In article , Arfa Daily wrote:

"N Cook" wrote in message
...
Arfa Daily wrote in message
...

"N Cook" wrote in message
...
I can understand the other way round being a no-no but what is wrong
with
using switch mode power suply type electrolytics for mains smoothing of
high
voltage rails.?


The high voltage electrolytics in SMPSs are not specced for 40/50kHz.
That
would be the low voltage ones on the secondary, after the high voltage DC
has been chopped at 50kHz, and transformed down. The high voltage one(s)
are
merely the smoothers for the primary DC rail, derived from the normal
line
frequency (50/60Hz) input. There is absolutely no reason at all that
these
primary smoothers cannot be used for exactly the same purpose on normal
line
frequency transformer-derived supplies, such as might be found in valve
amplifiers, and indeed, I have used them for such purpose on many
occasions.

Arfa

I've never actually weighed or measured any but I always thought that V
for
V and uF for uF that the ones from SMPS were always lighter than the the
more general application ones. Therefore implying something different
about
their construction.


Whether or not they are made lighter ( and I suspect that in a lot of cases
it's that they are made smaller to better fit switchers and modern consumer
equipment physically ), if you think about it, they are just working at 100
/ 120Hz fundamentally, as their primary ( ha ! ) job is just to smooth the
main rail after the bridge, and the ripple from that bridge is just line
frequency x 2. Whether or not there is a transformer before the bridge, does
not have any basic effect on this. I can't immediately think of any other
differences that would cause the caps to *need* to be any different, but I
would agree that they do tend to be small compared to those found as
original in amps.

I'm not sure I go along with your ripple argument here.

The ripple in a SMPS input capacitor has two components. The AC current
into it at 100/120Hz, typically rather peaky as the input bridge
conducts for a fairly small angle when the mains exceeds the
capacitor voltage. The current being taken out of it goes into the
SMPS transformer, typically a sawtooth (with gaps) at the SMPS frequency.
As current in equals current out of the capacitor (when averaged),
the ripple thus has roughly equal components at both 100/120Hz
and the SMPS 40/50kHz. Thus this capacitor does need to be
rated for SMPS frequency ripple current.

However, I agree that a SMPS rated input capacitor should still be
fine for use generally at 100/120Hz (so long as its ripple current
rating is adequate for the application).

Regards,
Mike.