In article , Ken Scharf
writes:

Many newer eletrolytic caps have both a 'surge' and a 'working' voltage
rating. I don't know if this is just a new parameter, or if the newer
caps were engineered to handle a brief surge of voltage while
'charging'. In anycase using 150v caps in a circuit supplied by 120v is
NOT allowing for the peak voltage. I think in a half wave circuit
there is probably a lower peak since it is reached in double the time
between peaks. Even so, 180v or even 200v caps should be used. I've
seen filter caps now being made in a 160v rating for line service, but
that's still too low for safety (but maybe these units carry a higher
'surge' voltage rating).

Peak voltage reached is the same whether a rectifier is half-wave
or three-phase full wave. It is dependent on the AC input voltage.

Surge currents are serious concerns in semiconductor diode
rectifier circuits. Not just within the circuit but from the AC line
itself. To get really worried, dual-trace scope the AC input line
with second trace on the diode DC output, sync on the AC line
set to trigger a sweep above the normal AC peak voltage. Watch
during a time when other appliances in the house are working,
like food mixers and bread machines. Might be surprising...

For details of the guts of capacitors, the Cornell-Dubilier website
used to be good. Surge ratings depend on the foil, the electroyte,
size of the lead wires. Lots more different things for tantalums.
There's not much one can do if someone else makes them, just
read the specs and hope for the best.

For hobby work of onsies and twosies, if in doubt use the higher
working voltage ratings. That way there's less time wasted on
analyzing all the busted parts after being too economical.
Hobbyists don't have all the time in the world to shave designs
down to the nubbin to save pennies per unit on a thousand-lot
production run.

Len Anderson
retired (from regular hours) electronic engineer person