I'm working and learning about power supplies.
How does one size the voltage of an electrolytic cap after
the diode bridge? At say 24vAC from transformer,
then after diode bridge I read about 21v, but when
the cap is installed, it voltage reads closer to 35v.

Do I need a smoothing capacitor rated for 25v, or 35v?
Maybe a little extra on top?

 

Rectifying an AC voltage gets you a DC voltage that is 1.414 times the AC voltage..
and
http://forum.allaboutcircuits.com/threads/full-wave-rectifier-smoothing-capacitor-value.67094/

and you want a capacitor with a voltage HIGHER than the max voltage..

 

Thanks! 50v cap should do it.

 

Rules of thumb for high reliability design in military and industrial applications:
Capacitor voltage rating is at least twice the peak voltage applied.
Transistor voltage rating is at least twice the peak voltage applied.
Resistor power rating is at least twice the average or RMS power applied.
Diode current rating is at least twice the peak current applied.

Etc.

ak

 

I recently repaired an old radio shack 12v 2A supply for a friend. It had single 2200uf
16v smoothing cap. I thought that was low, but I wasn't sure. I had a 35v 4700 in my
junk box, so I replaced it with that. That got me to wondering whether the ac
voltage or the peak voltage was the one in consideration. I guess the lesson is
this: I shouldn't let radio shack engineering influence proper application.

 

radio shack 12v 2A supply. It had 16v smoothing cap.

A 16 volt capacitor will survive just fine with 12 volts DC applied to it. Were you thinking the 12 volt DC label meant Radio Shack started with a 12 volt AC transformer?

 

Yes, that was my assumption.
Come to think of it, that unregulated radio shack
power supply sure couldn't keep
the voltage up under even a modest load.

9v transformer * 1.414 = 12.726v...

Oh, the irony.

 

Yes, that was my assumption.
Come to think of it, that unregulated radio shack
power supply sure couldn't keep
the voltage up under even a modest load.

9v transformer * 1.414 = 12.726v...

Oh, the irony.

Note that the secondary voltage of an unloaded transformer will be about 25 percent higher than the voltage at the rated load. So your 9-volt transformer could put out as much as 16 volts into an open circuit.

 

Note that the secondary voltage of an unloaded transformer will be about 25 percent higher than the voltage at the rated load.

True.
Maybe not exact, but true.
Two of the difficulties in designing linear voltage supplies are the allowable variation of the power line voltage and the transformers' voltage output change from unloaded to loaded conditions. Over-designing for these variables is a big factor in how much energy is wasted in a linear regulator.