
To measure ripple from the
switching circuit, I put a 1 kΩ
load on the output of the TC962.
With the J3 jumper installed,
I measured 50 mVpp at 24 kHz. With the jumper removed,
I measured 90 mVpp at 12 kHz. Increasing the value of C4
should reduce the ripple.
■ FIGURE 2A.Voltage mirror circuit board.
■ FIGURE 2B. Solder side.
Assembly is easy. Start with the LM317 (refer to
Figure 2A). First, apply a bit of solder to the two pin pads
and the copper that will be under
the chip. (That’s a bit of solder, not
a big glob.) Position the D-PAK
chip onto the pads and hold it in
place with a finger; use just
enough pressure to hold it steady.
Then, apply your soldering iron
to one of the two small pins. The
solder you applied earlier should
melt and flow over the pin.
Remove the iron, and let the pin
cool for a few seconds. Now, do
the same to the other small pin. At
this point, you no longer need to
keep your finger on the chip.
Apply solder to the metal edge
of the D-PAK that is lying on the
patch of copper. Make sure the solder flows evenly and
bonds the metal edge to the copper. Let it cool for a few
■ FIGURE 3. Assembled board.
August 2012 29