Sorry in advance for a question with a possible obvious answer:
2 electrolytical capacitors in series, each rated for max 25 V, is it safe to apply 35 V at the 'outside' of these two capacitors?
If it is, is the maximum allowed voltage 50V?
I think it is but I am not sure. Thanks.
2 electrolytical capacitors in series, each rated for max 25 V, is it safe to apply 35 V at the 'outside' of these two capacitors?
If it is, is the maximum allowed voltage 50V?
I think it is but I am not sure. Thanks.
You should have voltage-sharing resistor string alongside the series cap string, otherwise the voltage won't share evenly and could overstress one of the caps. Electrolytics are particularly problematic for putting in series as the value tolerance is so wide, and they change value over time.
I happened to have these, their size is small and it is interesting to know, I was curious.
These capacitors are part of a Sijosae Discreet Rail Splitter and I am considering to use a higher input voltage.
If you do series 'lytics, be sure to place resistors of equal value across each cap to swamp out variations in DC leakage current in the different caps if there is a DC component across the caps. Otherwise, the cap with lowest leakage current will slowly go into an overvoltage condition after the caps have charged up.
I always use parallel resistors to equalize out the voltages despite leakage current differences, but someone on a Dutch forum who usually knows what he writes about claims that it equalizes out automatically. Apparently the electrolytic that is going into overvoltage automatically starts leaking more, which solves the issue. I haven't dared to try and see if he is right yet.
I've seen commercial caps in series inside a cardboard package with no dividing resistors. I think it was 4 uF @ 700 VDC, made up of 2X 8 uF @ 350 VDC. We ran 'em to the max and they rarely failed. The cap voltages do equalize, more or less, because at higher voltage the leakage current increases, just as described above. I'd only do it with identical caps and it makes me nervous because it seems things could go south some years down the road when one or the other increases in leakage by a lot.
It seems I was right to ask the question in the first place, things are clearly not as simple as I hoped for.
Better not take a risk and stay well below the maximum voltage.
Better not take a risk and stay well below the maximum voltage.
Music Reference RM-10 is one of the example NOT using equalizing resistors on series-connected electrolytics.
Music Reference: Sound Design Philosophy
Music Reference: Sound Design Philosophy
It is possible, in general. But better to avoid that because of reasons above. We use this only for a high voltages when there isn't caps with large enough rated voltage.
At turn on, there can be a substantial reformation current flow which subsides to the normally identified leakage level. That reformation current can easily be more than 10x the leakage level. An electrolytic cap also is designed to allow +10% overvoltage for a short duration, which can help during turn-on when even with balancing resistors the voltage sharing may not be to the resistor tolerances due to capacitor differences (especially if the e-caps are not the same make/vintage or someone has replaced just one of the ecaps during service).
Which is probably why once starting to service an old amplifier with equally old capacitors, it is wise to replace them all at the same time? And probably not only the capacitors but also other components like carbon resistors that vary over time. And stay within the safety margins of the components whenever possible.
If I interpret correctly.
If I interpret correctly.
Last edited: