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.

INA186: estimate error due to VREF source resistance

Part Number: INA186

I have a group of current sensors all connected to the same +Vs/2 output reference voltage (Vref) source

Due to wiring, a maximum of 1 Ohm series source resistance (Rs) has to be taken into account

The value R of the four resistors used in the differential output amplifier of the INA186A1 are not specified in the data sheet

If Rs << R, then Vout = Vref - Vref x Rs/2R + Vcm x Rs/2R

Vcm is the first stage common output voltage. The error is zeroed if Vcm = Vref. The worst case is met when Vcm is close to either Vs rails.

At Vs = 5 V, the output voltage error is less than 0.125 mV if R is greater than 10 kOhms (with Rs = 1 Ohms), meaning 0.005 mV input error at a gain of 25.

This could be acceptable for a 12 bit acquisition system as it only represents 50 ppm of the of 2.5 V full scale

Please confirm the lowest possible R value for the INA186A1 and the impact of 1 Ohm source resistance on Vref

Thank you

Francois

  • Hey Francois,

    Welcome to the forum. Thanks for the question.

    Exactly how the first stage capacitively couples the differential shunt voltage is quiet complex so it is not straightforward to derive a transfer function when REF pin is loaded down with resistance. The second stage feedback resistors (R) should be much greater than 10kOhm, more in the range of hundreds of kOhms, even for the A1 variant.

    Anyway, based upon my experience with the device, a1-Ohm VREF source impedance will have a negligible effect on device performance. REF current data from this application note shows that when INA190A5 (same amplifier topology as INA186) is referenced to mid-supply (Vs/2), that the current sunk (going into) REF pin is only ~30nA. Even if we assume an increase x20 when going from A5 (500V/V) to A1 (20V/V), the REF current is still only 600nA which would result in a 600nV offset due to 1-Ohm of VREF loading.

    Now obviously there will still be theoretical gain error and degraded CMRR, but these should still be negligible for 1-Ohm.

    Sincerely,

    Peter Iliya

  • Very useful answer, thanks Peter, less than 1 uA into/from VREF terminal will obviously not impact the design !