• State Resolved
  • Locked Locked
  • Replies 3 replies
  • Subscribers 48 subscribers
  • Views 1336 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 Input Protection Diodes

shadders
Expert 1065 points

Part Number: OPA192

Hi,

I want to confirm that the OPA192 opamp will require input protection for the inputs when powering up the device.

I am using the opamp as a DC servo in a class A/B amplifier. After investigation, the opamp is outputting 6volts on the negative input. I had replaced one of the previous back to back 10volt zeners as it seems to have failed.

I am using +/-15volts to power the opamp. What may have happened is that I applied the +/-30volts to the power amp, which applied a low current signal with 10uA magnitude, at 10volts to the negative opamp input without power being applied to the opamp.

I can see from the datasheet absolute ratings that you must not apply to any input, more than 0.5volts greater than the power supply. Obviously this has occurred, but at low current = 10uA. I examined other posts :

"https://e2e.ti.com/support/amplifiers/f/14/t/623990?tisearch=e2e-sitesearch&keymatch=input%2525252525252520protection%2525252525252520opa192"

From the post, it is not clear whether the 10mA maximum current limit and +/-0.5volts greater than +/-Vsupply has to be met simultaneously. That is, can you exceed +0.5volts greater than the supply, as long as the current is below 10mA ??

As such, I assume that the cause of the opamp failure is the excess low current voltage, and to mitigate this in the future, that schottky diodes should be added to the opamp inputs, connecting to the power supply inputs of the opamp ?

In the final build of the amplifier, I can ensure that the power supply for the DC servo is applied to the opamp before any power amp power supply.

Thanks and regards,

Shadders.

  • 0
    TI__Guru* 93105 points

    Hello Shadders,

    You mention the input current was limited to 10 uA when applied to the OPA192 inverting input. If the polarity was positive, then the upper ESD diode going from the input to VDD as shown in datasheet Figure 65 could turn on. If forward biased the diode would provide current to the VDD pin, and anything that had in the circuit that could provide a return path to ground. Ten microamps shouldn't damage anything in the OPA192 input circuitry as the specifications allow an input current up to 10 mA.

    I can see from the datasheet absolute ratings that you must not apply to any input, more than 0.5volts greater than the power supply. Obviously this has occurred, but at low current = 10uA.

    The 0.5 V limit is in place to prevent a user from forward biasing the internal ESD diodes which are not intended to be turned on in use. The diodes won't allow 10 mA to flow with 0.5 V of forward bias. They would actually require more than 0.5 V to conduct to a level that could support 10 mA. So the critical limit here is the 0.5 V.

    As such, I assume that the cause of the opamp failure is the excess low current voltage, and to mitigate this in the future, that schottky diodes should be added to the opamp inputs, connecting to the power supply inputs of the opamp ?

    See if you can identify some low leakage Schottky diodes to use at the inputs. Schottky diodes tend to have high leakage current and can ruin the low bias current of the OPA192 when they are connected to the op amp inputs. Since the Schottky diodes have a lower forward voltage compared to silicon switching diodes they will turn on at a lower voltage than the internal ESD diodes.

    The Schottky diodes are only part of the solution. It is necessary to add a transient voltage suppressor (TVS) diode from each supply pin to ground. They need to be selected such that they remain off during normal powered operation, but turn on before the supply pin maximum voltage is exceeded. These can be small, lower current TVS diodes because your circuit doesn't involve a lot of power.

    Regards, Thomas

    Precision Amplifiers Applications Engineering

  • 0
    Guru 140200 points

    Hi Shadders,

    can you show a schematic?

    Kai

  • 0 in reply to Thomas Kuehl
    Expert 1065 points

    Hi Thomas,

    Thanks for the reply. You have confirmed that there are diodes on the input, and any voltage appearing on the input greater than the power supply will cause the diode to conduct.

    I have simulated the 30volts input through a 1meg resistor to the inverting input, and i can see the 0.6volts diode voltage at 30uA current through the 1meg resistor. I have used an LT1028 as opposed to the OPA192, as i do not have the model - and have not had the time to create.

    It seems that the zener diodes do not conduct given that the current is insufficient (the ones used require 50uA). I set the voltage sources for the opamp to zero to simulate the condition i thought caused the fault.

    I have located low leakage diodes - BAS70 with 100nA leakage current.

    When the power supply is connected to the DC servo circuit - there is the 1meg ohm input, and series 2x 500kohm resistors as per the circuit - which from the simulation shows that the 30volts input to the circuit will not cause a fault - 0.6volts seen at the input, 30uA current.

    The opamp is faulty as it is providing 6volts approx on the negative input to the zener diodes - and the zener diode is getting hot - the previous one must have burned out. The opamp is not getting hot. I am not sure how this has occurred, but the opamp is faulty - so, in the future build i can add the low leakage diodes - although, it does not seem to be needed.

    For the transient voltage suppressor - i can add a simple zener.

    From the circuit - can you see if there is any condition that the simulation is not processing that could cause 6volts approx to be output from the negative input etc ?

    Thanks and regards,

    Richard.