#2 Check out application note AN543
on the Microchip website. It explains
general-purpose tone generation and
offers assembler source for the
PIC17C42 which I'm sure you can get to
run on your 18F part.
Santa Ana, CA
[#4102 - April 2010]
Car Autoranging Ammeter
Need a circuit for an autoranging car
ammeter. In-line cable shunt output at
the battery. Resistor short protection at
the battery for a pair of sense wires
running to the circuit and meter under
the dash. Car battery powered, no
microprocessors, low range 0 to 10 amps,
high engine start range for starting
current, d’Arsonval meter, not digital.
Want to see relatively fast amperage
changes. Should this go in the negative or
positive side of the battery? Accuracy on
0-10 amp range say 10% or better; 20%
or better on the engine start range. KISS.
The simplest solution would be to
use two shunts, two sets of short-circuit
protectors, and two d'Arsonval meters.
The two shunts would be rated something on the order of 20 amperes and
200 amperes full scale, or as appropriate.
I'd use 50 mV shunts (rather than 100
mV shunts) because the alternator
output voltage will be regulated with
reference to the load side of the low
current shunt and this introduces only a
0.4% error in actual battery charging
voltage. Both meters would be rated
50 mV full scale, but the low current
meter would need to be a zero-center
instrument (- 50 mV to + 50 mV).
However, one instrument with the
ability to switch between two current
ranges was requested. I believe that the
scheme illustrated in Figure 1 addresses
that challenge. The battery, ignition
switch, starter, and starter relay are
considered to be pre-existing equipment
in the car. A negative-ground system is
The two shunts are as above,
delivering 50 mV for full-scale ratings of
20 amperes and 200 amperes, or as
appropriate. They are located at or very
close to the positive post of the battery.
The low current shunt is inserted into the
existing lead that connects to all of the
non-starter loads in the car, plus the
alternator. The high current shunt is
inserted into the non-grounded starter
Fuses F1-F4 are 1/4" diameter
0.1 ampere rapid-acting devices (type
AGC or GBB) contained within in-line
weatherproof fuseholders. Resistors
were specified, but resistors affect the
voltage delivered to the indicating
instrument, and with only 50 mV to play
with full scale, any loss is significant.
Fuses are better.
M1 is a 50 mV (full-scale) zero-center instrument of appropriate size
and accuracy with its scale hand-calibrated to match the shunts.
K2 is a fast-acting three-pole double-throw insrument relay having gold
contacts and a 12 volt DC coil. Gold
crossbar contacts would be the best.
We're switching millivolt signals and
voltage barriers due to dry-contact
phenomena cannot be tolerated.
K1 is a sensitive single-pole relay
(one normally-open contact is required)
having a low current 12 volt DC coil. It
exists because the contacts in K2 are
small and the starter relay is an unknown
device, so K1 exists to switch the load
presented by the starter relay. It also
adds delay which favors the circuit.
In operation, K2 is normally not
energized hence M1 is connected across
the low current shunt. When the car
engine is to be started, the operator
rotates the ignition switch to the START
position which energizes relay K2. The
uppermost set of contacts on K2 close
and energize relay K1; and through K1,
the starter relay; and through the starter
relay, the starter motor.
But at the same time that K2 causes
K1 to be energized, its remaining two
sets of contacts have likewise transferred
such that meter M1 is now connected
across the high current shunt, hence M1
will monitor the starter motor current
once it is energized.
When the engine begins to run, the
operator releases the ignition switch key,
breaking the connection to the START
Since the START contact carries the
full load of the starter relay coil — and
also the coils of K1 and K2 — the starter
motor will be disconnected rapidly and
before the contacts on K2 transfer back
to their de-energized position, and thus
M1 can be connected safely to the low
Peter A. Goodwin