Hard Processor System Technical Reference Manual: Agilex™ 5 SoCs

ID 814346
Date 4/01/2024
Public

A newer version of this document is available. Customers should click here to go to the newest version.

Document Table of Contents

12.3.1.5.3. Introduce Board Trace Delay

RGMII specifies simultaneous generation of the clock and data by the transmitting source which requires introducing a skew between clock and data. The skew can be achieved by matching PCB trace routing delay in the transmitting or receiving node. The minimum delay needed in each direction is 1 ns, but it is recommended to target a delay of 1.5 ns to 2 ns to keep timing margin.

The downside of this method is that the clock skew requirements for RGMII can vary depending on the specification of the Ethernet PHY device being used and the specific RGMII version. In general, the skew tolerance for RGMII is on the order of a few nanoseconds, which is a relatively tight tolerance. This means that any PCB trace length mismatches or other components in the signal path that introduce delays or phase shifts can have a significant impact on the overall clock skew and, therefore, the reliability of the interface.

To ensure reliable data transmission in an RGMII interface, it is essential to meet the required clock skew specifications. This can be accomplished by using phase-locked loops (PLLs), delay lines, or other timing adjustment techniques to compensate for any delays or phase shifts introduced by the PCB traces or other components in the signal path.

Additionally, proper PCB trace length matching and timing analysis tools such as Signal Integrity (SI) and Time-Domain Reflectometry (TDR) simulations can be used to verify the clock skew and ensure that it is within the required tolerances.

Figure 305. Using PCB Trace Delay to Introduce Clock Skew