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.

ADS1192 External Voltage Reference with Internal Reference Buffer

Other Parts Discussed in Thread: ADS1192, REF5025, OPA350, OPA333, THS4281

Hello

I have a question if ADS1192 can configure to use External Voltage Reference using Internal Reference Buffer?

The idea is to connect External VREF like REF5025 (better performance than the internal one) to VCAP1 and enable Internal Reference Buffer (Set CONFIG2 Bit 5 ON and Bit 4 is OFF).

I found that block diagram of the internal reference, Figure 26 on page 22 in the datasheet. The block has R1 after internal Bandgap (100kohm) and if 2.5V voltage is applied to VCAP1, internal buffer operate the 2.5V to VREFP without disturbing and be disturbed by the bandgap.

I would like to save external reference buffer amp.

Best regards,

Masa 

.

  • Hey Masa,

    Unfortunately that configuration is not supported. The device was not designed to have an external voltage applied to the VCAP pin. Its purpose is just to free the internal reference buffer input from high frequency noise. If the internal reference buffer is powered on, the internal bandgap reference must be used. If you decide to stick with an external reference, let us know if you need help choosing a buffer.

    Regards,
    Brian Pisani
  • Hi Brian,

    Thank you for answering.

    I am working with below.

    https://e2e.ti.com/support/data_converters/precision_data_converters/f/73/t/412866

    The application requires low power consumption and OPA350 5.2mA Iq is not acceptable.

    Considering a composite of THS4281 and OPA333 it will reduce Iq (750uA + 20uA) however it still not be enough.

    I have a question,

    if only OPA333 (20uA Iq) is used for REFBUF, how much performance will be lost.

    I simulated TINA below.

    Frequency performance was not good but noise.

    The external VREF is planned to use 2.5V+/-0.175%(max25 ppm/C).

    I wonder if the VREF performance will be lost by OPA333 and may be no benefit not to use internal reference (2.42V+/-0.5%(typ 45ppm/C)).

    (U

    Best regards,

    Masa

  • Hey Masa,

    I would not recommend using the OPA333 because in addition to the noise performance at high frequency, it has a relatively low unity gain bandwidth in comparison to the OPA350. It may seem counterintuitive that an amplifier which is used to drive a DC voltage needs a high bandwidth, but with a delta sigma ADC that is the case. Then the ADC samples, it draws transient current from the reference. The capacitors on the Vref pin are used to meet that momentary demand in current, but they also must be replenished quickly such that the voltage at the reference pin is stabilized before the next sample. The modulator samples the input at a rate of 512 KHz on the ADS1192 to the reference buffer amplifier must have a significantly higher bandwidth than that.

    Clearly this exemplifies a trade-off between power and performance when considering how to drive the ADC reference. I'd like to point out that one added benefit of using the internal reference with this device is the "worry free" aspect with using it. I'd advise doing some bench performance testing with the internal reference and see if you can get away with using that option.

    Regards,
    Brian Pisani
  • Hi Brian,

    Thank you very much fo your advices. i am understanding that 10uF cap is the real reference driver and REFBUF requirs much hiigher frequency to fill it. 

    The external refernce with high speed buffer requires current consumption. then need to backup to the worry-free internal refernce.

    An issue of the internal reference is temperature drift, 45ppm/C. I looked at  Internal reference vs Temperature graph (Figure 2 on page 11) in ADS1192 Datasheet.

    Is it realistic to compenastae the voltage drift by temerature measured because the curve is very linear? And is there any design recomenndation to reduce the internal reference temperature drift? 

    Best regards,

    Masa  

  • Hey Masa,

    Great questions! One possible strategy for compensating for the temperature drift in the internal reference would be to measure the internal reference voltage across temperature for a few devices (as many as you deem to be statistically significant). Then hard code the average reference voltages for different temperatures into your application processor. If the application is capable of measuring temperature by some other means, then you can use the hard-coded reference voltage data to compensate for the average device drift in the final application. Does that seem like a workable solution?

    Brian Pisani
  • Hi Brian,Thank you very much. Masa