• State Resolved
  • Locked Locked
  • Replies 11 replies
  • Subscribers 63 subscribers
  • Views 1987 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.

TLC59116 ESD Problem

Andreas Kurz
Prodigy 100 points
Other Parts Discussed in Thread: TLC59116

We have a problem with the sporadic shutdown of TLC59116.

Now it´s clear. The problem is electostatic discharge.

I tried many things but nothing works. Electostatic discharge generates a shut down of the TLC59116. A RESET at the processor starts the TLC59116 again.

Could someone send me a suggestion how i can protect the TLC59116 from electostatic discharge?

Or should we change something in the software?

Regards,

Andreas

  • 0
    Guru 23000 points
    Hi Andreas,

    Protecting against Static Discharges can be very difficult.
    Do you know the source of the discharge?
    How is your board protected and connected to the outer world?

    Leo.
  • 0 in reply to Leo Bosch
    Prodigy 100 points

    Hi Leo,


    sorry for the late answer.

    Contact Discharge is the source. I use an encoder with a metal shaft. If we touch the shaft a Discharge event occur and the TLC59116 made an Reset. The Controller on the other Board is not affected.

    The Metal Shaft of the Encoder ist the only way to the outer world. I think the TLC59116 is very sensitive. I think the problem is found at the I2C lines.

    I protect the I2C lines with 823 560 501 01 from WÜRTH. It is a Bidirectional Suppresssordiode with 5V and 100pF. But that dont work.

    The Capacity is now a problem for the speed of the I2C. But that could be solved.

    Is there anybody out there which uses the TLC59116 and had the same Problem as I?

    Thanks

    Andreas

  • 0 in reply to Andreas Kurz
    Guru 23000 points

    Hi Andreas,

    If the encoder gets a discharge not only the I2C line are affected but all connections gets the same discharge spike in respect to earth, even Vcc and GND.

    Now it comes to how is everything connected; You say two boards, are there two power supplies? How are the boards connected together, including GND and Vcc?

    Is the encoder metal cased or plastic.

    What is connected to the shaft?

    It is possible the shaft himself is charged-up and by touching you discharge it.

    An solution could be to separate connect the shaft/housing nut to the board power input ground.

    Leo.

  • 0 in reply to Leo Bosch
    Prodigy 100 points

    Hi Leo,
    i will send a photo.
    On one Board is the power supply and the TLC´s are on the other board. Connected with 14pin ribbon cable 150mm
    the encoder is metal
    to the shaft there is a knob (wheel) connected. it is aout of aluminium.
    No the shaft isn´t charged up. If i have dischrged myself then its all o.k.
    I have separated the encoder housing from the planes (see the photo)

    regards

    Andreas

  • 0 in reply to Andreas Kurz
    Guru 23000 points

    Is the encoder housing unconnected?

    Somewhere there must be a connection from your electronic to earth. Can the discharge goes from encoder to MCU board via 14p flat to TLC board trough TLC to LED (string) to earth?

  • 0 in reply to Leo Bosch
    Prodigy 100 points
    No, the encoder is connected over a bidirectional suppressordiode and a few ceramic Capacitors to GND.
    The GND ist connected via 14p flat cable to MCU Board.
    I tried a direct earth connection from shaft but nothing works. I have only the hope that a continously updating of the TLC solves my problem.
  • 0 in reply to Andreas Kurz
    Guru 23000 points

    Andreas Kurz said:
    I have only the hope that a continously updating of the TLC solves my problem.

    But this doesn’t solve the discharge problem and your TLC and/or other parts can be destroyed in time.

    Without a schematic overview of your parts and their connections it’s difficult for me to see where the problem can be. The photo doesn’t show much.

  • 0 in reply to Leo Bosch
    Prodigy 100 points
    Hi Leo,

    I send you my schematic and my PCB Board
  • 0 in reply to Andreas Kurz
    Prodigy 100 points

    The highlighted line is the SDA and nearba is ths SCL line.

  • 0 in reply to Andreas Kurz
    Guru 23000 points
    Sorry, still not enough! I must be able to see the whole picture, where the discharge current can go. You are too much focused on the I2C bus only.

    This, what you show now, is a board with display and the encoder and not the TLC I guess.
    To complete; Where are the MCU, TLC, LED’s and at which point you supply power. From AC adapter?
  • 0 in reply to Andreas Kurz
    Guru 23000 points
    A few other remarks;
    Decrease the I2C pull-up’s to max 4.7KOhm, use on each end of the bus a pull-up so it comes to each bus line have two pull-up’s of 4.7K (2.35K effective). Place the pull-up’s as much as possible on the end of the lines/traces.

    Place at each I2C a series resistor of ~300Ohm in the SDA & SCL line, close to the I2C device. This to avoid high voltage spikes.
    Keep capacity of I2C bus lines low, locate them on the outer side of the flat cable, if possible both. Use two Vcc or GND lines between SDA and SCL to avoid crosstalk.
    Decouple the wires surround the I2C bus lines (+8V) with capacitors of 10-100nF at each connector pin or one if the pins are connected via a plane (copper pour).

    D50 is useless, at the moment he shorts the circuit it’s already far over 3.3V and everything will already been blown. A ceramic capacitor of 100nF on this place will give better result.