INA169: INA169 offset/gain calibration technique with examples

Part Number: INA169
Other Parts Discussed in Thread: MSP430F149

I am using INA169NA/3K chip current measurement IC. I am getting current sense data in my ADC (12 bit) MSP430F149 through that chip. I need to calibrate offset and gain errors of INA169NA chip as all chips do have some errors in them and external components also affect it as well.

My loads are various lamps of various watts, but I am trying to calibrate it using a 10 OHM resistor, so I am trying to get, my board supply 12V to the load. so it consumes 1.2A. There are two modes as per load resistor, low gain and high gain, I want to measure offset and gain in both regions.

gLowGainOffset: 000001

gLowGainADCIsense: 000129

gHighGainOffset: 000008

gHighGainADCIsense: 000549

Now, I am calculating gain, system voltage: 3

LOW Gain: 1.2Amps/((gLowGainADCIsense-gLowGainOffset)*3/4096) = 12.8
HIGH Gain: 1.2Amps/((gHighGainADCIsense-gHighGainOffset)*3/4096) = 3.028
Based on the above gains, If I am using any X load in high gain then I would simply multiply Isense value with the gain, to get the current. For example, ((549-8)*3/4096)*3.028 = 1.2Amps approx, this way calibrated gain can give me accurate current for any other load too.
Amplifier input supply 3.3V, Rs = 15mOhm, Vin+ = 12V

is this the correct way of calibration?

Thanks

  • Hello Divyesh,

    Welcome to the forum and thanks for your question. I apologize for the delay. The website was under maintenance last evening.

    I think there could be some problems with the calibration. Most importantly you need some precise source to consistently achieve an accurate calibration for an amplifier. What is the accuracy of the 12-V source and tolerance the 10-Ω load setting resistor? Calibrated error could be dominated by these instrument limitations. You will want to possible also for temperature if using a resistor.

    Secondly, calibrating gain should really be done with two measurements in the 90% and 10% of the full-scale output range and then linearly interpolating between them to solve for slope (true circuit gain).

    Lastly, how the offset of the circuit was determined is also important. You actually need the two 90% and 10% FSR output measurements in order to determine the true linear offset of the amplifier and both measurements need to be precise enough to meet system error requirements.

    Hope this helps,

    Peter