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.

LMH6643: Maximum allowable current into output pin during ESD

Part Number: LMH6643
Other Parts Discussed in Thread: STRIKE

Hello,

I am designing ESD protection for the LMH6643 for an event on the output pin. I would like to know the maximum allowable current into the input output pin during such an event.

The absolute maximum voltage on input/output pins is specified as a 0.8V difference (~1 diode drop) relative to the supply voltage:

The maximum recommended current through input pins during an input overvoltage is given in application notes as 10 mA:

The ESD protection circuit for the input is also shown:

Presumably the output pin is similarly protected. Should it be assumed that it is also not recommended to exceed 10 mA into the output pin during an ESD event on the device output?

Thank you.

  • Apologies but images got dropped from the post. I believe it is still answerable.
  • Hi Mosel,

    the 10mA maximum input current specification is not only to protect the ESD diodes but also to prevent damage to the internal protection scheme against too high differential input voltages.

    The usual way to protect the output of an OPAmp against ESD is to clamp the output signal to the rails by the help of two Schottky diodes, BAT54 e.g. A small resistor between this clamping point and the OPAmp output limits the remaining current flowing into the OPAmp during an ESD event. The series termination resistance (50R or 75R), often seen at the output of a HF-OPAmp, can ideally be used for this.

    Kai
  • Hi Kai,

    I am indeed implementing a protection scheme like the one you have described, along with some additional protection via TVS diodes and a ferrite bead on the front end. However, my spice simulation still shows a 150 mA pulse into the output pin lasting about 60 ns during the ESD event of interest. I am wondering whether the device can tolerate this. Reducing that current to 10 mA requires upping the series resistor to 1 kohm or more, which is not really an option since the load is only 50 ohm and the power loss would be excessive.

    For reference I attached the image of waveform. Negative current is flowing into the op-amp output pin.

  • Hi Mosel,

    This looks equivalent to a temporary output short circuit. According to the Absolute Maximum Ratings, the limiting factor for this specification is junction temperature. This part is designed to supply short circuit current for up to 1.5ms, so I am not concerned about 60ns.

    Best regards,

    Sean
  • Hi Mosel,

    during a 2kV ESD (HBM), which the LMH6643 is designed for to withstand, a current of more than 1A is flowing through the internal protection scheme for a very brief period. So, your 150mA ESD current spike should be acceptable.

    Kai
  • Sean,

    In a short circuit scenario, wouldn't the primary concern be current flowing through the BJTs and not through the ESD clamping diodes? From the datasheet:

    (Note: I am presuming the output rail is clamped similar to the input rails based on other information in the datasheet. This is not shown in a diagram)

  • That is a sensible argument, but what then to make of this statement in the datasheet? Perhaps this only refers to the case of a comparator where a differential overvoltage may frequently occur and the ESD diodes will be exposed to high current for long periods?

    Edit: I am presuming the output rail is clamped and protected similar to the input rails based on other information in the datasheet and hence the current flow limitation would be similar, but a diagram is not provided in the datasheet.

  • Please refer to my previous comment above.

  • Yes, if you are using feedback and the output is not railed out like a comparator, the BJTs can supply way more current than the ESD diodes. Are you using feedback? If so, these BJTs will try to correct the ESD strike, as they would a short circuit.
  • Hi Sean,

    Yes. I am using a feedback configuration.

    So in reference to your feedback and my discussion with Kai above, is the 10 mA maximum on the output pin a continuous limit? Kai pointed out that a 2 kV ESD would induce a transient current up to 1A and the device is stated to be able to handle an ESD event of that voltage.

  • Hi Mosel,

    yes, the 10mA maximum input current specification is a continuous limit. But TI recommends in many posts to keep the input current well below this 10mA. This current limit is ok for rare and short lasting events but not for extended periods under normal operation.

    Again, the ESD protection scheme (input and output) is designed to withstand 2kV ESD (HBM). So, when your additional protection circuitry keeps the stress to the OPAmp below this level, everything should be fine.

    Kai
  • Excellent. Thank you Kai.