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: LMH6643

Part Number: LMH6643
Other Parts Discussed in Thread: STRIKE

Hello everyone, and happy new year!

So, in relation to this topic, I would like to make more or less the following schematic same ESD protection circuit (TVS + Ferrite Bead + Schottky diodes).

- The power supply of my AOP is V+=12V and V-=GND

- Output voltage = 10V

- Same ESD protection circuit as mentionned (TVS + Ferrite Bead + Schottky diodes)

- Important: The load on the AOP needs to draw up to 20mA (which is covered by the LMH6643) 

I am still not sure about the current limiting resistor which is placed between the AOP output and the schottky diodes. The value of this resistor, in order to be under 10mA on the output pin, would limit the current that the load needs.

Do I have to limit this current, or is it sure that the AOP can resist a surge of current that would occur during the ESD tests?

Thank you!

  • Hi Mariem,

    I cannot see a schematic. Please post again.

    Kai
  • Hello Kai,

    Thank you for responding.

    So it is a basic circuit, I did not choose the components yet but here is an approximation:

    - TVS Vrwm=12V, Vclamp=19.9V (and Vbr=13.3V)

    - Schottky: I was thinking about BAT54

    I might be calculating in a wrong way but isn't  Io=(Vtvs-0.8)/Rlimit ? (Io the current in the outpout pin)

    I was thinking Vtvs=19.9V in case of ESD test. In that case, Rlimit should be around 2k

    Here is an approximatif schematic.

  • Hi Mariem,

    you use the LMH6443 which is a HF OPAmp. Such a HF OPAmp does not like any load capacitance at the output. The resistor placed directly at the output of LMH6643 can isolate some load capacitance and stabilize the OPAmp. How much resistance can you afford here?

    What is your signal bandwidth? Can your circuit really accept an output capacitance of 100nF? And don't forget that the SMJ12CA also has a big junction capacitance.

    Kai
  • Hi Mariem,

    Are you trying to protect the output of the amplifier from an ESD strike? If a positive ESD is dumped on the output node, the high speed amplifier will quickly sink it. The output is designed to cancel-out errors and drive high speed transients. Usually you have to protect the amplifier inputs, which have no ability to dissipate charge, but high speed amplifiers naturally protect their own outputs. You can run a simulation with a switched capacitor to see how much discharge it can really take.

    Best regards,

    Sean
  • Hello Kai,

    I only need an AOP that can amplify a signal up to 10V and provide 20mA.

    I took this AOP because it meets the characteristics that I need: Slew rate> 50V/µs, GBW>40MHz, rail-to-rail.

    My question is, what is the resistance value that I need to put on the AOP output in order to protect it (as the current that is sinked by the internal leds can not be >10mA) while also not preventing me from drawing the 20mA needed.

    Thank you!

  • Hello Sean,

    Yes I am trying to protect it from ESD Strike, in a circuit that provides a 0-10V, 20mA output.
    So if the AOP can protect it's own output, that means I can put a 100R or a little more resistance value to bring some sort of isolation of the capacitive load, even if it doesn't help reduce the current that goes into the output down to the needed 10mA as said in the datasheet (in case of ESD)?

    Thank you!

  • Hi Mariem,

    I have given an answer in a similar post which could be interesting for you:

    e2e.ti.com/.../753821

    By the way, an OPAmp which isn't powered, cannot protect itself via the output stage, because the output transistors are turned-off then and cannot sink any current. So, your Schottky diode clamps from the output of OPAmp to the rails are certainly helpful.

    Kai
  • Hello Kai,

    Thank you it is very helpful. I think I will put a 75R or 100R resistor as you've suggested in that post.

    Just one last question. I think there is an error in my schematic. Shouldn't this resistor (Rlimit) be inside the feedback loop in order avoid a voltage drop, and should I take it into consideration for the calculation of the feedback resistor value?.

    Thank you again.

  • Hi Mariem,

    if you want to put the current limit resistor into the feedback loop of LMH6443 you would need to very carefully compensate the OPAmp. You would need to restore the phase margin by the help of well known "dual feedback" circuit discussed in this training video:

    training.ti.com/ti-precision-labs-op-amps-stability-6

    Kai
  • Kai,

    Thank you that was perfect.
    Your answers as well as the training videos you referred me to were very helpful and well explained.

    Thank you again for your help!