English
Language : 

MC33219A Datasheet, PDF (22/28 Pages) Motorola, Inc – Voice Switched Speakerphone
Freescale SMeCm33i2c1o9Anductor, Inc.
1) Design the hybrid, ensuring proper interface with the
phone line for both DC and AC characteristics. The return
loss must be adjusted to comply with the appropriate
regulatory agency. The sidetone should then be adjusted
according to the intent of the product. If the product is a
speakerphone only (without a handset), the sidetone gain
(GST) should be adjusted for maximum loss. If a handset is
part of the end product, the sidetone must be adjusted for the
minimum acceptable sidetone levels in the handset.
Generally, for the speakerphone interface, 10–20 dB
sidetone loss is preferred for GST.
2) Check the acoustic coupling of the enclosure (GAC in
Figure 31). With a steady sound coming out of the speaker,
measure the rms voltage on the speaker terminals and the rms
voltage out of the microphone. Experience has shown that the
loss should be at least 40 dB, preferably 50 dB. This should
be checked over the frequency range of 20 Hz to 10 kHz.
3) Adjust the transmit path for proper signal levels, based
on the lowest speech levels as well as the loudest. Based on
the typical levels from commonly available microphones, a
gain of about 35–45 dB is required from the microphone
terminals to Tip and Ring. Most of that gain should be in the
microphone amplifier to make best use of the transmit
attenuator, but the maximum input level at TAI must not be
exceeded. If a signal generator is used instead of a
microphone for testing, the circuit can be locked into the
transmit mode by grounding CPT (Pin 3). Frequency
response can generally be tailored with capacitors at the
microphone amplifier.
4) Adjust the receive path for proper signal levels based on
the lowest speech levels as well as the loudest. A gain of
about 30 dB is required from Tip and Ring to the speaker
terminals for most applications (at maximum volume). Most
of that gain should be in the receive amplifier (at RXI, RXO) to
make best use of the receive attenuator, but the maximum
input level at RAI must not be exceeded. If a signal generator
is used for signal injection during testing, the circuit can be
locked into the receive mode by grounding CPR (Pin 10),
although this is usually not necessary since the dial tone
detector will keep the circuit in the receive mode. Frequency
response can generally be tailored with capacitors at the
receive amplifier.
5) Check that the loop gain (i.e., the receive path gain +
acoustic coupling gain + transmit path gain + sidetone gain)
is less than 0 dB over all frequencies. If not, “singing” will
occur: a steady oscillation at some audible frequency.
6) a) The final step is to adjust the resistors at the level
detector inputs (RLI and TLI) for proper switching response
(the switchpoint occurs when I1 = I2). This has to be the last
step, as the resistor values depend on all of the above
adjustments, which are based on the mechanical, as well as
the electrical, characteristics of the system. NOTE: An
extreme case of level detector misadjustment can result in
“motorboating”. In this condition, with a receive signal
applied, sound from the speaker enters the microphone, and
causes the circuit to switch to the transmit mode. This causes
the speaker sound to stop (as well as the sound into the
microphone), allowing the circuit to switch back to the receive
mode. This sequence is then repeated, usually, at a rate of a
few Hz. The first thing to check is the acoustic coupling, and
then the level detectors.
b) Starting with the recommended values for R1 and R2 (in
Figure 2), hold a normal conversation with someone on
another phone. If the resistor values are not optimum, one of
the talkers will dominate, and the other will have difficulty
getting through. If, for example, the person at the
speakerphone is dominant, the transmit path is overly
sensitive, and the receive path is not sensitive enough. In this
case, R1 (at TLI) should be increased, or R2 (at RLI)
decreased, or both. Their exact value is not critical at this
point, only their relative value. Keeping R1 and R2 in the
range of 2.0–20 k, adjust them until a suitable switching
response is found.
c) Then have the person at the other end of the phone line
speak loud continuously, or connect to a recording which is
somewhat strong. Monitor the state of the circuit (by
measuring the CT versus VB pins, and by listening carefully to
the speaker) to check that the sound out of the speaker is not
attempting to switch the circuit to the transmit side (through
acoustic coupling). If it is, increase R1 (at TLI) in small steps
just enough to stop the switching (this desensitizes the
transmit side). If R1 has been changed a large amount, it may
be necessary to readjust R2 for switching response. If this
cannot be achieved in a reasonable manner, the acoustic
coupling is too strong.
d) Next, have the person at the speakerphone speak
somewhat loudly, and again monitor the state of the circuit,
primarily by having the person at the other end listen carefully
for fading. If there is obvious fading of the sound, increase R2
so as to desensitize the receive side. Increase R2 just
enough to stop the fading. If this cannot be achieved in a
reasonable manner, the sidetone coupling is too strong.
e) If necessary, readjust R1 and R2 a small amount
relative to each other, to further optimize the switching
response.
Transmit/Receive Detection Priority
Although the MC33219A was designed to have an idle
mode such that the transmit side has a small priority (the idle
mode position is closer to the full transmit side), the idle mode
position can be moved with respect to the transmit or the
receive side. With this done, the ability to gain control of the
circuit by each talker will be changed.
By connecting a resistor from CT (Pin 7) to ground, the
+ ƪ * ƫ circuit will be biased more towards the transmit side. The
resistor value is calculated from:
R
RT
VB
DV
1
where R is the added resistor, RT is the resistor normally
between Pins 6 and 7 (typically 15 kΩ), and ∆V is the desired
change in the CT voltage at idle.
By connecting a resistor from CT (Pin 7) to VCC, the circuit
+ ƪ * ƫ will be biased towards the receive side. The resistor value is
calculated from:
R
RT
VCC –VB
DV
1
R, RT, and ∆V are the same as above. Switching response
and the switching time will be somewhat affected in each
case due to the different voltage excursions required to get to
transmit and receive from idle. For practical considerations,
the ∆V shift should not exceed 50 mV.
Disabling the Idle Mode
For testing or circuit analysis purposes, the transmit or
receive attenuators can be set to the ON position, even with
steady signals applied, by disabling the background noise
monitors. Grounding the CPR pin will disable the receive
background noise monitor, thereby indicating the “presence
22
For MorGeoIntfoo:rwmwatwio.fnreOenscTahlies.cPoromdMuOcTtO, ROLA ANALOG IC DEVICE DATA