It depends on how the loop is constructed. Some multiphase converters have one feedback loop per phase, while others have a single feedback loop for all phases. Using a single loop for all phases means that the loop stability is taken starting at the multiple of the frequency of an individual phase. I think in the single feedback loop case, excess parasitics on one phase will affect stability of all other phases, and I do not know how the PWM controller accounts for this.
OK, so commenting before I have to go. its lower noise and higher current, because you have the resistance and voltage (and thermal) noise cut in half, plus you can use smaller, higher quality inductor parts as its higher FS and easier to filter (although more challenging to lay out). does it make sense to use just because you want lower noise with a higher voltage drop than is practical for linear, even if you dont really need a higher current?
unavoidable power solution for the PC motherboards, you can see the implementation of this type of buck converter on any motherboard. Thanks Zach ! you are the king Actually, it would be perfect to see the layout tips(both PDN and EMC) for this because handling high currents is usually tricky.
I have not seen any isolated multiphase converter available on the market. This is currently a research topic. If you look on Google you will see many scholarly articles looking at the design of isolated multiphase converters. I think the reason this has not been achieved yet is due to the difficulty of isolating all the phases with the same transformer. This is because you have a lot of switching phases in parallel. You could isolate all phases separately but this makes the system very large, so it would be desirable to do this through a single transformer in a half-bridge configuration or resonant LLC configuration.
