• State
  • Locked Locked
  • Replies 3 replies
  • Subscribers 16 subscribers
  • Views 268 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

TS5MP646EVM: process to terminate the shield wires to the EVM board?

May
Genius 9880 points
Part Number: TS5MP646EVM

Hi Team,

We have received inquiry regarding TS5MP646EVM that they not getting output. Below is the inquiry.

" I am trying to use a TI TS5MP646EVM  evaluation module to test if I can choose between 2 NVIDIA computers and one camera. I pulled 4 mipi lane pairs and the clock signals out of the camera cable and ran them through the EVM board. I am not getting any output. What is the process to terminate the shield wires to the EVM board?"

Regards,

Maynard

  • 0
    TI__Genius 10974 points

    Hi Maynard,

    If you are not getting any output from the device, have you checked the following?

    1) Are your logic levels correct? Please note this is an active low enabled device (device active when OE pin is low), with an SEL pin as well which will select the appropriate channel path.

    2) What voltage is present on the input? Have you measured on the board to make sure the signal is making it to the input of the switch?

    3) Make sure your supplies are greater than or equal to the signal you are trying to pass, otherwise the switch will not be able to pass the signal.

    If possible, a schematic of the setup would be helpful as well to help debug this.

    Thanks!
    Bryan

  • 0 in reply to Bryan Marshall
    TI__Genius 9880 points

    Hi Bryan, 

    Below is the response from customer.

    "

    I will work on the items that were suggested link

    I do not have a schematic per se, I will explain what I am trying to accomplish.
    I have a Leopard Imaging LI-IMX264-MIPI-M12 camera connected to a Leopard Imaging Systems LI-NANO-CB / Nvidia Nano stack via an IPEX FAW1233 30 pin micro coax cable.
    I pulled the 4 data lane and clock signals (10 wires total) out of the IPEX cable bundle and connected them to the TS5MP646EVM Board then back to the LI-NANO.
     
    FROM THE CAMERA TO THE TS5MP646EVM Board:
    The 4 Data Lane signals connect to TP3 through TP10.
    The clock signals from the camera connect to the TP11 and TP12.
    FROM THE TS5MP646EVM Board to the LI-NANO-CB / Nvidia Nano stack:
    The 4 Data Lane signals from the EVM board go from TP20 through TP27 on the EVM to the corresponding pins on the LI-NANO.
    The clocks go from TP13 and TP14 to their respective pins on the  LI-NANO for Channel A operation. I do not have anything connected to Channel B at this time. I am trying to test a Proof of Concept.
    The wires are micro coax and each have a shield. I do not have the shields connected to the EVM board. I am sure this is also causing issues. I am also inquiring how this is typically done.
    I have the Select and Enable jumpers set:
    J1:
    Lo= Channel A output.
    J2:
    Lo= Enable.
    I am powering the EVM Board with an external power supply set to 3.3 volts on + on TP1 and 
    - on TP17.
    My ultimate goal is to be able to switch from one MIPI camera to 1 of 2 Nano's. I am only working with one Nano currently on Channel A.
    "
    Regards,
    Maynard
  • 0 in reply to May
    TI__Genius 10974 points

    Hi Maynard,

    Nothing is jumping out to me from the setup. Can the customer try doing a basic setup to confirm the part is operating properly? Essentially, check that you can pass any signal from input to output? Just using basic supplies/inputs without the external cameras. This would help confirm the part is operating properly at least.

    Lastly, the shielding usually provides a low impedance path to GND, however, they would need to make sure the GND on the EVM side of the cable and the other side are the same potential, otherwise you risk creating a ground loop.

    Thanks,

    Bryan