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.

TPS23881: the parameter of Detection_Time_Tdet of sifos AF/AT test for the BT port

Part Number: TPS23881

HI all,

             The version R0B of PCB  has come back, Because of  some reasons, we can’t have the sifos BT test now. We use the BT loader to test the eight BT ports. They all can work normally.

              When we have the sifos AF/AT test for the eight BT ports, the parameter of Detection_Time_Tdet for 42 and 43 port is fail ,the value is 750 which you can see the report. Also We use the AT loader to test the eight BT ports. They all can work normally.

             Now, I have some questions as follow,

  • The attachment is the design of PSE card, we have move the sensor resistors and MOSFETs near the TPS23881 chip,can you help me to find some other resons?
  • If the parameter of Detection_Time_Tdet is fail, is there any question for customer use ?

 

Thanks.

  • HI Jingjing,

    Can you re-attach the report and design? I can't seem to open it from my end. 

    Regarding the Detection_Time_Tdet, the TPS23881 has a range of detection duration as highlighted below:

    This parameter is non-configurable and the PSE device should always follow this, assuming there's no issue from the hardware or firmware. 

      When we have the sifos AF/AT test for the eight BT ports, the parameter of Detection_Time_Tdet for 42 and 43 port is fail ,the value is 750 which you can see the report. Also We use the AT loader to test the eight BT ports. They all can work normally.

    Just to clarify, are you saying the other ports connected to the same TPS23881 didn't see this issue with detection time, only port 42 and port 43? 

    If so, what are the other port's detection duration and are all ports 4-pair configuration or 2-pair? 

    Once you re-attach the design file and the report I can help further debug this case.

    Thanks and Regards,

    Raymond Lin

  • Hi Raymond,

                  I don’t know how to add the attachment in the website ,so I send the mail to you. The attachment is the design and the report. U13 and U14 in the design are used for the eight BT ports.

                 The 142000003346H is the R0B version and the 14200000316S is the R0A version. The result of the R0A and R0B is similar. The fail port is same.

                 Also I have a question, when the BT port works as a AF/AT port ,which lane is supply ? 1,2,3.6 or 4.5,7,8?

  • Can you give me a mail address,I can not send you 

  • Hi Jingjing,

    I sent a friend request through e2e, you can send me your files through messaging. 

    Depending on how the PD and PSE are set up, if a 4-pair BT port connects to a 2-pair AF/AT port, either pair-set 12/36 or 45/78 will be used as the power lane. In most cases, pair-set 12/36 are used for 2-pair power but from the PD side if the hardware is set so pair-set 45/78 are used on the ethernet port to the PD then it will use that pair-set.

    On the PSE side you can read into certain registers to see which channel is being used to power on the 2-pair PD. For example, register 0x30 through 0x3F will give you information on each Channel's voltage and current reading where channel 1 refers to pair-set 12, channel 2 to pair-set 36, etc. Register 0x0C through 0x11 will tell you the status of the channels with detection, requested classification, and power status. For hardware testing you can also measure the voltage reading between Vdrain and GND to see if that channel is being used or not. 

    Hopefully that answer your question regarding BT and AT/AF port operation, let me know if you have any further questions or concerns. Once you send me the attachment through direct message I can help with debugging. 

    Thanks and Regards,

    Raymond Lin

  • Hi Jingjing,

    I've received the files, let me take a look into the schematic and test results. 

    Can you provide a timeline of when you need results back by? 

    Thanks and Regards,
    Raymond Lin

  • Hi Jingjing, 

    Nate will be taking over with supporting you on this thread, he'll be reviewing the schematic and providing feedback. 

    Thanks and Regards,
    Raymond Lin

  • Hi  Raymond,

           This question is very urgent,we must finish all the test of board at the end of this month.And we will give the machine to our guest next month.The attchment is some part of the switch board.

           Thanks.

  • Hi Raymond,

                we use 23881'S 0 and 1 port for the RJ45's 12 and 78 port.Can we change the sequence as 23881's 1 port connecting to the RJ45'S 12 port and 23881's 0 port connecting to the RJ45'S 78 port.

                 Also ,Do  the two ports of BT port  must be continuous as 0,1 or 2,3 or 4,5 or 5,6?

  • Hi Jingjing,

    I have some questions.

    Can you tell me which of the 7 PSEs is connected to port 42 and port 43?

    we use 23881'S 0 and 1 port for the RJ45's 12 and 78 port.Can we change the sequence as 23881's 1 port connecting to the RJ45'S 12 port and 23881's 0 port connecting to the RJ45'S 78 port.

