3 Replies Latest reply: Jul 22, 2014 7:28 AM by EstiS RSS

    CN0191: 20-Bit, Linear, Low Noise, Precision, Bipolar ±10V DC Voltage Source

    CFTL

      Add your comments about the Circuit Note found at http://www.analog.com/en/circuits-from-the-lab/CN0191/vc.html.

        • 1. Re: CN0191: 20-Bit, Linear, Low Noise, Precision, Bipolar ±10V DC Voltage Source
          CFTL

          Questions or comments about CN0191, leave us your feedback here.

          • 2. Re: CN0191: 20-Bit, Linear, Low Noise, Precision, Bipolar ±10V DC Voltage Source
            eyeow84

            I am looking for high precision and great consistency millivolt source design reference for thermal couple sensor. This circuitry design is pretty same as my current design.

             

            I have some questions below:

            1) Is this circuit applicable or fine-tuned to output voltage reference source range like +2V to -2V instead of +10V to -10V?

            2) The AD5791 has a good linearity error long term stability in the specification. It could achieve about 0.11~0.19 LSB after 1000 hours at specified room temperature. In real application, will the output op amp AD8675A contribute any impact to the output consistency?

             

            Thanks,

            eyeow84

            • 3. Re: CN0191: 20-Bit, Linear, Low Noise, Precision, Bipolar ±10V DC Voltage Source
              EstiS

              Hi eyeow84,

               

              Once the voltage reference values are in the ranges specified in the datasheet, the DAC should work as expected. We cannot guarantee the functionality of the part at values out of specs.

              A positive reference voltage range of 5V to VDD-2.5V, and a negative voltage reference voltage range of VSS+2.5V to 0V are possible reference inputs for the AD5791.

              The value of this voltages and the configuration of the output amplifier allow the user to obtain different output voltage spans (Unity Gain/ Gain of Two Configuration).

               

              Looking at the reference/output buffers, there should be no major issues/impact on the output.

              • The AD5791 is linear to 1LSB over its entire operating temperature range, the temperature drift of the amplifiers should not be large to impact the performance of the AD5791 in any way, the low 0.6uV/C max drift will account for approx. 0.1ppm/C error in the system.
              • Input bias current of 2nA max is low so the effect on INL is negligible.
              • To complement the low noise of the AD5791, complementary products should have NSD specs less than 9nV/rtHz. The AD8675/AD8676 offers 2.8nV/rtHz @ 1kHz.

               

              Regards,

              Estibaliz