English
Language : 

MC33219A Datasheet, PDF (20/28 Pages) Motorola, Inc – Voice Switched Speakerphone
Freescale SMeCm33i2c1o9Anductor, Inc.
APPLICATIONS INFORMATION
Switching and Response Time Theory
The switching time of the MC33219A circuit is dominated
first by the components at CT (Pin 7, see Figure 2), and
second by the capacitors at the level detector outputs (RLO,
TLO).
The transition time to receive or to transmit mode from
either idle or the other mode is determined by the capacitor
at CT, along with the internal current sources (refer to
Figure 25). The switching time is:
+ DT
DV CT
I
When switching from idle to receive, ∆V = 150 mV,
I = 90 µA, the CT capacitor is 15 µF, and ∆T calculates to
≈ 25 ms. When switching from idle to transmit, ∆V = 100 mV,
I = 50 µA, the CT capacitor is 15 µF, and ∆T calculates to
≈ 30 ms.
When the circuit switches to idle, the internal current
sources are shut off, and the time constant is determined by
the CT capacitor and RT, the external resistor (see
Figure 25). With CT = 15 µF, and RT = 15 kΩ, the time
constant is ≈ 225 ms, giving a total switching time of ≈ 0.68 s
(for 95% change). The switching period to idle begins when
both speakers have stopped talking. The switching time back
to the original mode will depend on how soon that speaker
begins speaking again. The sooner the speaking starts
during the “decay to idle” period, the quicker the switching
time, since a smaller voltage excursion is required. That
switching time is determined by the internal current sources
as described above.
When the circuit switches directly from receive to
transmit (or vice versa), the total switching time depends
not only on the components and currents at the CT pin, but
also on the response of the level detectors, the relative
amplitude of the two speech signals, and the mode of the
circuit, since the two level detectors are connected
differently to the two attenuators.
The rise time of the level detector’s outputs (RLO, TLO) is
not significant since it is so short. The decay time, however,
provides a significant part of the “hold time” necessary to
hold the circuit (in transmit or receive) during the normal
pauses in speech. The capacitors at the two outputs must
be equal value (±10%) to prevent problems in timing and
level response.
The components at the inputs of the level detectors (RLI,
TLI) do not affect the switching time, but rather affect the
relative signal levels required to switch the circuit, as well as
the frequency response of the detectors. They must be
adjusted for proper switching response as described later in
this section.
Switching and Response Time Measurements
Using burst of 1.0 kHz sine waves to force the circuit to
switch among its modes, the timing results were measured
and are indicated in Figures 17–21.
a) In Figure 17, when a signal is applied to the transmit
attenuator only (normally via the microphone and the
microphone amplifier), the transmit background noise
monitor immediately indicates the “presence of speech” as
evidenced by the fact that CPT begins rising. The slope of
the rising CPT signal is determined by the external resistor
and capacitor on that pin. Even though the transmit
attenuator is initially in the idle mode (–16 dB), there is
sufficient signal at its output to cause TLO to increase. The
attenuator control circuit then forces the circuit to the
transmit mode, evidenced by the change at the CT pin. The
attenuator output signal is then 6.7 dB above the input.
With the steady sine wave applied to the transmit input,
the circuit will stay in the transmit mode until the CPT pin gets
to within 36 mV of its final value. At that point, the internal
comparator (see Figure 27) switches, indicating to the
attenuator control circuit that the signal is not speech, but
rather it is a steady background noise. The circuit now begins
to decay to idle, as evidenced by the change at CT and TLO,
and the change in amplitude at TAO.
When the input signal at TAI is removed (or reduced), the
CPT pin drops quickly, allowing the circuit to quickly respond
to any new speech which may appear afterwards. The
voltage at CT decays according to the time constant of its
external components, if not already at idle.
The voltage change at CP2, CPT, and TAO depends on
the input signal’s amplitude and the components at XDI and
TLI. The change at CT is internally fixed at the level shown.
The timing numbers shown depend both on the signal
amplitudes and the components at the CT and CPT pins.
b) Figure 18 indicates what happens when the same signal
is applied to the receive side only. RLO and CPR react
similarly to TLO and CPT. However, the circuit does not switch
to idle when CPR finishes transitioning since the dial tone
detector disables the background noise monitor, allowing the
circuit to stay in the receive mode as long as there is a signal
present. If the input signal amplitude had been less than the
dial tone detector’s threshold, the circuit response would have
been similar to that shown in Figure 17. The voltage change
at CT depends on the setting of the volume control (Pin 19).
The 150 mV represent maximum volume setting.
c) Figure 19 indicates the circuit response when transmit
and receive signals are alternately applied, with relatively short
cycle times (300 ms each) so that neither attenuator will begin
to go to idle during its “on” time. Figure 20 indicates the circuit
response with longer cycle times (1.0 s each), where the
transmit side is allowed to go to idle. Figure 21 is the same as
Figure 20, except the capacitor at CT has been reduced from
15 µF to 6.8 µF, providing a quicker switching time. The
reactions at the various pins are shown. The response times at
TAO and RAO are different, and typically slightly longer than
what is shown in Figures 17 and 18 due to:
– the larger transition required at the CT pin,
– the greater difference in the levels at RLO and TLO due
to the positions of the attenuators as well as their decay
time, and
– response time of the background noise monitors.
The timing responses shown in these three figures are
representative for those input signal amplitudes and burst
durations. Actual response time will vary for different signal
conditions.
NOTE: While it may seem desirable to decrease the
switching time between modes by reducing the capacitor at
CT, this should be done with caution for two reasons:
1) If the switching time is too short, the circuit response
may appear to be “too quick” to the user, who may consider
its operation erratic. The recommended values in this data
sheet, along with the accompanying timings, provide what
20
For MorGeoIntfoo:rwmwatwio.fnreOenscTahlies.cPoromdMuOcTtO, ROLA ANALOG IC DEVICE DATA