Source Core

Table 4-8:SPI-4.2 Control Word Mapping to 64-bit User Interface

R

 

Associated SPI-4.2 Lite Control

 

Control Word

Word bits on TDat (Qualified by

Associated Source FIFO Signal(s)

 

TCtl=1)

 

 

 

 

Start of Packet (SOP)

TDat[15] =1, TDat[12]=1,

SrcFFSOP, SrcFFAddr[7:0]

 

TDat[11:4] <== SrcFFAddr[7:0]

 

 

 

 

New Burst (address change

TDat[15] = 1, TDat[12] = 0,

SrcFFAddr[7:0]

without SOP)

TDat[11:4] <== SrcFFAddr[7:0]

 

 

 

 

 

 

End of Packet

TDat[14:13] = 10

SrcFFEOP, SrcFFMOD[2:0]

(EOP, even bytes valid)

 

When TDat[14:13] = 10:

 

 

MOD = 000 if data bits 63-0 have valid data

 

 

MOD =110 if data bits 63-16 have valid data

 

 

MOD =100 if data bits 63-32 have valid data

 

 

MOD = 010 if data bits 63-48 have valid data

 

 

 

End of Packet

TDat[14:13] = 11

SrcFFWEOP & SrcFFWMod[2:0]

(EOP, odd bytes valid)

 

When TDat[14:13] = 11:

 

 

MOD = 111 if data bits 63-8 have valid data

 

 

MOD = 101 if data bits 63-24 have valid data

 

 

MOD = 011 if data bits 63-40 have valid data

 

 

MOD = 001 if data bits 63-56 have valid data

 

 

 

End of Packet

TDat[14:13] = 01

SrcFFErr, SrcFFEOP,

(EOP, abort, error condition)

 

SrcFFMOD[2:0]

 

 

 

Transmitting Training Patterns

Training patterns are transmitted at startup (after reset) until the core acquires synchronization on the FIFO Status Channel. Subsequently, if the parameter DataMaxT or AlphaData are not zero, the core will transmit AlphaData training patterns at least every DataMaxT cycles.

The core continuously monitors the number of data cycles since the transmission of the last training pattern. Once a DataMaxT interval of SPI-4.2 bus cycles has completed, the current transfer is terminated on the next burst boundary, and training patterns will be transmitted on the SPI-4.2 bus (AlphaData number of times). Once the training patterns have completed, the SPI-4.2 Lite core will resume transmission of data on the data bus.

The control signal TrainingRequest (see Table 2-11) is provided for you to request that training patterns be sent out of the Source SPI-4.2 interface. When the TrainingRequest signal is asserted, the transmission of data is halted on the next burst boundary and training patterns are transmitted on the SPI-4.2 Interface.

If the static configuration signal AlphaData[7:0] (see Table 2-15) is set to zero, and the TrainingRequest signal is asserted, the Source core will transmit a complete training pattern sequence. The core will continue to transmit training patterns until TrainingRequest is deasserted. When it is deasserted, the core will halt transmission of training patterns after the current sequence is complete.

If the static configuration signal AlphaData[5:0] is set to a non zero value, the Source core sends the number of training patterns defined by AlphaData every time it detects a rising edge on the TrainingRequest signal.

SPI-4.2 Lite v4.3 User Guide

www.xilinx.com

81

UG181 June 27, 2008

Page 81
Image 81
Xilinx UG181 manual Transmitting Training Patterns

UG181 specifications

Xilinx UG181 refers to the User Guide for the Xilinx 7 Series FPGAs, which offers a comprehensive overview of the architecture, capabilities, and features of these powerful field-programmable gate arrays (FPGAs). Designed to cater to a wide range of applications, Xilinx 7 Series FPGAs are widely adopted in industries such as telecommunications, automotive, aerospace, and consumer electronics.

One of the main features of the Xilinx 7 Series FPGAs is their use of advanced 28nm technology, which enables them to achieve high performance while maintaining low power consumption. This fine process technology not only ensures better power efficiency but also allows for increased logic density. The 7 Series includes several families, such as Artix-7, Kintex-7, and Virtex-7, each tailored for specific application demands ranging from cost-sensitive solutions to high-performance data processing.

Xilinx 7 Series FPGAs also incorporate a rich set of programmable logic resources. This includes Look-Up Tables (LUTs), Flip-Flops, and Digital Signal Processing (DSP) slices that have been optimized for various arithmetic functions. With several thousands of logic cells available, designers can implement complex algorithms and systems directly in hardware for improved performance over traditional software solutions.

In addition to their logic capabilities, Xilinx 7 Series FPGAs feature an array of high-speed serial communication interfaces. These include support for technologies like PCI Express, Gigabit Ethernet, and Serial RapidIO, which facilitate efficient data transfer and integration into enterprise-level systems. The presence of high-speed transceivers also makes them ideal for applications that require fast data handling like video processing or high-frequency trading.

Furthermore, these FPGAs offer extensive memory options, including support for a wide range of external memory interfaces. This versatility allows for the integration of high-bandwidth memory solutions, which is essential for performance-intensive applications. With the introduction of the Memory Controller IP, users can easily connect various memory types, ensuring flexibility in system design.

Finally, Xilinx has made significant strides in development tools for 7 Series FPGAs, providing a robust ecosystem for design engineers. With design suites such as Vivado and SDK, users benefit from a comprehensive platform for deciding, simulating, and implementing designs efficiently. The combination of advanced hardware capabilities and powerful software tools solidifies the position of Xilinx 7 Series FPGAs as a preferred choice for custom digital hardware design across various industries.