Use ferrite cautiously in precision analog power supplies
Tips to use of ferrite in precision signal conditioning.
Case Study:
Recently, I was working with client for modifying his existing precision analog design. While my job was to work on analog designs but my eyes caught use of ferrite in input of high PSRR LDO which was powering this precision analog chain.
As ferrite was paired with decoupling capacitors, I was curious to know if that was causing any resonance.
Running the quick simulation it revealed there is huge resonance near 78K. Now that was serious as the SMPS used has near 50K to 125K variable switching (as per load).
A time domain simulation with input signal of 78K at 50mVp-p gave output of 264mVp-p.
Doing practical troubleshooting, original SMPS had near 60mVp-p noise+ ripple which was getting amplified to near 220mVp-p before input to LDO. The high PSRR LDO was doing its job & output of LDO had 6mVp-p ripple.
Damping the ferrite with given circuit reduced input of LDO to 26mVp-p & output to near 0.8mV. Please noteferrite’s L+ 1uF further reduced the ripple & a drastic improvement in ripple & noise needed for precision signal conditioning. The images below shows the modified circuit & response
The client noticed the impact on ADC reading accuracy & stability both.
Ferrites are usually good arsenal against EMI but sometimes designers unknowingly create unwanted things along with other circuit components & their parasitics.
So the next time you use EMI ferrites in any precision supplies, please check & analyze well if that is not harming the accuracy of the signal chain.