Tool/software:
Hi Teams,
We have a complex USB design used HD3SS3212 mux to swicth 2 USB3.2 Gen.1 host to 1 device.
The block-diagram as below.
How about the DC blocking capacitors we allocated?
From HD3SS3212 spec shows that it would need bias resistor to provide USB signal DC bias when both side of mux have place capacitors.
But we do done this test, it could work. Does it have any potential risk?
BR.
Phil
Hello Phil,
If the common mode voltage in your system is higher than 2V. The AC coupling caps need to be placed on both sides of the switch with a biasing voltage smaller than 2V. Without the bias you might have issues in your system.
The capacitors need to be on the TX and RX sides, which is what you have in your diagram; but you might have too many.
The red areas should have a pair of AC caps, the rest are not needed and can be removed. Add a capacitor on the RX lane of the PCIE.
Thanks,
Nir
Hi Nir,
There is no any DC bias coming from HD3SS3212 right? PCIe bridge has 1.2V bias, hub need to block DC from other source.
So I need to reserve the cap on the right side and remove all the caps on the left side, right?
Phil
Hello Phil,
No, there is not.
The mux requires AC coupling caps in both sides (input/output) on the RX and TX lanes like in the diagram you provided.
So you still need a cap here (yellow) and can remove the other ones.
Each side requires only one pair per lane of decoupling caps from the mux perspective.
Thanks,
Nir
