• State Resolved
  • Locked Locked
  • Replies 9 replies
  • Subscribers 48 subscribers
  • Views 672 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.

LMV358A: Overvoltage on input

Greg
Expert 2071 points

Part Number: LMV358A
Other Parts Discussed in Thread: TPS62A01

I have a few questions about overvoltage on the input of the LMV358A op-amp.  I watched one of TI's videos on this subject, and here is a screenshot from the video:

Input Resistor To Prevent Damage

In this diagram, are the TVS diodes included because the +/- 15V supply is a linear regulator, and linear regulators are unable to sink the current?  If a buck regulator is used, would the current have a path to ground?  I will be using the TPS62A01, which operates in forced PWM all of the time.  The block diagram of the device shows high and low-side FETs, so it seems if the output voltage rises, the low-side FET would turn on more often which would pull current out of the inductor.  But that process may be slow, so is that the reason for needing the TVS diodes?

I don't need high accuracy, so I am willing to use a very high-valued series resistor in the signal path, such as 1 MOhm.  This would greatly reduce the current, even for voltages of several hundred volts.  How low would the current need to be in order to allow the Absorption Device in the op-amp to absorb the extra current, making the TVS diodes unnecessary?

Since the LMV358A has some internal ESD protection, will a 1M series pulse-proof resistor on the input be sufficient to prevent ESD from being an issue?  Or would I need to add a TVS diode on the input as well?

Thank you for the help.

Best regards,

Greg

  • +1
    TI__Mastermind 33723 points

    Hey Greg,

    Thanks for taking the time to look through our videos! You are correct, the external TVS is there because of the linear regulator's inability to sink current. Your supply voltage will rise according to the current through the ESD diode and the capacitance on the supply pin. This will be governed by I = C*dV/dt. As long as the input current does not charge the supply faster than the TPS62A01 can sink current and discharge it, and the pin voltage at V+ is kept below absolute maximum, then the main concern is current through the ESD diode.

    For short durations, the recommended current is <10mA. If there was going to be an extended condition for the ESD diode to conduct (for example, the input voltage is always on, but the amplifier is not), your recommended current is <1mA.

    I would size the resistor to be large enough to keep the current below 10mA, and ensure that the converter can sink enough current to keep the voltage steady below absolute maximum.

    Best,
    Jerry

  • 0 in reply to Gerasimos Madalvanos
    Expert 2071 points

    Thank you Jerry.  Voltages beyond the supply voltages will be accidents, and not likely to last more than 5 or 10 seconds.  I just need to ensure the circuit will not die.  With a 1M resistor, even 500V would result in about 0.5 mA.  The max acceptable voltage I expect will be measured will be about 30V. I don't think I will need to include TVS diodes on the supply if the current is in the range of 0.5 mA or less for short durations.

    Regarding ESD protection, how would I go about analyzing it?  The LMV358A has some ESD protection.  Is this protection separate from the 10 mA that is allowed through the steering diodes?  If a 20 kV zap was applied top one end of the resistor, the IC would see about 20 mA of current for a short period, assuming it presents itself as a low impedance.  Is that a simplified approach that would tell me if the resistor is sufficient or not?  The duration of the 20 mA would be very short, so many the op-amp would survive?

    Best regards,

    Greg

  • +1 in reply to Greg
    TI__Mastermind 33723 points

    Hey Greg,

    By max acceptable voltage range, do you mean on the input before the 1MEG resistor, or is this on the supply pin?

    The ESD protection is separate from the EOS protection. The input diodes are designed for ESD stress, but are resilient to much lower levels of EOS stress.

    The LMV358A is rated for 2kV HBM and 1kV CDM, which is primarily for ease of manufacturing. If this is connected off-board with a cable/connector and you would like to protect the inputs from ESD, it is recommended to include ESD protection diodes at the connector to avoid damaging the device.

    Best,
    Jerry

  • 0 in reply to Gerasimos Madalvanos
    Expert 2071 points

    Hi Jerry,

    I'm sorry, I didn't use the best choice of words.  What I meant by that was the max typical voltage a user would try to measure, which implies an input voltage that doesn't activate the steering diodes.

    I plan to include a TVS diode on the input and not populate it unless I need to.  But I'd like to iron that out ahead of time if possible.

    During an ESD event, a 1M resistor would form a voltage divider with the op-amp.  20 kV applied to one end of the resistor would result in a voltage at the input of the op-amp.  Would that voltage be low enough to fall within the 2kV HBM and 1kV CDM ratings?  Alternatively, if the impedance of the op-amp is very low, one could assume the voltage at the input would be very low, and potential damage would be caused by the 20 mA of current that would flow for the duration of the ESD event.

    I appreciate any insight you can give.  Is there an ESD e2e forum I should ask this question in?

    Best regards,

    Greg

  • +1 in reply to Greg
    Guru 322000 points

    The ESD ratings are tested directly at the pins.

    If you haver a series resistor, you can ignore the ESD ratings, and you need to ensure only that the clamping current is not too large.

  • +1 in reply to Clemens Ladisch
    TI__Mastermind 33723 points

    Hey Greg,

    I agree with Clemens comments. Here is a simulation representation of the ESD tests, so it seems that a large series resistance before the pin would help with the robustness.

    Best,
    Jerry

  • 0 in reply to Clemens Ladisch
    Expert 2071 points

    Thank you Clemens for that information.  Since ESD events are very short in duration, is the 10 mA spec one that I need to observe?  Much of the energy of the ESD spike will be dissipated as heat in the external resistor.  Assuming a resistor value of 1M, I2R is significant.  If the IC can handle 10 mA continuous, it seems 20 mA for a short duration would be tolerable.

    Regards,

    Greg

  • 0 in reply to Gerasimos Madalvanos
    Expert 2071 points

    Thank you Jerry.  1MΩ is certainly much much greater than any of the resistors in those simulation models.

    Regards,

    Greg

  • 0 in reply to Greg
    Guru 322000 points

    Such a small device does not have much thermal mass; you must never exceed the absolute maximum rating of 10 mA.