Filter Caps

Discussion in 'General Electronics Chat' started by ELECTRONERD, Jul 19, 2009.

  1. ELECTRONERD

    Thread Starter Senior Member

    May 26, 2009
    1,146
    16
    Hello,

    For power supplies, after the rectifier diodes filter the AC, the capacitor smooth's out the ripple. I've heard that the larger the capacitance value, the better. Is this true? I suppose a applicable amount would be needed until you don't need any more capacitance.
     
  2. hgmjr

    Moderator

    Jan 28, 2005
    9,030
    214
    The sizing of such a capacitor typically depends on the amount of current demanded by the load since it is the load that will discharge the energy that has been stored onto the capacitor during the charge cycle. Erring on the side of larger-than-needed is certainly preferable to erring on the side of too small.


    hgmjr
     
  3. millwood

    Guest

    the issue is actually extremely complicated and requires a lot of highly specialized knowledge to analyze.

    on the surface, higher capacitance, under a given load condition, means smaller ripple on the supply rail. that part is always true.

    however, the higher capacitance also means the diodes are conducting during a shorter period of time. That means much higher peak current through the diodes, with much more high frequency content.

    This creates interference problems on the rail and throughout the amp, which typically have very strong psrr at mains frequency at practically none at high ferquencies (>1 or 10k).

    and if you factor in parasitics, or magnetic imperfection, the issue is even more difficult.

    so you basically trade off between low capacitance thus high ripple content that your amp can reject and high capacitance thus low ripple content but more high order harmonics that your amp cannot reject.

    I usually stay around the rules of 1000uf capacitance for every 10w of output power, and dampening of diodes. If I am working on a high priority amp, I will use soft recovery diodes as well.
     
  4. ELECTRONERD

    Thread Starter Senior Member

    May 26, 2009
    1,146
    16
    Thanks for the quick replies. Now I know more about power supplies.
     
  5. millwood

    Guest

    power supplies, especially the simple unregulated types, are also some of the most difficult ones to understand and analyze. I typically go by my rule of thumb approach, :)
     
  6. ELECTRONERD

    Thread Starter Senior Member

    May 26, 2009
    1,146
    16
    Whew! You answered fast! Your definitely on the ball ;). Thanks a lot millwood, I'll keep that in mind.
     
  7. ELECTRONERD

    Thread Starter Senior Member

    May 26, 2009
    1,146
    16
    I have another question:

    What would be an acceptable amount of ripple if I am going to have 5A of current?

    Thanks!
     
  8. Jony130

    AAC Fanatic!

    Feb 17, 2009
    3,957
    1,097
    Since C=Q/V=(I*t)/V so we can do approximate calculations of a capacitance:
    Capacitor will be discharge by half of the period 1/60Hz=16.7ms---->8ms
    The discharge current is equal 5A and ripple voltage, say 4V (I do not know the output voltage)
    C=(5A*8ms)/4V=10000uF
     
  9. ELECTRONERD

    Thread Starter Senior Member

    May 26, 2009
    1,146
    16
    The output voltage will be 24V
     
  10. Jony130

    AAC Fanatic!

    Feb 17, 2009
    3,957
    1,097
    In this case, you can increase the capacitor, so that ripple voltage will be 10%....5% of output voltage. But remember when you increase capacitor inrush current also increases
     
Loading...