Most of the circuits I've looked at for choke-input power supplies use two series caps to ground rather than one cap in each LC node. Is this generally a safety measure to ensure high enough max voltage specs in case the load is removed or critical current isn't met?
In addition, there are typically resistors paralleled across the caps. Are these bleeder resistors to draw extra current, ensuring critical current is met?
I'm looking at THIS circuit as an example.
In addition, there are typically resistors paralleled across the caps. Are these bleeder resistors to draw extra current, ensuring critical current is met?
I'm looking at THIS circuit as an example.
In general, yes: E-caps are technologically limited to ~450V, or a little bit more, and if you take into account the PT's regulation, mains voltage variation and no-load operation, this rating could easily be exceeded. Sometimes, it simply boils down to component availability.
In general, no: these ensure that the voltages across the caps are balanced, even if one has more leakage than the other.
However they can simultaneously play the role of a bleeder, and of discharge monitor if you put a LED in series with one of them (a valuable addition, as I can certify..)
The 220K are not a sufficient bleeder for a big power supply. The 6K0 to the right (in Fig 1) do that job.
As you are figuring, the rated loaded voltage is 518VDC, the no-load voltage can go to 820VDC, economic e-caps go to 450V or so, you have to have two 450V caps in series even at nominal voltage. Two caps never have the same leakage, and the voltage will not split equally; the 220K make the split more equal.
500V e-caps exist but won't last long at nominal 518V steady, not take 820V even briefly.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.