PBL 3786
polarity as Rxdet) at the CMP input.
If the microphone input signal contains
breaks as breath pauses and such, the
voltage at Txdet decreases. If the voltage
at any one time during a transmit period
becomes less than one half of the level
across C4 (absolute values) the circuit
will start a rapid charge of C4 towards
the reference voltage. The negative
charge of C4 caused by normal speech
because its many breaks does not affect
the speech switching. See fig. 17.
The activation of noise detection in
noiseless environment is prevented by a
treshold (50mV) in the Txdet.
When using the application according
to figure 19, pin -C is used as a negative
floating-point supply for the amplifier.
The output signal of the loudspeaker
amplifier is referred to +L. The capacitor,
C, makes it possible for the amplifier to
handle high peaks which are above its
constant-drive capability.
The optimal design, without using a
stable supply, is to balance the biasing
of the circuit against the DC characteris-
tics of the speech circuit working in
parallel. See fig. 20. The capacitor C is a
charge reservoir for the loudspeaker
amplifier.
Tone Ringer
The tone ringer drives directly a 50 ohm
loudspeaker without an output trans-
former. This is done by using short, high-
current pulses. A threshold in the
activation of the circuit prevents LD-
dialling from starting the ringing (anti-
tinkle). See fig. 25.
Basic mode: The ringing tone in the
basic mode is set by capacitors; the
main frequency at pin 13 and the shift
frequency, ƒS at pin 12 (see figure 26).
Smaller adjustments in the frequencies
can be made by a series resistor to the
capacitors. The approximate frequency
calculation formula is:
In the receive mode some of the
Function of the current feed
system for the audio power
amplifier
loudspeaker output signal will be sensed
by the microphone. In order not to treat
this as noise, the background noise
detector goes into a hold state thus rem-
embering the level of previous transmit
or standby period.
ƒ = 1/(1.3 • 10 6 • C)
Tone ringer sound output
The impedance in the speech circuit
transmitter output is low because of the
feedback loop. This low impedance
causes when a current is added to the
output the dc level is not affected and
the current through R is unchanged. If
this added current is taken from the line
VL, the line current IL is not influenced as
the current through R is not changed i.e.
the current through R is constant. See
fig. 21 and 23.
The volume control resistor (see fig. 26)
on pin 14 controls the width of the output
pulses, hence the energy to the louds-
peaker. The lower the resistance, the
wider the pulse and the higher the
energy.
CTR input
A voltage at this input (normally V+ - 600
mV) controls the comparators that set
the gain in the receive and transmit
channels continuously from full speech
control mode (50 dB attenuation
between channels) to duplex mode (both
channels fully open). Input left open
renders full speech control. See figures
14 and 16.
In full duplex mode, the CMP input is
clamped to the reference voltage in
order to avoid anything at this input to
interfere with the function.
If current is drained out of the CTR in-
put, the gain is reduced in both channels
channels, and a mute mode is achieved
(at approx. 35 - 40 µA). See fig. 16.
Decreasing the voltage at the input
below 0.9 V, a shut-down state is
reached, where the transmit and receive
amplifiers are switched off and the curr-
ent consumption is reduced. See fig. 18.
Microprocessor mode: The toneringer
circuit (see fig. 27) can be programmed
to interface a microprocessor. It will feed
the processor with power during the
presence of ringing signal on the line.
Two different configurations can be
used:
IL = IR = IB + IC
This added current + IC in our case is
the charge current to the capacitor C
and functions in the system as explained
above.
To set the minimum current needed
for undistorded function of the speech
circuit at the lowest line voltage IBmin is
done as follows:
The current through RE • 50•IE = IC is
the charge current of the capacitor C
under the charge condition. (logic state).
The control logic enables charge when
VC - VRE ≤ 0 and cuts the charge current
when VC - VRE > 0
1. With volume control at pin 14. The
TTLsgnal, 1-4 Vpp (referred to
groud) is fed into pin 13 through a
100 kΩ resistor. Pin 12 is connected
to pin 17 through a 1MΩ resistor to lift
it up. The DC supply to the processor
has to be solved in an alternative way.
2. Providing the processor with DC
power. The TTL signal 1-4 Vpp is fed
into pin 13 through a 100 kohm
resistor. A resistor to ground at pin 12
programs how much DC supply
current is available at pin 14. In this
case the output volume can be
optimized by connecting a capacitor
across the loudspeaker output, it
increases the output pulse width.
IC + IB = IR ICmax + IBmin = IR is the lowest
specified current the telephone must
work.
Loudspeaker Amplifier
The loudspeaker amplifier drives directly
the 50 ohm loudspeaker, which is also
used by the tone ringer. The amplifier is
designed to work under a number of
different biasing conditions.
IB
IR
IBmin
IBmin + ICmax
1 - 50R
R
=
=
The IBmin has to be calculated taking
into account both the quiscent dc. level
and the highest specified swing at the
lowest line voltage. x 50 is the internal
current mirror factor.
Note: The capacitor across the loud
speaker influences the frequency
characteristics of the power amplifier
for speech. The maximum DC supply
current that can be taken out of pin
14 is dependent of the amplitude and
impedance of the ringing signal on the
line.
The highest output swing is obtained if
pin -C is connected to ground and pin +L
is connected to a stable supply. The bia-
sing in that case could be provided either
from a mains supply or via a coil from the
telephone line. See figure 19. Current
consumption is directly proportional to
the voltage between pins +L and -C.
The single-ended loudspeaker
amplifier has an internal gain regulation
that prevents distortion in case when not
enough current from the line is available.
12
ERICSSON [ ERICSSON ]
