DSP is widely used in many fields because of its fast processing speed, low power consumption, good performance, large voice storage capacity based on TMS320C5402DSP chip and good communication quality.

The speech analysis system realized in this design has the following advantages:

1. Audio data takes up less resources.

2. High-quality communication level

3. The development difficulty is low

4. The interface circuit between voice chip and DSP is simple.

5. Small size

In the process of completing the thesis, I first consulted the relevant books in the library, studied how to design a voice recorder based on TMS320C5402DSP chip, and then designed various modules needed by the system, and made a detailed study on the chip. Secondly, design, program and debug application software on the computer and experimental board with reference to relevant materials, and then carry out joint debugging of software and hardware under the guidance of teachers; Finally, I sorted out and summarized the graduation design materials and finished my graduation design thesis.

In the whole design process, this paper first introduces the working principle of voice recording and playback system based on TMS320C5402DSP chip, gives the overall design scheme and working block diagram, and then gives the hardware design scheme of the system; In the hardware design, we use TLV320AIC23DSP chip as the core audio recording and playback interface device, combined with TMS320C5402DSP chip, voice data storage FLASH memory and so on to basically complete the hardware design process of voice recording and playback; Finally, the software design of voice recording and playback system based on TMS320C5402DSP chip is introduced. The software part is mainly programmed in C language under CCS environment. The analog voice signal input from the outside is sampled by the high-fidelity voice chip AIC23 and stored in the storage space of the external expansion memory. Then these stored digital speech signals are read into DSP through the buffer serial port MCBSP2 of DSP, and the high-frequency parts and other noises in the speech signals are filtered by the FIR digital low-pass filter. Finally, these speech signals are transformed by FFT.

The design of this speech analyzer can complete the collection, playback, storage and spectrum analysis of speech, and basically realize the speech analysis function. With the development of technology, the speech coding scheme combining TMS320C5402DSP and TLV320AIC23 will have a better application prospect.