Post Go back to editing

recommended bypass capacitor value for ada4522

What are the recommended bypass capacitor values for the ADA4522.

The only thing I see anywhere is 0.1u, located "close" to the power supply pins.

Is this adequate or would additional capacitors (2.2u, 10u...) improve performance.

  • Hi Kevin,

    The value of 0.1 uF is somewhat outdated. Modern conditions impose new requirements. The full impedance value of the standard 0.1 uF does not satisfy the filtering of modern mobile UHF interference. I would recommend trying to position a 100 pF quality ceramic capacitor as close to the power pins as possible. Then a little further to place a 0.1 uF. A little further on, an electrolytic capacitor. There is no exact capacitance value, different experts recommend different empirical numbers from 1 to 47 uF. I find it difficult to answer exactly.

    Do you have digital components in your system? And what is the supply source?



  • I have quiet power supplies (line transformer w/LT1086's). There is very little digital anywhere near the ada4522's, plus it is running on a separate power supply. The board will be shielded & filtered, so there is unlikely to be much RF present.

Reply Children
  • Looking at the PSRR for ADA4522 graph we can conclude that the positive rail is the weak point here. Since the LT1086 does not impose restrictions on the output capacitor, you can suppress low-frequency interference with good modern LOW-ESR electrolytic capacitors. Start with 47 uF, for example. Note that the data sheet recommends tantalum capacitors, but in my opinion this is also an outdated recommendation. Tantalum capacitors have a low HF impedance due to design features, but now there are alternative offers for them. 

    Study the specification for your chosen capacitor, and learn where its impedance will become too high in the frequency domain. Reinforce this area with a ceramic capacitor. Use the graph in this article for the initial selection of capacitance, but then study the specifications to find out how far from the PSRR range is the self-resonant capacitor. My intuition says that it should be at least a decade higher in the frequency domain.

    Will this help you?