manufacturer who has the aqueous
system to clean it off the boards. For
us home shop workers and hobbyists,
our best choices are the old
fashioned rosin-based fluxes such as
Kester 44 or one of the "no wash"
solders that are available.
I have a question about a circuit
in "Herding Data Over Bridges" in the
February 2013 Design Cycle. In
Schematics 3 and 4, Fred Eady details
a PIC18F46J13, and there is a
separate crystal (X2, 32.768 kHz) to
drive the internal RTCC. The PICs I
use in my circuits (PIC16F887) are
much more primitive, but when using
Timer1 to provide an accurate
timekeeping function, I have found
that linking it to the 4 MHz main
oscillator crystal is just as accurate as
using a separate 32.768 kHz crystal.
One clock I designed only gained
one second over 12 months (inside
my very climate-controlled house)!
Why do you include the second
crystal? Does the RTCC of that PIC
not have any option to run off of the
Judy May W1ORO
Hi Judy! Thanks for reading
The RTCC component of the
PIC18F46J13 was designed to run
independent of the core processor.
That means the RTCC can continue to
run and keep accurate time, even if
the microcontroller is sleeping.
To be able to do this, the RTCC
clock source must be a stand-alone
The PIC18F46J13's RTCC can be
clocked at 32.768 kHz by placing a
32.768 kHz crystal across the Timer1
T1OSCO and T1OSCI pins.
The RTCC can also be clocked by
an internal RC oscillator (INTRC). The
RTCC clock source is selected by the
RTCOSC configuration bit
CONFIG3L<1>. There is no provision
to clock the RTCC from the CPU
When the INTRC clocks the
RTCC, the tolerances of the internal
RC oscillator determine the stability
of the RTCC timing. The INTRC clock
frequency is stated to be around 31
kHz. You will find that the crystal
clock option is a better choice for
accurate long-term timekeeping. To
assure maximum timing accuracy in
crystal mode, a calibration register
(RTCCAL) is provided for fine tuning
and can adjust the RTCC timing up to
± 2. 64 seconds per month.
PICing My Brain
I have a question about Thomas
Henry’s circuit in "PIC Trainer from
Surplus Parts" in the February 2013
issue. I love the article! Mr. Henry
has done a thorough job coming
up with such a wide array of
components that can be connected
to the PIC. With regards to the serial
port, a series protection resistor has
been included. I do not understand
how this would protect the
microcontroller from the possible
+ 12 volt and - 12 volt signals. The
resistor will limit the current, but
wouldn't the microcontroller pin still
be exposed to voltage that far
surpasses the absolute maximum
ratings, risking damage to the MCU?
Judy May W1ORO
A quick check of the PIC
datasheet shows that most of the pins
feature input protection diodes which
will effectively clamp the voltages.
The series resistor limits the current
through them. I've never had any
trouble with this scheme when
sensing RS-232 signals.
52 October 2013
READER FEEDBACK Continued from page 7