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.

XTR106: XTR106 Internal Bandgap Reference Voltage?

Part Number: XTR106

Hello,

I'm attempting to work out the tuning functions of some of our heritage sensor circuitry. One of the methods we use to tune sensor span/gain is to adjust the bridge excitation voltage [pin 14 (Vref 5)], by having a couple circuit elements return current to pin 13 (Vref 2.5). Per page 4 of the datasheet, the XTR106 uses an internal combination of a bandgap voltage reference and a voltage divider to determine the pin 14 output voltage (our bridge excitation). Could I please get a typical voltage/resistances for that bandgap reference and the voltage divider?

Thanks,

Brian

  • Hi Brian,

    If I understood your question correctly, you are requesting the typical bandgap voltage and the typical values of the internal resistors at the feedback on the REF Amp on the simplified Functional block diagram of page 4, in order to adjust the excitation voltage of the bridge sensor using external resistors. Is this correct?

    The XTR106 can only provide an accurate 2.5V (by shorting pin 13, pin 14) or 5-V reference on pin 14. However, the reference output is not really intended to be a an adjustable excitation bridge reference beyond these two voltage values. The XTR106 internal resistor ratios and their internal resistor drift is very well matched among themselves, where the internal resistors are ratiometric, achieving the 0.05% reference output accuracy and 20 ppm/C drift. However, the problem is that the internal resistors absolute accuracy will vary significantly from device to device and lot to lot variation (the internal resistors absolute accuracy could vary up to 20% over device and process variation).

    Therefore, when attempting to use precision external resistors to adjust the reference, although the external precision resistor absolute value may be accurate, the absolute value of the internal resistors may vary up to 20% over lot to lot process / device variation, and this mismatch in the ratios and drift between the external components and internal resistors will cause the reference to have significantly degraded accuracy and drift. 

    A possible solution may be to use two resistors in series with the bridge sensor, with one resistor on top and one at the bottom of the bridge, to adjust the excitation voltage seen by the bridge sensor (similar to figure 5).

    Thank you and Best Regards,

    Luis

  • Hi Brian,

    if a datasheet does not contain further going data, this usually means: Don't touch, only use as intended by the manufacturer.

    In this case the internal resistances cannot only show huge production tolerances, as already mentioned by Luis, but the bandgap with the buffer is also part of the complexly working linearization circuitry. Also keep in mind, that the functional diagram of datasheet only shows a very simplified schematic. The shown resistances might not exist in this form.

    So, I would leave it untouched.

    Kai
  • Hello Luis and Kai,

    Thank you both for the replies. Luis, your understanding is correct. Though, we also use diodes for a thermally varying forward voltage, this adjusts the reference voltage thru temperature (which is then fed to Vref 2.5). There is minimal original design documentation, however I believe the motivation for going this route might have been a more linearized response thru temperatures, while also reducing the impact on the Wheatstone bridge excitation. As you mentioned though, this comes with major impacts to overall accuracy and long term stability.

    Thank you both very much for the help!

    Brian

    PS Kai, I agree; and, for a new design I'd definitely do it differently.