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.

INA138: Is it possible to increase differential input voltage range by adding resistors in series with In+ and In-?

Part Number: INA138

The data sheet advises that the maximum differential input voltage that gives linear measurements is 500mV, giving an output of 100uA. If I wish to measure a larger differential voltage, it seems to me that if I add 20k precision resistors in series with In+ and In-, I can now measure up to 2.5 volts for the same 100uA linear output current range. Apart from the obvious issue of degraded common mode rejection from mismatched resistors, are there any issues with doing this?

The X-Y problem: I'm designing a variable high voltage power supply using a linear pass element, and a tracking switching preregulator to minimize dissipation in the pass element. My basic design is to use the linear element to control output voltage directly, and use the voltage across the pass element as the feedback for the switching preregulator. Thus, the preregulator output will track the output of the linear element, plus a constant dropout voltage programmed by the components in the feedback loop.

The power supply output voltage is high enough that I'm looking at using the technique outlined in your TIDU833 application note to pass the measured dropout voltage from the high side pass element down to the low side where the switching regulator lives. I realize I can adapt the technique to work with a normal voltage-output instrumentation amplifier but the vbe of the common base bjt adds output offset which makes it a bit more of a pain so I'd prefer to avoid that if I can.

  • Hello Stephan,

    Thanks for considering to use Texas Instruments. We will have a response to your issue on the following business day (starting 8:00AM Monday, U.S.A Mountain Time)
  • Hi Stephan,

    There is nothing preventing you from adding a pair of resistors to the input, thus changing the system gain. Sounds like you’re fully aware of the potential drawbacks of doing so.

    The technique outlined in the application note is tailored toward current out devices, if you have a way of making it work with a voltage out devices, that’s great. Otherwise you may consider isolated current sensing techniques.

    Regardless, I would pay attention to loop bandwidth and make sure selected current sensing device/architecture can satisfy the requirement and provide stable operation.

    If you don’t have it already, you might want to try TINA-TI to verify your circuits as you proceed.

    Regards, Guang

  • Hi Stephan,

    We haven’t heard back from you in a while, if you have further questions please let us know. Otherwise we’ll consider you have all the information needed and the case closed.

    Regards, Guang

  • Hi Stephan,

    the usual argument against such an approach, is that the internal resistors on die can have a different temperature coefficient compared to the external resistors. Also, sometimes the resistors on die might be very precise relative to each other, but can show an absolute error of up to 20...30%, due to wafer production tolerances.

    But here, according to section 7.1 of datasheet, the internal resistors seem to be "precision thin film resistors trimmed in production to an absolute tolerance". So, adding 20k resistors to the inputs might work.

    Kai