Re: Capacitor size? Me again

I think Andy's correct here.

Big electros are normally rated at -10/+180% of nominal value (this
does vary across manufacturers). If you put 2 caps in series where one
has a much higher value than the other, then model each cap as ESR in
series with capacitive reactance, you'll end up with a vast difference
in voltage across each cap when charging current is supplied. As each
cap cannot exceed its rated voltage, you'll end up only safely being
able to use about 150% of the individual voltage rating as the total
supply voltage, not the 200% previously mentioned.

The R strings only work where the R value is low compared with the
reactances, ie. where DC is being discussed. When 100/120Hz currents
are being discussed it is no longer viable, and the charging currents
we're discussing are quite large.

Putting caps in parallel to reduce effective ESR and increase
capacitance and ripple current rating is a standard technique. Putting
them in series increases ESR, decreases capacitance and leaves ripple
current rating unchanged. It's not often done.

Also, could I make a request on behalf of the more pedantic of us?
There is no such term as amperage; the correct term is current.

My $0.02, adjusted for inflation.

davidh

> awander@v... writes:
>
> And the only way you can double the voltage rating by

series-connecting the

> caps is if you can be sure they are EXACTLY IDENTICAL to each

other. This is

> very unlikely, and it means that the voltage will never be split

equally

> across them.
>
>
> Yes, and this is why one puts identical "voltage distributing

resistors"

> across EACH electrolytic cap. wired in series. Those R's then help

"even out"

> the voltage-split. Look it up!
>
>
> [Non-text portions of this message have been removed]