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.

REF5020 noise spec

Other Parts Discussed in Thread: ADS8320, OPA353, OPA350

Hi,

We are planning to use REF5020 as a reference input to characterize CT front end device for which integrated noise is a very critical spec. So we want to know REF5020 integrated noise till say 10M. I simulated REF5020 model using TINA and found it reaches to 33uV as shown in below plot which seems to be quite high. So I was wondering whether noise has been modeled correctly or not. So could any one have idea about integrated noise of REF5020 till 10MHz?

Thanks

Regards,

Shabbir

Design engineer

Texas instruments, India

+91-9243532152

  • Shabbir,

    First, the noise performance of REF5020 is a function of a capacitor placed on TRIM/NR pin (see below) where adding 1uF capacitor decreases the overall output noise by a factor of 2.  Also, adding RC filtering at the output with corner frequency below ~60kHz will decrease the total noise-see voltage noise spectral density simulation plots below.  The REF5020 macro-model provided on TI website may not be precise but is relatively close to correctly simulate the output voltage noise.

    I have simulated the REF5020 output voltage noise spectral density with CNR of 0 and 1uF - see results below;

    The integrated noise for the two cases is shown below: ~105uV for CNR=0 and ~50uV for CNR=1uF.

    Of course, the simulations above do NOT include the effects of any additional RC filtering at the REF5020 output and thus any such additional pole below ~60kHz will result in a decrease of the total noise of REF5020. 

  • Marek,

    Thanks for your quick reply. It seems TINA model and your simulation results matches very nearly. Initially we were avoiding putting any filter at the reference output because our device takes some current from reference and because of which putting resistance in the path will introduce some drop in the reference and will result in some error. But it seems we have to put external filter to avoid noise. We will analyze it more and will see possibility of putting any buffer which will provide device current.

    Thanks

    Regards,

    Shabbir

  • What is a better choice for the reference voltage for the 16-bit ADC converter ADS8320 chip, REF5020I or REF3220 or LM4040A20?

    thanks in advance
    Sasa
  • Sasa,

    REF5020 has the lowest noise and lowest drift but nevertheless requires op amp driver for filtering and speed in order to obtain ADS8320's optimal performance.  The best op amp for the driver in such application is OPA350 or OPA353 capable of direct driving capacitive load in excess of 10uF.

     

  • Thank you for your prompt response.
    I deliver to you the basic scheme, based on your suggestions. What is your opinion.


    Regards, Sasa

  • Sasa,

    You will most likely need to add filtering to optimized the solution for ADS8320 - please see below.

  • Marek, thanks for the advice.
    Best regards,
    Sasa