Start
• It's All
in the Grounding • Transient
Voldatge Supression • Modem
and Power-Line Protection • Unplug
the Computer! •
Automatic Thunderstorm
Switch
• Black Box
It
•
Waiting for Summer
•
Sources
BLACK
BOX IT
The
sample data distributed with the sensor demonstrated
that the effective number of hits per minute the sensor
picked up during a thunderstorm ranged from 30 to
well over 300 (within a 100-mile radius, I suppose
this is acceptable).
Personally,
I’d be heading for the cellar if I saw an indication
of a 300 hits-per-minute storm, but Jeff was fascinated
at knowing the actual quantity. He included eight
LEDs to provide a visual display of the hits per minute
as a power of two.
Using
this method, the first LED indicates two hits; the
second, four hits; the third, eight hits; and so on.
The last LED indicates 256 or greater hits per minute.
The LEDs are off under clear-sky conditions.
To
count hits from the sensor, Jeff used the T0CK1 input
on a PIC16C54 microcontroller. Figure 3 shows the
circuitry for this simple display. Besides the eight
directly driven LEDs and T0CK1, there are three bits
for configuration and a single-bit alarm output. The
configuration bits choose how the alarm output functions.
|
Figure 3—The optical pulses are received from the lightning sensor
and converted to a hits-per-minute LED indicator.
|
The
output can be a 250-ms momentary pulse or continuously
low during an alarm condition. The alarm trigger point
is selected using the first two configuration bits.
The four combinations select the hits-per-minute turn-on
point of any of the upper four LEDs as its trigger
level.
The
software’s main loop contains a 3-s counting
period, followed by a total of the last 20 periods
(total over the last 1 min.). The total counts are
transposed into a byte with the proper bit high to
enable an LED indicating the appropriate range.
Because
you want to know if the storm is moving toward or
away from your location, it’s important to know
about the past. Therefore, the PEAK count is displayed
as a steady-state LED.
The
PRESENT count is indicated by XORing the present count
with the LEDs such that if PEAK and PRESENT are the
same (as in a storm moving toward you), the LED flashes.
However,
once PRESENT starts dropping, the peak-count LED remains
steady and the present-count LED flashes. Now, you
can tell immediately which direction the storm is
heading. The PEAK value can be reset by pressing the
reset button.
Although
Jeff’s software only takes up about a quarter
of the 16C54’s available code space, all of its
registers are used. The majority are taken up by the
20 table entries doing the 3-s counting samples.
Only
five other registers are used by the code for the
rest of the functions. When no LEDs are on, the circuit
requires only about 3 mA (add about 10 mA per LED).
Obviously,
this whole circuit could operate from three alkaline
batteries, but if it’s mounted where you plan
to view the thunderstorm’s progress, trickle
charging four NiCd batteries would be better. After
all, when you need the information most to either
pull the plug or tell you that conditions are all
clear, you don’t want to depend on a tired set
of batteries.
Figure
4 is the switch’s connect/disconnect section.
This particular configuration accommodates AC power,
coax, and phone lines.
|
Figure 4—The hits-per-minute converter also generates a "lightning
alarm" output. This signal causes the
circuit to physically disconnect the AC power,
cable, and phone-line connections to the protected
appliance.
|
The
16C54’s alarm output (set for steady-state output
mode) drives an optoisolated triac switch controlling
the AC power relay. The circuit’s normal condition
is for the alarm output to be high and the relays
energized. A small DC power supply, connected in parallel
with the AC power relay coil, controls the coax and
phone relays.
The
switches’ output connections are made to the
normally closed contacts. When the power is off or
an alarm condition exists, the relays are deenergized.
This connects the equipment side of things to ground,
where it provides the greatest protection.
While
the normal spacing of the relay contacts is less than
¼², should a high-voltage surge enter the line side
of the switching unit, any place it arcs within the
relay will be at ground. Certainly, if you incorporate
MOV suppression in addition, such voltage levels shouldn’t
even exit at the relays.
We
didn’t include those MOVs and surge-suppression
components on the schematic, but we indicated their
proper placement. If you’re already using protection
devices on these lines, you may not need to include
them inside as well.
