Preliminary

AS2520/21/20B/21B

Rev. 5.1

Page 7

May 1999

This approach has some disadvantages. It requires a

high degree of discipline, since the three state

channel control gives a very distinct half duplex with

a relative high switching time constant to avoid

chopper effects. Furthermore, the system is very

sensitive to the environment,- noise, line conditions

and acoustics (echo).

Apart from keeping a distinct discipline, the user can

not do anything to minimise the effect of these

constraints, since the parameters of the voice

switching

(thresholds,

time

constants,

noise

discrimination, etc.) can not be changed or adapted

to the actual conditions by the user.

The dynamic voice control system of the AS252x

have

been

designed

to

overcome

the

above

constraints. The basic philosophy behind the AS252x

is that telephone circuits should not have any

automatic regulations preventing the user from

having all information about the actual conditions

which should enable her/him to act accordingly, i.e.

to comply with the given constraints.

Now, assuming subscriber A has a handsfree

telephone and is calling subscriber B, who has a

normal telephone. The B subscriber does not

necessarily know that A is using a handsfree

telephone and will therefore not automatically comply

to the discipline of a half duplex conversation. Hence,

the disadvantages by using half duplex should apply

to the A subscriber only.

Secondly, if A is in a noisy environment, the B

subscriber should hear it, so that he speaks up to

increase the signal-to-noise ratio at the A subscriber.

The traditional 3-state switching system has two

major drawbacks: first of all, when no one is talking,

the circuit is in idle state and the B subscriber gets

the feeling that the line is dead, since the background

noise

does

not

activate

the

voice

switching.

Secondly, the B subscriber does not speak up, since

she/he does not hear the background noise.

The concept of the AS252x, however, does not

exclude

the

human

factor,

but

provides

the

information about the actual conditions to the user

and allows her/him to act accordingly, i.e. to speak

up, to change the volume, etc.

In more technical terms, the AS252x works in the

following manner:

When no signal is applied neither from the line nor

from the microphone, the circuit is in the only static

state, which is transmit channel full open and receive

channel attenuated by up to 30 dB.

The advantages of using the transmit state as the

static (idle) state are that the B subscriber hears an

open line (the line is not dead), does not miss the

initial word of a sentence when the A subscriber

starts talking, and hears the level of the background

noise at A´s end which will actuate her/him to speak

up accordingly.

When the A subscriber starts talking, the circuit

remains in the static state.

The dynamic state of the voice switching can only be

activated by the receive signal. Applying a receive

signal above a certain level will cause the circuit to

enter the dynamic state.

VOL

± 10 dB

AGC

PEAK

DETECTOR

2/4

Figure 5: Channel Control System

The signal for controlling the channel attenuation is

taken after the sidetone amplifier. With the volume at

0 dB (neutral) the threshold for entering the dynamic

state (V

is 15 mV assuming that V

figure 5).

In the dynamic state the channel attenuation is

controlled by a voltage controlled amplifier. The

attack time is 4 ms/6 dB and the hold time is 200 ms.

A speech compression is activated when a transmit

signal with a high amplitude reaches a level

corresponding to approximately 460 mV on the line.

1.50

1.25

1.00

0.75

0.50

0.25

0.00

0

50

100

150

200

250

300

Microphone Input Signal

(mV)

(mV)

Sidetone Cancellation: 11 dB

Volume Control: 0 dB (neutral)

Figure 6: Speech Compression