HD3SS3220: Not recognizing USB 3.0 with development boards.

Hi Erwin,

Could you help provide some block diagrams of the setups you're using? I'm having a hard time visualizing what these setups look like off of description alone, so that would help me with figuring out where the issue may be.

Erwin Goris said:

Then if i add the mux chip and a USB C 20Gbit/s cable to an USB C poort of the PCI expentionboard, suddenly the USB 3.2 part of the USB7252C is not recognized anymore.

Does this PCIE expansion board use USB-A or USB-C connectors? Is there any way you can confirm what the capacitance on the lines could be, according to our EVM and what other capacitors may be on the lines?

Is there any way to monitor the CC lines to ensure there is an active connection?

Thanks,

Ryan

Hello Ryan,

To clarify my setup please see my hand sketch below.

The PCIE card has both A and C ports.

I have no further information on Capacitance on the lines of this expantionboard.

I attached 2 sheets of my design that is currently not functioning.

Looking forward to your reply.

Best,

Erwin.

PST-Hub UFP.pdf

USB Pipeline.pdf

Hi Erwin,

Got it, thanks.

Just to confirm, this cable is 20Gbps, but the connection itself is being done over USB3 at 5Gbps or 10Gbps, correct?

Is there any way you can monitor the DIR and CC lines on the HD3SS3220 EVM? I know it may be hard, considering the boards size. I would like to check those and make sure an active CC connection is being made, If there is a connection I believe there should be a voltage on one of the CC lines, and the DIR pin should be flipping accordingly.

Thanks,

Ryan

Hey Ryan,

I ordered an 20Gbps cable just to be shure this wouldn't be the bottle neck. the max speed needed is 10Gbps.

I measured the DIR pin with an oscilloscope and it clearly switches from high to low when i flip the USB-C cable. this seems to work.

Hi Erwin,

Then it sounds like the HD3SS3220 mux is switching properly most likely.

Is it possible to monitor the CC pins and ensure there is a connection thereThere should be a voltage on one of the lines indicating a connection, what is the voltage you're seeing there?

Is it possible to monitor the TX or RX lines and look for any activity on those lines? If there is a connection, you should be able to see activity on the RX and TX lines. If not, and you only see small pulses on the TX lines of the USB-A connector on the HD3SS3220 board, then that would indicate no connection is being made, If you don't see any signal on those pins at all, then the mux might not be flipping correctly.

Thanks,

Ryan

Hey Ryan,

On R19 and R21 i measure 1,8V and it flips when i rotate the USB C plug.

If i have no Hub attached and i  measure on C1 or C3 i have a pulse of 500mV at 83,33Hz with a pulse with of about 20uSec.

If i connect the hub i measure noise Swings +/-50mV.

Looking forward to your reply.

Best,

Erwin

Hi Erwin,

Erwin Goris said:

If i then connect the same setup to my motherboard USB-type C poort (5Gbit/s) it recognizes the the complete hub with USB 3.2 but i am limeted by the bandwith of the Root hub.

Just to confirm from the original question, is this also with the HD3SS3220, or is this system without the HD3SS3220 in the path? If the HD3SS3220 is not in the path and is able to work properly, then my concern is that the capacitance on the lines might be outside of spec. The HD3SS3220EVM uses capacitance values higher than what is listed in the spec in anticipation of there being other capacitors in series, so if there are not, that capacitance may be causing the USB3 connection to fail. You could test this on the HD3SS3220EVM by removing the capacitors and shorting the pads across.

Thanks,

Ryan