This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

TPS65987EVM: TPS65987EVM problems

Part Number: TPS65987EVM

I’m an FAE in the Israeli team,

I am currently supporting a customer,

The customer is using the TP65987EVM with display port sink board from 10G expansion EVM attached (image attached),

He is facing the following three problems:

 

  1. In the pictures he sent me(attached) you can see the two CLK levels on the I2C line in between the FTD and the TSP65987,

From seeing the pictures you can see that there is a collision on the line and he does not understand why this happens.

 

  1. The board is configures as DRP ( dual role power)  ( and a UFP-D for Display port) so when it is connected to power source The board is charging the laptop and when he disconnects the power source the laptop is the power source for the board.

The problem is that when he connects the power back it does not change back to become a charger of the laptop unless he reconnects the usb-c . What should he     change in the configuration to enable this to happen.

 

  1. The display port function is working with the laptop but not with my smartphone. It is a Samsung S10e and it works with any adaptor that is on the market but not with the TI board.

 

 

Can you please help me with the problems he is facing?

  • Hello Victor,

    For your three problems:

    1. I am not sure what lines you are looking at. I2C2 (which is used for FTDI) has a pullup, which is not reflected in the scope shot. Also,  I2C2 is only used in debug mode through the GUI for the EVM. 

    2. There could be some equalization issues occurring when the phone is driving the super speed lines. You could try doing 4 lane DisplayPort if your phone can support that. 

    3. Check to see if GPIO21 is set to Barrel Jack Detect Event. If it is not, enable it. This will check to see if the barrel jack has been connected and should enable a power role swap when it is.

    Thanks,

    Emma

  • Hello Emma,

    this is my reply from customer:

      1. I2C(which is used for FTDI) has a pullup, which is not reflected in the scope shot. Also,  I2C2 is only used in debug mode through the GUI for the EVM.

    As can be seen in the Scope picture it is not a problem of pullups. You can see in a train of clocks that for some bytes it is 3.3V amplitude and for some it is about 1V. I don’t have any explanation except bad IC.

    May be it is easier and cheaper to send a new EVB.

     

      1. There could be some equalization issues occurring when the phone is driving the super speed lines. You could try doing 4 lane DisplayPort if your phone can support that.

    It is not a problem of picture through Display port . it is a problem of negotiation . you can see on the pictures that when I connect the laptop the led of the DP (d13) is lit even when DP cable is not connected to EVB.

    On the second picture when the smartphone is connected the Led is off.

    If the debug function was functioning I could capture the status and analyze it but unfortunately it does not function.

     


     

     

     


     

     

     




     

      1. Check to see if GPIO21 is set to Barrel Jack Detect Event. If it is not, enable it. This will check to see if the barrel jack has been connected and should enable a power role swap when it is.

    That solved this problem, thanks.

     

  • Hello Victor,

    Glad to hear the GPIO event solved one of your issues.

    I2C works through pull ups, so if the lines are not pulled up, the communication will not work. If I2C2 is broken, you should order a new board. Your I2C communication usually look like this on the TPS65987EVM. 

    Thanks,

    Emma

  • Thank you Emma,

    regarding the second question my customer asked, can you please help me answer his comment(resending)?:

    1. There could be some equalization issues occurring when the phone is driving the super speed lines. You could try doing 4 lane DisplayPort if your phone can support that.

    a. It is not a problem of picture through Display port . it is a problem of negotiation . you can see on the pictures that when I connect the laptop the led of the DP (d13) is lit even when DP cable is not connected to EVB.On the second picture when the smartphone is connected the Led is off.

    If the debug function was functioning I could capture the status and analyze it but unfortunately it does not function.

     

  • Hi Victor,

    As discussed above, please order a new board for your customer or ensure that I2C is pulled up by soldering connections. I2C needs a pullup to work and if their I2C lines are not pulled up, the communication will not function properly. 

    Thanks,

    Emma

  • Hell ,

    I have ordered a new board for the customer(TPS65987EVM) as you have advised,

    He is still having the same issue with the display port function as it is working with the laptop but not with his smartphone.

    It is a Samsung S10e and it works with any adaptor that is on the market but not with the TI board.         

     

    The debug problem that was on the previews board is happening on the new one.

    See attached project files.

    Can you help me with this?

     tps65987 4th.zip

    Thank you for the support.

     

  • Hello Victor,

    Do you or your customer have a PD analyzer to read the PD messaging being sent between the phone and EVM? The first step for displaying video over Type-C is to negotiate and accept the display port alternate mode. In your previous post, you state "on the second picture when the smartphone is connected the Led is off". The purpose of this LED it to indicate that HPD (hot plug detect) has gone high. So the fact that this is not going high shows that a display port connection is established and HPD is not going high (cause for the display issue) or that their is no DP alternate mode negotiated between the two devices. Both would cause for the video not to display, and you need a PD analyzer to help isolate where the issue could be occurring.

    This is the one I typically recommend customers use.

    https://www.totalphase.com/products/usb-power-delivery-analyzer/

  • Hello Adam, 

    Yes the customer has total Phase PD analyzer and attached are the captured negotiation of 3 scenarios

    1. His Samsung S10e Smartphone with an Cypress based HDMI Dongle.
    2. His Samsung with TI EVM with no DP output. (it looks there is no negotiation for DP at all)
    3. MacBook Pro with the same EVM with same configuration with DP displayed .

     The configuration in the last post for checking.

    samsung s10e with TI evm.zip

  • Hello Victor,

    With the Samsung and TI EVM, the Samsung device never sends the VDM:Discover Modes message, so the DP alternate mode is never discovered and entered. With the MacBook pro and TI EVM, the MacBook does send this discover modes message, so the two devices are able to discover and enter into the DP alternate mode. I am having difficulties opening the Samsung and dongle file so I cannot review them fully, but if this test works, it is probably because the dongle is sending the VDM:discover modes message and the two devices are able to enter into the DP alternate mode.

    Since the Samsung phone is the one acting as the DFP. it should be the one to send out the request for VDM alternate modes.

  • Hi Adam, 

    My customer answered the following:

    "

    I have checked with another product which is based on Cypress component with the same Samsung S10e and it works. Here is the log file from cypress analyzer that succeeded to charge and send DP signals.

    I can’t accept that with Cypress PD controller it works and with TI not.

     "

    samsung usb lock.zip

  • Hello Victor,

    I will take a look at these logs and get back to you soon

  • Looking through the log you sent, the Samsung S10e does not negotiate a DP alternate mode with the Cypress device. All it shows is a power contract being negotiated, some data role swaps, and then a hard reset. Also, the excel file and the PD log do not match.

    You can also change the configuration of the TPS65987D to see if that helps. In the Display Port Capabilities register (0x51), there is a bit called "DisplayPort Mode Auto Entry Allowed". Make sure this bit is enabled

  • Hi Adam, 

    He is asking how can he force the configuration to fixed 4 lane DP alternate mode. No usb switching only fixed DP.

    Thanks.

  • The default configuration allows for DP 4 lane. This is done through the UFP and DFP pin assignments in the DP configuration. These pin assignments are defined by the DP alternate mode.

  • Hi Victor,

    To help debug this issue, can you please share the following PD logs with the "DisplayPort Mode Auto Entry Allowed" bit enabled on the TPS65987DEVM? You can use the application customization tool to update this register. 

    1. TPS65987DEVM with the Macbook
    2. TPS65987DEVM with the Samsung S10e
    3. Cypress PD controller with the Macbook
    4. Cypress PD controller with the Samsung S10e

    Also what device is the customer using that has the Cypress PD controller to negotiate a DP connection with the Samsung phone? Is it an EVM as well?