Power Supply Cap Values


2005-07-01 10:52 am
hi all,

I'm a newbie which needs some guidance on capacitance value selection.

I've a CDP that has a non-regulated power supply controlling the transport, servo system etc and it is has a pair of 4700uF 16V caps after the rectifier (for smoothing purpose I supposed?).

Would it be beneficial then to change the caps to something like 10,000uF? Having a larger cap means that I would have a larger energy reservoir but what would be the disadvantage then?

Hope you guys can help me out.

Hi there...here's an example.



  • ripple cals.jpg
    ripple cals.jpg
    13.5 KB · Views: 321
Dr.Gone said:
My guess is that the inrush current will be more, so you may not want to go with too much capacitance...

Yes. Remember that there is a relatively short time for the rectifier diodes to charge that cap (only when the AC voltage from the xformer exceeds the cap voltage), so the charge goes by pulses. If you go to 10.000 uF, those pulses increase significantly in peak current value. I would think going from 4700 to 10.000 is probably alright, but it increases the stress on the rectifier.

Even if the ripple is lower, there is still the question of what it does to the final sound. You probably have to try it and see (hear) if you like it.

Jan Didden
mzzj said:
power factor is somewhat better.

Ignore that one for the moment.....with cap input filtering and high step down (220VAC to 12VAC) voltages with lowish current no real improvement will result. The winding resistances is high enough to soften peaks.
In practice the transformer will run slightly warmer, but others mentioned diodes etc are more important as more peak current is being handled.
You will have to decide the trade off.
Rule of thumb....1V p-p ripple is roughly design standard on full load.

Recent power factor legislation in CA requires everyone to re-think the whole practice of peak current/ cap input filtering (low pf factor) in high current power supply designs with better efficiency.



Rule of thumb