0

enter image description hereI have an op amp circuit using the OPA2626 that has a built-in input/output ESD protection of +/-3000V human body model. The circuit is 3.3V single supply, non-inverting, with only an external battery connection. However, it can have a large electromagnetic input to the amp due to a parallel coil with magnetically induced voltages of 10-40mV in normal operation. The main threat may be a magnetically induced voltage spike at the sensor, but do not know at what amplitude. The sensor may see some electrostatic spikes also from nearby objects albeit will have some ferrous-like shielding.

The op amp circuit input is low impedance under 100 ohms. The op amp circuit output is clamped with a zener at +3.3V and continues into an AVR (Atmega) digital pin that is pulled down. The input for the Atmega is between 10 KOhm-100 KOhm.

What add-on ESD should I consider for the input and the output?

Found this excellent post, but still a bit confused about the solution for the op amp w/ESD vs op amp w/out ESD as far as protecting that mA amplitude gap below the ESD's operating range:

"Therefore, the most we would need to do is limit the ESD diode current to <100mA and all will be well. If we suppress the 8kV to about 250V (as shown by the test curves above), then as the supply is negligible relative to that, an input resistance of 250V/100mA = 2.5k would do the job. Note, however, the source of our input bias current: The amplifier input bias current is the leakage current of these ESD diodes. This leakage is a function of the temperature and common mode voltage of the amplifier, as shown in the Typical Performance Chacteristics."

ESD protection for OpAmp inputs and outputs

Edit: added the circuit. L1 will have a ferrous EMI protection from the outside influence. The output of the op amp is limited to the upside at 3.3V via zener diode.

TommyS
  • 135
  • 10
  • YOu have an EMI crosstalk problem and inadequate shielding, isolation, orientation. Why do you ask aboutESD? We call this an XY problem – Tony Stewart EE75 May 27 '20 at 21:58
  • There are two threats: ESD and EMI-induced voltage spikes. I believe both could lead to higher voltage/current spikes effecting the same parts of the circuit mainly the OP AMP and the AVR. So far the circuit is on a breadboard and functions OK. But I hope to avert any problems once built that could come from the way of the sensor or from stray electrostatic discharge. – TommyS May 27 '20 at 22:10
  • 1
    ESD is an impulse than can be attenuated with a ferrite bead, or CM choke or choke. This raises Z(f) to limit current into Zin=100 ohms. But if you added small pF then you have a 2nd order filter. But why not fix the EMI issue and define the ESD environment instead if that is your only concern. (Eg. Cable length, connection, dust level , floors RH, etc, HBM model or cable model? ESD and EMI can be Common Mode and/or Differential. Depending on your physical system and Earth bonded or floating etc... – Tony Stewart EE75 May 27 '20 at 22:14
  • Thank you for the suggestions. I added the proto-circuit diagram to the original post above. The cable length will be only a couple of inches and shielded. No earth-ground. Ambient between -40F-150F (direct sun). RH cannot be controlled. – TommyS May 27 '20 at 22:31
  • 1
    Zin is actually 27k but reduced. By negative feedback and gain.. Raise Rin to limit current to desired level with clamp diodes from Vin- to Vin+ to form a true current limiter. – Tony Stewart EE75 May 28 '20 at 00:36
  • 1
    Also examine CM crosstalk – Tony Stewart EE75 May 28 '20 at 00:56

0 Answers0