Abstract: In a handsfree telephone system of the type described in our application No. 584,579, filed in the U.S. on Feb. 29, 1984, the method used to avoid howling due to feedback from the loudspeaker to the microphone is to increase the attenuation to speech in the quiescent direction as compared with that in the active direction. This involves monitoring successive speech samples in the two directions with each sample compared with a preset threshold, which also takes account of the system's noise levels. This is effected under processor control, as in the adjustment of the attenuation, since the system uses digital speech transmission.It has now been found that it is not necessary to monitor all speech samples to derive adequate information about the occupancy conditions of the two speech directions. In the present systems only one speech sample in four is used. This enables one microprocessor to serve two lines alternately.

Abstract: In a loudspeaking telephone arrangement, there are separate channels for outgoing speech (A-Tx) and incoming speech (A-Rx), and the channels are sampled at intervals. The results of these samplings, which represent the amplitudes of the speech in those channels are each compared by a microprocessor with a preset threshold. The background noise level is also sampled and the result used to adjust the speech channel threshold. Each channel has an attenuator, and that is adjusted so that attenuation is reduced in the presence of speech and increased in the absence thereof. The adjustment on the basis of background noise enables the current state of the channels to be taken into account.