The Rev2.1 versions are starting to reach us through the supply chain and we are seeing severe instability on the clk out pin (31). The instability includes a random 100% modulation of the clock.
There has been no such problem with Revs 1.9 and 1.10.
We use that clock to drive an FPGA with 1CMOS input load via a track that is less than 1 inch long.
The MOD54415 has a ground-plane around the whole module and another for VCC3V3.
There are 0.1 ceramic capacitors near to the VCC3V3 pins.
Has anyone else seen this problem?
What is Netburner's actual recommendation for external supply decoupling?