• State
  • Locked Locked
  • Replies 7 replies
  • Subscribers 30 subscribers
  • Views 1711 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.

TM4C - TM4C1230 EMI problems

michael starinki
Prodigy 70 points

Hello,

I am using a TM4C1230 in an industrial control application.  The product requires certification with IEC 61000-4-2 (level 3 = +/-4KV contact) electrostatic discharge.  I am using a SM05-02HTG and a 0.33uF capacitor for suppression across the power rail that connects to an external sensor.  This power rail is from a 5V supply which is separate from the 3.3V supply provided to the controller, but both share the same ground plane.  Occasionally during an ESD test, the controller will stop functioning (partially) and requires a power cycle to return to normal operation.  What I have noticed is that following anESD event, the device stops communicating over its serial port and one of its status LEDs stops working.  The other three LEDs on the product continue to flash at the predefined rate, so I am certain that the controller is still running and the clock frequency is still correct.  I can also elicit similar behavior when I charge the ESD gun near the back side of the circuit board close to the microcontroller.

Has anyone else experience odd behavior like this during an ESD test, and can you give me any ideas where to look for problems?

I have identified several potential mistakes, but I am not certain if this will have a significant impact.

1)  The decoupling capacitors on net Vddc were mistakenly placed near pin 25 as opposed to near pin 56 of the controller.

2) The external crystal is located near the offending input

  • 0
    TI__Guru**** 244400 points
    Hello Michael,

    The TM4C123x device have an issue with ESD latchup due to the 5V tolerant ESD structure. This is described in the errata as well.

    Regards
    Amit
  • 0 in reply to Amit Ashara
    Prodigy 70 points
    I reviewed the errata but I am confused. Does this mean that if I "Zap" the digital ground in my circuit, any and every GPIO pin can be subjected to latchup? Holy smokes, batman!

    Is this prevalent in the TM4C1294 as well?
  • 0 in reply to michael starinki
    TI__Guru**** 244400 points
    Hello Michael,

    On the TM4C1294 there are only 2 IO's with 5V tolerant circuit (USB pins) which have this issue.

    Regards
    Amit
  • 0 in reply to Amit Ashara
    Prodigy 70 points
    I have a design that uses the TM4C1294 as well.

    In one of the other forums, you stated that the static issue "is a known issue on TM4C129 and requires the use of caps on the USB DP-DM lines." I have been unable to identify these capacitors values anywhere. Can you please elaborate on the values?
  • 0 in reply to michael starinki
    TI__Guru**** 244400 points
    Hello Michael

    In the other thread I had stated the errata. If the pins are unused, then it is safest to connect them to GND via a resistor. If USB function is being used, a Low Pass Filter needs to be in place to control the rise and fall time of the incoming signals and these need to be sized as per the requirement of the design and their placement must be kept as close as possible to the pins.

    Regards
    Amit
  • 0 in reply to Amit Ashara
    Prodigy 70 points

    Amit,

    Thank you for the info on the TM4C1294

    Regarding the TM4C1230

    The addition of an RC network and pull-down resistor to every input seems extreme to correct this problem.  I understand the reasoning, but it is very impractical to implement in most products.  I have not seen this kind of an issue before and it makes me wonder what the best option to proceed with our design.

    Is there an alternate component that fits this footprint/pinout but does not suffer from latch-up?

    Will this problem show itself differently among individual TM4C1230s?

    If I change the design and accommodate every input with a 1K pull-down or an appropriately sized RC, how confident are you that this will resolve the symptoms during the ESD test.

    Would you be available sometime on Monday for a discussion of these and other alternatives?

  • 0 in reply to michael starinki
    TI__Guru**** 244400 points
    Hello Michael

    No, there is no alternate component which is pin compatible. The only mitigation could be TM4C129x class devices which have this limitation only on the USB VBUS and ID pins.

    Regards
    Amit