Recently, speech-coding are widely used in many practical areas; such as communications, medical care and education. The potentials offered by VLSI enable us to integrate many complex signal processing system on a single chip like VOCODER.

In this work, the play back synthesis part of an LPC VOCODER for Arabic language is implemented on a VLSI chip using VHDL entry system. The synthesis is achieved through a direct LPC filter with periodic pulse excitation for voiced frames and random excitation for unvoiced frames. The design has been optimized for efficient use of the implementation capacity and the quality of the synthesized sounds are checked for subjective quality and intelligibility.