# Design (Block diagram/topology) of high-current (I>100 A) power supply

I recently watched this video about the design of a high voltage (300 V) commercial power supply. I reckon that the block diagram is essentially a bridge rectifier followed by a DC/DC convertor like the following image shows.

Given the high voltage involved, it is implemented with a Phase-Shifted Resonant Convertor that uses soft-switching to reduce switching loss.

I was wondering, how are high current commercial power supplies designed, let’s say DC current above 100 A such as these and these one.

In theory, a bridge rectifier followed by a buck convertor would do the job, but I imagine it would be terribly inefficient (the same reason that the high-voltage supply isn’t implemented using a boost convertor).

Does someone know topologies are used in commercially available high current power supplies?

Many thanks in advance

• No, a bridge + buck would not do the job because most of these high-current power supplies require input-to-output isolation. A buck does not offer you any galvanic isolation from the mains. Even high-voltage DC lab supplies are generally mains-isolated, so they will all use passive rectification (possibly with PFC) and isolated DC/DC conversion. Commented Jun 10, 2020 at 14:17

## 2 Answers

You can use a multiphase buck converter such as this one: -

• Input is 30 volts to 70 volts
• Output is 12 volts at 180 amps (2.16 kW)

Note that this is a 6 phase device and the full circuit is shown in the data sheet on page 46.

To get down to the required input voltage range for the LTC7871 use an isolating forward converter to produce circa 50 or 60 volts (45 amps) from the power factor corrected rectified input AC voltage.

• @KenGrimes - are you done with this question now or do you need further clarification? Commented Jun 13, 2020 at 12:11

In my experience, good-quality lab supplies will use isolated DC/DC bridge conversion topologies, e.g. phase-shifted/resonant transition full bridge. These topologies allow for brick-wall current limiting, a wide voltage adjustment range, and are very resilient - the powertrain is quite robust.

Pure resonant topologies (like LLC) are less useful for lab supplies as you have a limited output adjustment range (you cannot regulate down to zero) and brick-wall current limit is very challenging unless you start adding post-regulation to the DC/DC converter, which adds cost and complexity.

As I commented earlier, you cannot just buck a rail from the rectified mains because you need galvanic isolation on any user-touchable rail in order to not get in trouble from the regulatory people (and also not to kill them by electrocution) - every lab supply I've seen has a mains rectifier and isolated DC/DC converter stages.