My general advice is to never connect anything but resistors in series.
If you do need 12 individual outputs, spanning from 0 to 600 V DC, there are some benefits of stacking the outputs of 12 converters in series:
Downsides would include:
- DC stress on the topmost transformers
- Dynamic behavior at startup and faults
Your other option would be to make just one converter with 12 outputs, at which point you have more control over the DC stress in the transformer. 600 V isn't that high, but your average enamel wire may only be rated for 300 V working voltage. With tape or insulated wire, you could have the last windings have ample voltage rating. Downside would be cross regulation with just one feedback. Without for example post regulation in the form of an LDO, you're down to how well the secondaries in your transformer are coupled to each other and how much load imbalance you have.
Assuming the transformer isn't a limitation, you need to consider fault cases. The foremost I can think of is either if one of the converters would stop operating by some fault or just an overload from whatever is connected downstream of it. If there are any loads from for example +600 V to ground (across all converters), depending on how much it behaves like an ideal current sink and how much inductance you have, will try to force the converter not supplying any current to go negative. The situation is identical to a series battery pack where one cell is depleted before the others. If the load across them keeps pulling current, the depleted cell will eventually go negative voltage. BMSes are used to prevent this very situation.
I would recommend you to put a Schottky diode across each output to prevent it from going too negative to damage anything (capacitors, feedback) and ideally, a power good signal from each converter, probably via optocoupler down to some central MCU or similar to shut down the other converters if this happens. A BMS of sorts.
The schematic below is very simplified. As it stands, it won't allow the converters to boot. You need to inhibit the power good function for some time until they are expected to be up and running. It's just to give you an idea on how to catch problems. D4 and D14 are Schottky diodes. It may be possible to hook into the existing feedback loops and determine if a converter has lost regulation, but it's the same startup behavior you need to catch as they all start from zero yet there is nothing wrong.
simulate this circuit – Schematic created using CircuitLab
