TUSB212: TUSB212 CD pin is not behaving correctly

Hi Nadeem,

Is there any to break the connection between the hub and the redriver? Or is the only way to break the DFP to disconnect the device?

My current suspicion is that there is something with the hub to where if the host disconnect first, the pull-up/pull-downs that are active at the time it is removed are not disabled by the hub, causing the redriver to believe that a successful connection is still present and not disconnected. This is something I will try to test in the labs on one of our setups, but if possible, if there are any means to directly connect the B Receptacle to the A Receptacle and test whether the CD pin immediately enables and disables, that may be helpful.

Thanks,

Ryan

Hi Ryan,

     The design is very tight, no option to break the trace. We are seeing this issue with two different HUBs from different manufacturer. 

Hi Nadeem,

Are these two hubs the only two you have tested with? Have you tested any other hubs and have been unable to duplicate this effect?

On my side, I have tested this with one of our hubs being powered externally with a power source and connected to a host PC upstream, and a TUSB212EVM with a USB drive connected downstream. I tested both disconnecting the upstream first and disconnecting the device first, and in both cases, the CD pin disabled once either side disconnected.

If possible, could you monitor the DP/DM lines when the host is disconnected and see if DP/DM stay pulled-up?

Thanks,

Ryan

We have two boards with two different hubs. I haven't tested any other hub as I don't have a simple way to connect conditioner chip to external hub.

I checked the lines after removing the ufp, they were not pulled up but CD was still hi. 

Does tusb212 look for pullups or terminations to change CD pin status? If it looks for terminations then probing the lines will not give useful info.

Hi Nadeem,

Would it be possible to send the schematic surround the TUSB212 to review? I want to make sure there are no issues on that side of things.

Would it be possible to monitor the CD pin just to see if there is any weird activity on the like when the UFP or DFP are disconnected?

Additionally, if possible, could you try replacing the TUSB212 with the TUSB216 or the TUSB211A? These devices provide improved functionality and are P2P or in need of only very slight modifications. I would be interested to see if there is any change in CD pin on a different device.

Thanks,

Ryan 

Due to NDAs I can not share schematics on open forum.

I can send the relevant schematics to your official email (as we already have NDA in place with TI).

Would you please let me know your email.

Hi Nadeem,

I sent you a friend request over E2E. You can send me a PM over E2E, that way only I can access these design files, and they won't appear on this thread.

Thanks,

Ryan

I got email for friend request. When I click on "accept" link, the web portal says "You do not have any friendship requests to review.". Looks like link is not working.

Hi Nadeem,

I sent you a PM over E2E just now. Please let me know if you got it.

Thanks,

Ryan

I have sent you sch on that link.

Hi Nadeem,

I looked over the schematic, and I'm not sure I see any issues that may be causing this. Does this issue appear on both TUSB212's, or have you only tested on one TUSB212? Is the to controller connection on the CD pins going straight to the controller with no additional components or signals from the controller.

For samples, I would recommend reaching out to your local FAE. They can complete a sample request order for you and get those delivered.

Thanks,

Ryan

CD pin is going thru a buffer to increase its drive strength.

What is the difference between tusb216 and tusb216I. I thought that its just the operating temperature but this app note "SLLA593 – AUGUST 2022" table 1-2 confuses me. Why the test mode compatibility of these two chips is different? 

Hi Nadeem,

I'm working on getting the specifics to this answer- I believe there used to be some differences in the devices, but that might have changes recently, so I am working on figuring out what the case is.

For sure though, you can use the TUSB216I, or if you want, I'd also recommend the TUSB211A. It requires fewer pins for configuration, and provides the same functionality in terms of performance as the TUSB216/I.

nadeem zia said:

CD pin is going thru a buffer to increase its drive strength.

Does this have any effect on the pin before the signal is raised high? I'm guessing not, just want to make sure.

Thanks,

Ryan

Hi Ryan,

     When you did the experiment with hub, did you remove the DFP of hub OR the UFP of hub?

The problem happens when you remove the UFP of hub (the cable between hub and laptop/desktop). as shown blue cable below.

I removed the hub from one of our board and connect off the shelf hubs (tried belkin and uspeed hubs). The issue still happens every time. I looked at the D+/- signals both are at zero level.

If I disconnect DFP of hub then  CD behaves correctly. 

Now I have 4 (2 external and 2 onboard) hubs showing same problem. 

The CD pin for the experiments is open and just connected to scope (it is isolated from the buffer)

Hi Nadeem,

Right now, My current suspicion is that the TUSB212 is entering test/compliance mode due to sequencing of disconnects, which is causing the CD pin to stay high. I think testing with the TUSB216I may fix this, so lets give this a try once the samples get there. I'll keep trying to debug this on my side in the lab as well just to see if I can replicate this, though I haven't been able to yet.

Please let me know if you have any other questions or concerns as well!

Thanks,

Ryan