AR# 66565

LogiCORE DisplayPort v6.1 (Rev. 1) - Why is the lnk_fwdclk_p/n reference input clock grounded for the RX IP for UltraScale Devices?

Description

UltraScale devices required two clock sources for the DisplayPort RX IP. The first clock lnk_fwdclk_p is for the line rates 2.7 and 5.4, and this comes as a forwarded clock from the DP159 re-timer.

Because the forwarded clock frequency from DP159 for 1.62MHz line rate falls under the CPLL holes for the UltraScale devices, we need to provide a separate input clock of 270MHz.


Hence there are two reference clocks for the Display Port IP connected in the DP core as below

  • MGTREFCLK0 - lnk_clk_p/n - external fixed clock of 270MHz
  • MGTREFCLK1 - lnk_fwdclk_p/n - forwarded clock coming from the DP159

When I generate the DisplayPort RX core for any UltraScale Device and try to trace the lnk_fwd_clk connectivity to the MGTREFCLK1, it is connected to ground. 

What is the reason for this and how can I fix this?

Solution

This is a known issue in the Vivado 2015.4 DisplayPort v6.1 (Rev. 1) GUI.

This issue is resolved in Vivado 2016.1.

Please see (Xilinx Answer 66301) for a patch for the DisplayPort v6.1 (Rev. 1) in Vivado 2015.4.


Linked Answer Records

Master Answer Records

AR# 66565
Date 10/03/2016
Status Archive
Type General Article
Devices
Tools
IP