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.

5 Volts regulator with 0.001% or best accuracy / precision

Other Parts Discussed in Thread: ADS1115, LP2951

Hi everyone !

I'm currently using a Raspberry Pi 3 with two ADS1115 ADCs to take readings from a magnetic fluxgate sensor (FLC-100 from Stefan Mayer Instruments). Each sensor does have a typical consumption of 2 ~ 5 mA.  I do use three sensors. I'm using two channels of each ADS1115 to measure it in single-ended mode (that's why i'm using two ADC chips).

My problem is that yet with the best Raspberry 3 switched power suply available (model T5875DV,  from Stontronics) i'm getting some variations on the 5V supply for the three sensors (the decimal separator here in Brazil is the Comma):

The specs for the power source that i'm using, are:

Output Regulation:   +/-5%  
Line Regulation:      +/-2%
Ripple & Noise:       120mVp-p Maximum

As i've measured by looking at the chart, seems the Power Source is doing very well for it's specs...but unfortunately I do need more precision in my academic project because the output of the sensor is in voltage too, so a little noise on it's supply can mess up the magnetic readings .

I do need a maximum of 0.0001 Volts of variaton on the power source, so anyone knows of way to regulate this 5Volts output from my power source to a maximum of 0.001% accuracy / precision or better ?  It does need to have at least 35 mA of current output available to keep the current consumption from the three sensors with safety, being possible to power the two ADS1115 too. 

  • Fabio,


    I'm not sure that you would be able to find a voltage reference (or regulator) with 0.001% accuracy very easily. For a 5V reference, this would be an error of 50uV. I don't think you could find one readily available.

    A better solution would be to pick a low-drift reference, with a worse accuracy specification, but perform a calibration, accounting for the reference error with some scaling in any ADC measurement.


    Joseph Wu
  • Joseph,

    First of all, thanks for your answer...

    You say to choose a IC like REF5050 with 0.05% Accuracy ? In this datasheet:
    http://www.ti.com/lit/ds/symlink/ref5020.pdf

    It looks like to have 10 mA of output. I can use three of these, one for each sensor.

    Other one i've seen and it's available to purchase here in my region is the LP2951:
    www.ti.com/.../lp2951-n.pdf

    Looks like to have 0.5% accuracy.

    Wich one looks better for my case ?

  • Fabio,


    As I mentioned in the previous post, you want to consider the drift of the reference as well. The REF5050 has a drift of 3 or 8 ppm/C max (depending on high or standard grade). The LP2951 has a drift 100 to 150 ppm/C max.

    You should also consider the noise specification. Both list a certain amount of noise, within a certain bandwidth. I don't think the two report the noise for the same bandwidth, but I'm reasonably certain that the REF5050 has better noise performance as well.


    Joseph Wu