forked from M-Labs/artiq
1
0
Fork 0

wrpll: update OBUFDS_GTE2 comment

Seems O can fan out simultaneously to transceiver and fabric.
Kasli is using ODIV2 for no particular reason.
This commit is contained in:
Sebastien Bourdeauducq 2019-12-09 11:57:33 +08:00
parent 4148efd2ee
commit 05e2e1899a
1 changed files with 2 additions and 5 deletions

View File

@ -27,12 +27,9 @@ class DDMTDSamplerGTP(Module):
self.rec_clk = Signal()
self.main_xo = Signal()
# Getting the main XO signal from IBUFDS_GTE2 is problematic because:
# 1. the clock gets divided by 2
# 2. the transceiver PLL craps out if an improper clock signal is applied,
# Getting the main XO signal from IBUFDS_GTE2 is problematic because
# the transceiver PLL craps out if an improper clock signal is applied,
# so we are disabling the buffer until the clock is stable.
# 3. UG482 says "The O and ODIV2 outputs are not phase matched to each other",
# which may or may not be a problem depending on what it actually means.
main_xo_se = Signal()
self.specials += [
Instance("IBUFDS",