Re: capacitor in series
The resulting value (uF) of the network works as predicted.
Theoretically, you should get double the WV. Unfortunately, it doesn't work that way in the real world.
It is a problem of balance. If there are any differences between the characteristics of the two caps (always) the center point will shift accordingly. This is an age-old problem with caps and is much more pronounced with aluminum electrolytic caps.
Even the best caps are pretty bad compared to the theory.
In Aluminum Electrolytics (AEs), the electrolyte deteriorates without applied voltage and is refreshed when forward voltage is applied. Tantalum caps also heal to some extent as well. If this is not exactly synchronized in two caps in series, one cap will see the full network voltage. If this network voltage is above the working voltage of one of the caps, it's cleanup time. (If you haven't cleaned up after an aluminum cap breakdown failure, you don't want to. I have also been burned by pieces of exploding Tantalum caps 5 feet way! Tantalums typically burn holes through the circuit board. )
When designing something for a one time use in the lab, you can get away with about 70% of full WV + WV, but if you want to reproduce the circuit (manufacture it!) don't even bother. Your quality review won't give you any more than one cap's WV. but, your line test failures will be much more entertaining (not to safety people of course.)
There are legitimate reasons to put caps in series. Obviously, getting a non-standard value. Also there is a slight improvement in reliability if the networks applied voltage is below each cap's WV (sharing), but that is offset by the decrease due to simply having one more part to fail.
Ceramic caps exhibit a piezo effect increasing with applied voltage which means their value changes with applied DC voltage. Using two ceramics in series to halve the applied DC (V<<WV) to each will improve the stability of value (uF); but at the cost of board space, MTBF, and price of an additional part.