• State
  • Locked Locked
  • Replies 19 replies
  • Subscribers 29 subscribers
  • Views 2060 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

TM4C129ENCPDT: Strange problem

Ilya Mazurenko
Expert 1300 points
Part Number: TM4C129ENCPDT

Hello Team,

Our customer have strange problem with TM4C129ENCPDT.

In their device TM4C129ENCPDT doesn't launch. But after a little heat(for example using finger) it launches.
Could you provide some comments? Or additional questions that we need to clarify with the customer.

You can find the video at the link below(dropbox):
www.dropbox.com/.../TM4C problem.mp4

BR,

Ilya

  • 0
    TI__Guru 72500 points
    My guess is that it is a bad solder joint.
  • 0 in reply to Bob Crosby
    Expert 1300 points
    Hello Bob,

    This is the first advice i gave them before i posted question here.
    They said that better soldering didn't help.

    BR,
    Ilya.
  • 0 in reply to Ilya Mazurenko
    TI__Guru 72500 points

    Is this a single unit that fails this way? Were they able to make any of these custom boards that work correctly?

  • 0 in reply to Bob Crosby
    Expert 1300 points
    Hello Bob,

    This device is already in mass production. All components has industrial temp range.
    This problem occurs in 1/500 produced devices.
    And it was in summer period. The customer is worried that a problem can occur more often in the winter period.
    What can you recommend?

    I can ask the customer for additional information if needed.

    BR,
    Ilya.
  • 0 in reply to Bob Crosby
    Expert 1300 points

    I can also add thermoprofile:

  • 0 in reply to Ilya Mazurenko
    TI__Guru 72500 points
    Is the customer able to access the device by JTAG on a failing board when it is in the failed state?
  • 0 in reply to Bob Crosby
    Expert 1300 points
    Hello Bob,

    Yes, they able to access the device by JTAG on a failing board when it is in the failed state.

    Also the customer added:
    They have external watchdog on it (STM6822TWY6F), it resets every 1.6 sec, but this isnt help. Only heating with finger helps(as on video).
    For example - power on - doesnt launch - heating with finger - launched - working - power off and then on - doesnt launch again (heating with finger needed again).

    Ilya.
  • 0 in reply to Ilya Mazurenko
    TI__Guru 72500 points
    If they access the device with JTAG while it is in the failing state, what is the CPU doing? Can they tell which routine it is in by the contents of the program counter?
  • 0 in reply to Bob Crosby
    Expert 1300 points
    Hello Bob,

    I will ask them to check this. And return to you with further information.

    BR,
    Ilya.
  • 0 in reply to Bob Crosby
    Expert 1300 points
    Hello Bob,

    The customer has tried to check what you asked

    When they connect debugger(J-Link) to the device, it always launches.
    So this way they cant check what CPU is doing in the failing state.

    If the cable from end device is not connected to the J-Link then TM4C129ENCPDT doesnt launch after power on.
    If they connect other end of this cable to J-Link and start the debugger then TM4C129ENCPDT launches.

    Do you have any ideas how to solve their problem?

    BR,
    Ilya.
  • 0 in reply to Ilya Mazurenko
    TI__Guru 72500 points

    What do they have attached to the RST pin?

  • 0 in reply to Bob Crosby
    Expert 1300 points

    Hello Bob,

    Thank you for the reply!

    The customer provided their scheme(in attachment)

    Could you please provide some recommendations?

    CTL04.2.700.pdf

    BR,

    Ilya.

  • 0 in reply to Ilya Mazurenko
    TI__Guru 72500 points
    Is XP2 a jumper plug? If so, does the problem still occur with XP2 removed (open)?

    What is the physical length of the TCK trace from XP3 to the TM4C129?
  • 0 in reply to Bob Crosby
    Expert 1300 points

    Hello Bob,

    1) Jumper is always added (They remove it only when working under J-Link. If it is not removed then processor will be reseted every 1.6 sec by external watchdog and there is no enough time to load J-Link).
    Also if working without J-Link and jumper is removed, then processor will launch when power on. OR if it is cold and didnt launched(like on video), then heating with finger doesnt help - processor will not launch.
    So for launch in this case(after a little heat) it is necessary to send a reset signal to the input.

    2) Length is 31.5 mm - but this line is not involved during normal operation.


    I should note, the problem occurs only in ~1/1000 of end products. But manufacturer(our customer) already faced with the return of non-working end devices(from their customers) with this problem.
    They are very afraid that in winter time this problem will start to occur more often.

    BR,
    Ilya.
     

  • 0 in reply to Ilya Mazurenko
    TI__Guru 72500 points
    I wonder if the issue is that we are not getting a good level on reset from the STM6822T. Can they look at the voltage on the "RES" signal? It should be a full 3.3V.
  • 0 in reply to Bob Crosby
    Expert 1300 points

    Hello Bob,

    Thank you for the reply!

    I got the following answer from the customer:
    They dont think that the problem is from STM6822TWY6F and plugged jumper(connected output of STM6822TWY6F).
    STM6822TWY6F has open collector output.
    Also the customer have made 2 scope shots direct from the jumper (Processor doesnt launch and resets every 1,5 sec):

    1. Reset Signal


    2. Front of reset impulse


    BR,
    Ilya.

  • 0 in reply to Ilya Mazurenko
    TI__Guru 72500 points
    I agree, it does not look like an issue with the level on RST. Can they take a part that fails, and reprogram it with a very simple code that just toggles a pin. That way we can identify if the part starts executing any instructions from flash.
  • 0 in reply to Bob Crosby
    Expert 1300 points
    Hello Bob,

    The customer tested it with simple program.
    It seems flash still doesnt launch.

    Also cutomer clarified, currently this problem occurs with ~25 processors out of 1500 every month on their production.

    The customer can provide pcb file. Do you need it to provide some help?

    BR,
    Ilya.
  • 0 in reply to Bob Crosby
    Expert 1300 points

    Hello Bob,

    Also the customer added pcb files:
    top.pdf

    bottom.pdf
    What can i do to help you identify the problem?
    I have no idea myself. This business is very important for us.

    If needed, the customer is ready to send their end device(with this problem) to TI's engineering/quality team.
    Also i can create a return request for the failed processors.

    P.s. We discussed the problem, which occurs with some end devices(with TM4C) returned from end customers.
    But currently we got information, that 3.8% of processors doesnt launch at all after soldering (so the customer cant even debug it) on the customer's manufacture.
    If this happens, they desoldering non-starting processor and solder another one, which works then on the same board.