SBAA591 December   2023 REF54

 

  1.   1
  2.   Abstract
  3.   Trademarks
  4. 1Introduction
  5. 2Flicker Noise Measurement Method
  6. 30.1 Hz to 10 Hz Band Pass Filter Design for Reference Noise Measurement
    1. 3.1 Filter Schematic
    2. 3.2 TINA - TI simulations
    3. 3.3 Component Selection
  7. 4Measurement Setup
    1. 4.1 REF54250CDR Noise Measurement Setup
  8. 5Filter Board Design
    1. 5.1 Schematic
      1. 5.1.1 Schematics Images
    2. 5.2 Layout
    3. 5.3 Bill of Materials
  9. 6References

3 0.1 Hz to 10 Hz Band Pass Filter Design for Reference Noise Measurement

GUID-20230619-SS0I-GS04-JLPK-9MNMKHRC4HBN-low.svg Figure 3-1 Flicker Noise Test Setup Block Diagram

Minimum noise specification of REF54250 is 400 nVpk-pk. To have the measurement error be less than 1.5%, calculate the maximum allowed setup noise as follows:

Equation 1. S e t u p   N o i s e   < =   275 ×   √ ( 1.015 2 - 1 2 )   < =   47   n V p k - p k

The first stage high pass filter’s resistor and first stage pre-amplifier noise (referred to as setup noise) gets directly coupled to the reference noise signal. Consequently, these two elements are the major error contributors to the total setup noise budget. Reference noise is amplified 1000 times in the first stage so the impact from subsequent stages setup noises are minimal. Hence, we assign greater than 98% of the budget to this stage. The high pass filter is designed with a 50 Ω resistor so that KT noise contribution of the resistor is minimal.

Equation 2. F l i c k e r   n o i s e   o f   50   Ω   h i g h   p a s s   r e s i s t o r   n 50   Ω     =   18.92   n V p k - p k  

The next step is to design a low noise pre-amp with a gain of 1000 which can meet the target spec of noise from first stage op-amp. OPA189 is an excellent choice in TI’s portfolio of ultra-low noise and offset op-amps. We need to connect 8 op-amps in parallel to comfortably achieve the target spec of << 56nVpk-pk.

Equation 3. Opamp FlickerNoisenopa=100 nVpk-pk8=35.35nVpk-pk

The gain of 1000 is realized by 10 Ω and 10 KΩ resistors. This equals to 10 Ω in noise contribution. The noise contribution by the 10 Ω resistor gets divided by √8 times.

Equation 4. F l i c k e r   n o i s e   o f   10   Ω   g a i n   r e s i s t o r   n 10   Ω     =   3   n V p k - p k    

Total noise contributed by first stage is RMS sum of equation 1, 2 and 3.

Equation 5. N o i s e   f l o o r   o f   f i r s t   s t a g e   n F i r s t     = √   ( n o p a 2 +   n 50   Ω   2   + n 10   Ω   2 )   =   40.3   n V p k - p k  

Current noise of the op-amp is ignored in this calculation as the impact is insignificant (current noise density = 165 fA/√Hz). Impact of later stage noise gets divided by greater than or equal to 1000 times. Therefore, the impact is not significant. This circuit contributes less than 1.5 % error in REF54250CDR noise measurement.

The second stage is a Sallen-Key high-pass filter of 0.1 Hz with a gain of 10 and damping ratio of 1 which provides 60 dB over all roll-off at 0.1 Hz high-pass. The next two stages are a multi-feedback 10-Hz low pass filter with an overall gain of 10 to provide roll-off of 80 dB at 10 Hz as shown in Figure 4-6. The noise is gained 105 times and captured on the scope over a 10 second window with sampling frequency >> 20 samples a second to avoid aliasing.