• State TI Thinks Resolved
  • Locked Locked
  • Replies 3 replies
  • Answers 1 answer
  • Subscribers 48 subscribers
  • Views 643 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.

OPA192: OPA192 getting burned inputs

Svein Berg
Prodigy 10 points

Part Number: OPA192
Other Parts Discussed in Thread: LM2664, LM7705, TLV387

Tool/software:

Hi.

I have a problem with the OPA192.

OPA192 is used as a low impedance buffer to the input of the MCP2024 ADC.

From time to time, all of a sudden the OPA192 is defect.. The -5V is drawn down til almost ground and the current consumption goes to some 400mA.

I have been changing The OPA192 4 times now, and need an answer to WHY..

The first to OPA192 was blown with a different chargepump; The LM2664. So probably its not the cgarge pump itself.

Is there anybody with any clue to why this is happening?

  • 0
    Guru 323670 points

    What is the signal voltage range? Would it be possible that it becomes negative (which would be clamped by the ADC's input protection diode)?

    Can you check whether there is a short at an input or output pin?

    (The LM7705 would allow you to use a cheaper low-voltage opamp like the TLV387.)

  • 0 in reply to Clemens Ladisch
    Prodigy 10 points

    Input to OPA192 is 0-2V, so never negative.

    If not something strange happens during startup...

    No shorts. Sometimes the OPA have been working for hours and through multiple startups.

    Will have a look at the LM7705.

    Thanks..

  • 0 in reply to Svein Berg
    TI__Mastermind 28488 points

    Svein,

    When damage occurs to an op amp, the possible locations that can be damaged are: power supplies, input pins, and output pins.  In your case, the input pin has a 1M resistor in series with the pin.  A large resistor in series with the input is normally a good approach to protecting devices.  If a large overstress voltage is applied to the resistor it will limit the current flowing into the ESD diodes.  So, I don't expect that the overstress is on the input of the op amp.  The capacitance on the input is also helpful in protecting the input as it prevents rapid transients on the input.

    The power supplies can also be damaged by overstress voltage, but the OPA192 is a 36V device with 40V absolute maximum voltage.  Thus, the overstress voltage on the supply would need to be over 40V (across both supplies) to damage the device.  It seems unlikely with a +/-5V supply that the voltage would go beyond 40V but it may be possible that on power up or shut down that there is a transient.  Normally for all supplies I recommend a unidirectional Transient Voltage Suppressor (TVS) diode.

    The output is the only other place where the device could be damaged.  The schematic shows the output connected to MCP3421 but you mentioned MCP2024.  So, I am a little confused as to what is connected to the OPA192 output.  In any case, if the circuit the OPA192 is connected to has the potential to connect an signals beyond the OPA192 absolute maximum voltage, than this could be the source of the overstress.  Normally an ADC input doesn't output any voltage so this seems unlikely, but you should check to make sure that there isn't some abnormal behavior that causes the MCP to output an overstress signal.

    Another possibility is ESD.  If the PCB is mishandled or in an environment that has ESD present, than you may get ESD damage.  Again, I think this is unlikely.

    In summary, I don't see any obvious circuit design issues that would be a likely source of EOS.  I would recommend adding TVS diodes on the two supplies.  I think this is something that should always be added to every supply for any product.  You should look at all the op amp pins with a high speed scope to see the startup behavior and shut down behavior.  Normally, overstress issues happen on startup or shut down due to an inductive kickback that causes a rapid large voltage transient.  Event though the OPA192 is a high voltage device, I think the most likely source of the overstress is on the supply and I think TVS diodes would solve this issue.  I do not see any decoupling on the +5V supply.  Does this supply have decoupling (it is required)?

    Best regards, Art