Introduction

www.ti.com

1.1.2 CPU Core

The C55x CPU is responsible for performing the digital signal processing tasks required by the application. In addition, the CPU acts as the overall system controller, responsible for handling many system functions such as system-level initialization, configuration, user interface, user command execution, connectivity functions, and overall system control.

Tightly coupled to the CPU are the following components:

DSP internal memories

Dual-access RAM (DARAM)

Single-access RAM (SARAM)

Read-only memory (ROM)

FFT hardware accelerator

Ports and buses

The CPU also manages/controls all peripherals on the device. Refer to the device-specific data manual for the full list of peripherals.

Figure 1-1shows the functional block diagram of the DSP and how it connects to the rest of the device. The DSP architecture uses the switched central resource (SCR) to transfer data within the system.

1.1.3 FFT Hardware Accelerator

The C55x CPU includes a tightly-coupled FFT hardware accelerator that communicates with the C55x CPU through the use coprocessor instructions. For ease of use, the ROM has a set of C-callable routines that use these coprocessor instructions to perform 8, 16, 32, 64, 128, or 256-point FFTs. The main features of the FFT hardware accelerator are:

Support for 8 to 1024-point (in powers of 2) real and complex-valued FFTs and IFFTs.

An internal twiddle factor generator for optimal use of memory bandwidth and more efficient programming.

Basic and software-driven auto-scaling feature provides good precision vs cycle count trade-off.

Single-stage and double-stage modes enabling computation of one or two stages in one pass, thus handling odd power of two FFT widths.

1.1.3.1Using FFT Accelerator ROM routines

The C5505 includes C-callable routines in ROM to execute FFT and IFFT using the tightly coupled FFT accelerator. The routines reside in the following address:

Table 1-1.

 

Address

Name

Description

Calling Convention

 

0x00ff6cd6

hwafft br

Vector bit-reversal

void hwafft_br( Int32 *data, Int32 *data_br, Uint16

 

 

 

 

data_len );

 

0x00ff6cea

hwafft 8pts

8-pt FFT/IFFT

Uint16 hwafft_8pts( Int32 *data,Int32 *scratch, Uint16

 

 

 

 

fft_flag, Uint16 scale_flag);

 

0x00ff6dd9

hwafft 16pts

16-pt FFT/IFFT

Uint16 hwafft_16pts( Int32 *data,Int32 *scratch, Uint16

 

 

 

 

fft_flag, Uint16 scale_flag);

 

0x00ff6f2f

hwafft 32pts

32-pt FFT/IFFT

Uint16 hwafft_32pts( Int32 *data,Int32 *scratch, Uint16

 

 

 

 

fft_flag, Uint16 scale_flag);

 

0x00ff7238

hwafft 64pts

64-pt FFT/iFFT

Uint16 hwafft_64pts( Int32 *data,Int32 *scratch, Uint16

 

 

 

 

fft_flag, Uint16 scale_flag);

 

0x00ff73cd

hwafft 128pts

128-pt FFT/IFFT

Uint16 hwafft_128pts( Int32 *data,Int32 *scratch,

 

 

 

 

Uint16 fft_flag, Uint16 scale_flag);

 

0x00ff75de

hwafft 256pts

256-pt FFT/IFFT

Uint16 hwafft_256pts( Int32 *data,Int32 *scratch,

 

 

 

 

Uint16 fft_flag, Uint16 scale_flag);

 

0x00ff77dc

hwafft 512pts

512-pt FFT/iFFT

Uint16 hwafft_512pts( Int32 *data,Int32 *scratch,

 

 

 

 

Uint16 fft_flag, Uint16 scale_flag);

 

0x00ff7a56

hwafft 1024pts

1024-pt FFT/IFFT

Uint16 hwafft_1024pts( Int32 *data,Int32 *scratch,

 

 

 

 

Uint16 fft_flag, Uint16 scale_flag);

 

 

 

 

 

 

 

 

 

 

14 System Control

 

 

SPRUFX5A –October 2010 –Revised November 2010

 

 

 

 

Submit Documentation Feedback

Copyright © 2010, Texas Instruments Incorporated

Page 14
Image 14
Texas Instruments TMS3320C5515 manual CPU Core, FFT Hardware Accelerator, Using FFT Accelerator ROM routines, Address

TMS3320C5515 specifications

The Texas Instruments TMS3320C5515 is a highly specialized digital signal processor (DSP) designed for a wide range of applications, including telecommunications, audio processing, and other signal-intensive tasks. As part of the TMS320 family of DSPs, the TMS3320C5515 leverages TI's extensive experience in signal processing technology, delivering robust performance and reliability.

One of the main features of the TMS3320C5515 is its 32-bit architecture, which allows for a high level of precision in digital signal computation. The processor is capable of executing complex mathematical algorithms, making it suitable for tasks that require high-speed data processing, such as speech recognition and audio filtering. With a native instruction set optimized for DSP applications, the TMS3320C5515 can perform multiply-accumulate operations in a single cycle, significantly enhancing computational efficiency.

The TMS3320C5515 employs advanced technologies including a Harvard architecture that separates instruction and data memory, enabling simultaneous access and improving performance. Its dual data buses enhance throughput by allowing multi-channel processing, making it particularly effective for real-time applications where timely data manipulation is critical. The device supports a wide range of peripherals, facilitating connections to various sensors and communication systems, which is vital in embedded applications.

In terms of characteristics, the TMS3320C5515 operates at an impressive clock speed, providing the computational power necessary to handle demanding tasks. The device is optimized for low power consumption, making it ideal for battery-operated applications without sacrificing performance. Its flexibility in processing algorithms also allows it to be readily adapted for specific requirements, from audio codecs to modems.

Another noteworthy aspect is the extensive development ecosystem surrounding the TMS3320C5515, which includes software tools, libraries, and support resources designed to accelerate the development process. This allows engineers and developers to bring their projects to market more quickly while minimizing risk.

Overall, the Texas Instruments TMS3320C5515 stands out as a powerful DSP solution, equipped with features that cater to the needs of various industries. Its combination of performance, efficiency, and versatile application makes it an attractive choice for engineers working in signal processing.