INA500: INA500B differential impedance tolerance and drift over temperature

Hi Robert,

In my design I have resistors in series with the amp input, so there is effectively a voltage divider between the input differential impedance of the amp and the mentioned external resistors. The voltage at the amp's input terminals depends on the input differential impedance.

So, unrelated to the gain error of the amp, there is an error in the division ratio. The input diff impedance has high tolerance (-+15%), and the system is 12bit accuracy, so this error is not negligible.

I am considering calibrating the analog errors at room temperature, therefore I would like to differentiate between the unit to unit error (in room temperature), and the temperature drift.

Can you supply more information regarding the differential input impedance tolerance? Either as min max, or standard deviation.

Does the -+15% refer to min and max across temperature, or in room temperature?

Is there any data regarding temperature drift of the input differential impedance?

Is -+15% a tight tolerance?

Thank you,

Amit.

Hi Amit,

Amit Halfi said:

So, unrelated to the gain error of the amp, there is an error in the division ratio. The input diff impedance has high tolerance (-+15%), and the system is 12bit accuracy, so this error is not negligible.

I'd like to clarify that the datasheet says that the resistor values can vary together up to +/-15%. One internal resistor will not shift +15% and another shift -15%. 

I'll ask the design team about this. 

In the meantime, how much resistance are you planning on adding to the input? 

Best Regards,

Robert Clifton 

Hi Robert,

The series resistance is significant, compared to the 2.88Mega Ohm of the amp and the required accuracy. On the order of hundreds of kilo Ohm.

I would like to emphasize that I don't care about each resistor in the amp, I care about the "total" differential input impedance,

Specifically, room temperature tolerance and temperature drift, either as min\max or standard deviation.

Thanks, Amit.

Hi Amit,

I'm still waiting to hear back from the design team.

When you are using hundreds of kilo ohms on the inputs, we typically recommend using an instrumentation amplifier instead to eliminate the distortion caused by any mismatch on the input resistors as well as general gain and CMRR degradation that will occur. 

Is there any reason that you decided to use a difference amplifier rather than an instrumentation amplifier? 

Best Regards,

Robert Clifton