ts34118 Taiwan Semiconductor Company, Ltd. (TSC), ts34118 Datasheet - Page 15

• Manufacturer Part Number: ts34118
• Description: Voice Switched Speakerphone Circuit
• Manufacturer: Taiwan Semiconductor Company, Ltd. (TSC)

TS34118
Design Equations
2) To provide a reasonable signal level to the TLI2 level detector; and 3)to minimize any gain applied to broadband
II) Receive Gain
The overall receive gain depends on the incoming signal level, and the desired output power at the speaker. Normal
receive levels (independent of the peaks) at Tip/Ring can be 35mVrms (-27dBm). Although on long lines That level can
be down to 8.0mVrms (-40dBm).The speaker power is:
Where R
path. Experience has shown that§30dB gain is a satisfactory amount for the majority of applications. Using the above
numbers and Equation 1, it would appear that the resulting power to the speaker is extremely low. However, Equation 1
does not consider the peaks in normal speech, which can be 10 to 15 times the rms value. Considering the peaks, the
overall average power approaches 20-30mV on long lines, and much more on short lines.
Referring to Figure 10, the gain from Tip/Ring to the filter input was measured at 0.833V/V (-1.6dB), the filter’s gain is
unity, and the receive attenuator’s gain is 2.0V/V (+6.0dB) at maximum volume. The speaker amplifier’s gain is set at
22V/V (26.8dB), which puts the overall gain at §31.2dB.
III) Loop Gain
The total loop gain must add up to less than zero dB to obtain a stable circuit. This can be expressed as:
Using the typical numbers mentioned above, and knowing that G
determined:
An acoustic loss of at least 23 dB is necessary to prevent instability and oscillations, commonly referred to as “singing.”
However, the following equations show that greater acoustic loss is necessary to obtain proper level detection and
switching.
IV) Switching Threshold
To switch comparator C1, currents I1 and I3 need to be determined. Referring to Figure 11, with a receive signal V
applied to Tip/Ring, a current I3 will flow through R3 into RLI2 according to the following equation:
Where the terms in the brackets are the V/V gain terms. The speaker amplifier gain is divided by two since G
differential gain of the amplifier, and V3 is obtained from one the side of the output. The current I1, coming from the
microphone circuit, is defined by:
Where V
equations yields:
noise generated within the attenuator. However, to cover the normal voice band, the microphone amplifier’s gain
should not exceed 48dB. For the circuit of Figure 10, the gain of the microphone amplifier was set at 35V/V (31dB),
and the differential gain of the hybrid amplifiers was set at 10.2V/V (20.1dB).
G
G
MA
AC
PSPK =
I3 =
I1 =
V
<-[31+20.1+(-15)+0+(-40)+26.8]=-22.9dB (Equation 3)
+G
M
S
M
TX
is the speaker impedance, and the dBm term is the incoming signal level increased by the gain of the receive
= V
is the microphone voltage. Since the switching threshold occurs when I1=I3, combining the above two
V
R3
V
+G
L
M
L
10dBm/10u
HA
u
R1
×
ª
«
¬
+G
G
G
ST
MA
HR
R3
R1
+G
Rs
u
FO
G
(Equation 5)
>
+G
G
FO
0.6
HR
RX
u
+G
G
u
RX
SA
G
G
+G
FO
u
MA
AC
G
u
u
2
<0
SA
G
2
RX
º
»
¼
15-20
G
(Equation 1)
(Equation 4)
(Equation 2)
SA
@
(Equation 6)
TX
+G
RX
=-40dB, the required acoustic coupling can be
2004/09 rev. B
SA
is the
L