I'm currently working on a project which requires building an EV battery-charger as a starting point. At present, I wish to design it for a 220 V, 50 Hz grid source input. The nominal voltage of the battery pack is 82 V, and nominal charging current is 80 A.

The DC-DC converter has to be in current-controlled mode (for 80 A constant current). The top-level schematic is as follows:


However, many guidelines on the web suggest that buck converter is not a suitable choice for high power applications:

  1. https://www.we-online.com/web/en/passive_components_custom_magnetics/blog_pbcm/blog_detail_electronics_in_action_45887.php
  2. https://www.smps.us/topologies.html

Why won't a buck or isolated-buck be sufficient?

  • 2
    \$\begingroup\$ woah, a charging current at say 82 V of 80 A – that's technically very challenging. Are you sure this is your "starting point"? \$\endgroup\$ – Marcus Müller Aug 12 '19 at 11:15
  • 2
    \$\begingroup\$ So you are designing a 6.5 kW converter and need help on the best topology? Are you sure this is something you are capable of doing on your own? \$\endgroup\$ – Vladimir Cravero Aug 12 '19 at 11:28
  • \$\begingroup\$ @VladimirCravero We're a team of few people here ;) \$\endgroup\$ – Vivek Gangwar Aug 13 '19 at 5:56
  • \$\begingroup\$ I'd definitely consider buying something for this. I would probably be looking at several phase shifted bridges effectively in parallel to reduce component stresses but this is not a beginner project/ I have been designing SMPS 30 years plus and I would not consider this an easy project. \$\endgroup\$ – Warren Hill Aug 16 '19 at 8:02
  • \$\begingroup\$ \$ 82V \times 80A = 6560W \$ assuming 100% efficiency is nearly 30A from the wall. You won't get that out of a standard mains socket so you will need to get an electrician in first. You can get this power from single phase as for example some electric ovens draw similar current but are usually hard wired and have their own breaker.. \$\endgroup\$ – Warren Hill Aug 16 '19 at 9:22

I think you're underestimating your problems.

I'll list a few things I think you should've considered in your drawing.

  • Find a cable rated for 80 A in a vibrating environment, matching connectors.
  • find a solution for connecting your components. Calculate how wide your tracks would need to be on a PCB, allowing 50°C of temperature increase on a 70 µm copper PCB.
  • So, considering the input power needs to be at least the output power, do a calculation of the current you draw from the 220 V grid. Then calculate how much power the conducting diodes in your bridge rectifier dissipate.
    You want these to be synchronously switched FETs, most likely.
  • Calculate the necessary size of the input capacitor if you don't want your voltage to fall below say 150V across that under maximum current draw. Check the price on that, and how much it'll get warmer due to heating through ESR.
  • Dimension your system: Which switching frequency do you aim for?
    • The higher the frequency, the smaller your inductance needs to be (calculate the size, copper and core weight of the inductor you need).
    • But the higher the frequency, the more power you lose when switching the FET. Calculate how much waste heat you need to dissipate!
    • Try to find a price for that FET. Note down its gate capacitance. Evaluate alternative choices like IGBTs.
    • How will you drive that FET? Build a gate driver.
  • Calculate the losses in the wheel diode ("FWD") in your circuit based on the above switching frequency
  • After you've done that, you have done the basic design steps for a voltage converter. That's not a Lithium battery charger! You need a lot more logic and sensors to make it one. And a lot more failsafes to make it a non-explosive battery charger.
  • Take your plans, and show it to an EMI testing engineer. Bring Vodka.

The losses in semiconductors are what drives people to use other, resonant technologies.

  • 2
    \$\begingroup\$ Bring vodka !!... \$\endgroup\$ – Solar Mike Aug 12 '19 at 11:36
  • 1
    \$\begingroup\$ This. Losses are the real enemy here. Ignoring safety issues that come with non-isolated buck converters, you will simply have no efficiency. Interleaved bucks work for higher powers than regular bucks, but for 6.5kW you're going to need a PSFB (or LLC if you're feeling brave). I'll leave you with this: If that switch fails short, what happens to your output? \$\endgroup\$ – Stiddily Aug 12 '19 at 11:49
  • \$\begingroup\$ I think the technical enemy is not the physical one (losses) here: OP vastly underestimates the severity of the physical problems in all parts that seem easy in a low-power low-voltage buck, and the complexity of the technical solutions that these require, @Stiddily. So, the enemy is unexpected complexity paired with high financial, technological and human risk. \$\endgroup\$ – Marcus Müller Aug 12 '19 at 11:55
  • \$\begingroup\$ And don't forget a suitable PFC ahead, no country will let connect to the grid without. \$\endgroup\$ – carloc Aug 12 '19 at 12:05
  • \$\begingroup\$ @carloc I did mention to bring vodka, didn't I? (it's not an EMI problem, but if anyone, then EMI engineers know their grid filters) \$\endgroup\$ – Marcus Müller Aug 12 '19 at 12:05

Not the answer you're looking for? Browse other questions tagged or ask your own question.