    I don't understand what you mean by 23881's 0 and 1 port. TPS23881 does not have a 0 port. Can you draw a block diagram of how the TPS23881 ports are connected to RJ45 port 12 and RJ45 port 78?

     Also ,Do  the two ports of BT port  must be continuous as 0,1 or 2,3 or 4,5 or 5,6?

    Are you referring to ports or pair-sets? TPS23881 pair-sets are arranged 1-2, 3-6, 4-5, and 7-8.

    Best Regards,

    Nate

  • Hi Raymond,

                   According the picture,as working a BT mode,the TPS23881 'S 1-2 pair are connecting the transformer centertaps of net 1-2 and net 7-8 in the RJ45 connector.Is there any suqence request that the tps23881's 1 port must connect to the net 12 of RJ45 connector and the  tps23881's 2 port must connect to the net 78 of RJ45 connector. Can we change the sequence?

                 Also ,do you find some problem in the PCB layout?

                 Thanks.

                        

  • Jingjing,

    This requires further investigation. I will have an answer for you tomorrow. 

    Best Regards,

    Nate

  • Hi Jingjing,

     

    When the BT ports are tested using the AF/AT test, a failure is to be expected. BT standards for testing are different than AF/AT standards because BT standard devices have a different Maintain Power Signature (MPS) than AF/AT standard devices do, and the Sifos tests use this MPS to help determine pass or fail.

     

    Best Regards,

    Nate

  • Hi Nate,

             Why some BT ports can pass the sifos AF/AT test,some BT ports can not pass the test. Also the BT port can use as the AF/AT port ,the protocol is compatible.If we want to pass the sifos AF/AT test, how to test?

               Thanks .

  • Hi Jingjing,

    Due to the differences in first class event timing, BT devices should not pass AF/AT tests. (first image from BT standard, second image from AT standard)

    It is true, BT standard devices are backwards compatible and can operate as AF/AT devices, but the Sifos tests are designed to work only with the devices that their standards correspond to. A BT device cannot pass an AF/AT test. 

    Best Regards,

    Nate

  •       According the picture,as working a BT mode,the TPS23881 'S 1-2 pair are connecting the transformer centertaps of net 1-2 and net 7-8 in the RJ45 connector.Is there any suqence request that the tps23881's 1 port must connect to the net 12 of RJ45 connector and the  tps23881's 2 port must connect to the net 78 of RJ45 connector. Can we change the sequence?

         Also,we change the sequence in the sofeware,then we have a sifos AF/AT test,the result is not same.We ask our software engineer,they say there is the sequence.

        Then I want to know what is the right sequence.

         Thanks.

  • Hi Jingjing,

    There must be a channel to send power and a channel to receive power. It is standard for channels to be set up this way. 

    The signal over Alternative A will be sent slightly earlier than the signal over Alternative B, so if they are swapped it is expected that the results would not be the same. However, results from AF/AT Sifos tests on BT ports will not be valid in either configuration. 

    Best Regards,

    Nate

  • Hi Nate,

          In our schematic diagram, the 42 port is a BT port using 4-Pair Wired mode.We use the TPS23881'S 1 port as Alternative A is connecting to the transformer centertaps of net 1-2 and use the TPS23881's 2 port as Alternative B is connecting to the transformer centertaps of net 7-8.

          The 44 port is also a BT port using 4-Pair Wired mode.We use the TPS23881'S 4 port as Alternative A is connecting to the transformer centertaps of net 1-2 and use the TPS23881's 3 port as Alternative B is connecting to the transformer centertaps of net 7-8.

       We use the auto mode in the software and we don't have the port re_mapping of all the ports. According the signal over Alternative A will be sent slightly earlier than the signal over Alternative B.

        Now ,do we must have the port re_mapping to the 44 BT port ?

        Thanks.

  • Hi Jingjing,

    You won't need the port remapping, but you will need to use the 4-Pair Wired and Port Power Allocation Register, register 0x29. 

    This register sets whether the channels are in 2-pair or 4-pair configuration.

    Best Regards,

    Nate

  • Hi Nate,

             Now the value of register 0x29 is 0xFF.I want to know when we need to have the port remapping?

             Also when we have a sifos BT test ,do we need to have the port remapping and is there anything need to set in the software?

             Thanks.

        

  • Hi Jingjing,

    For port 44, try remapping channel 3 to channel 4 and channel 4 to channel 3. Odd numbered channels get priority, so it may cause some issues for the test if channel 3 is used for Alternative B. 

    Bets Regards,

    Nate