1

I am looking at an EV charger, a very simple 3-phase AC unit. Basically, it's a power outlet with a bunch of relays and a microcontroller.

As I understand it, the input of an EV Onboard Charger (OBC) will normally have a significant input capacitance, which means a significant inrush current will occur when the relay turns on. That might mean a significant risk of contacts being fused.

How do they prevent this from happening? Do they switch each phase separately at minimum voltage? Or are there other tricks?

TQQQ
  • 831
  • 4
  • 16
  • What EV charger? You think "a power outlet with a bunch of relays and a microcontroller" will make the onboard EV charger work? – Rodo Aug 07 '23 at 16:54
  • @Rodo no, that would be an AC wallbox. Like this https://m.alibaba.com/product/1600791478243/New-Energy-Charging-Pile-32A-7kw.html?__sceneInfo=%7B%22cacheTime%22%3A%221800000%22%2C%22type%22%3A%22appDetailShare%22%7D&from=share&ckvia=share_0ebf1f976cdd4cd6bc120dd8a03fd6e5 – TQQQ Aug 07 '23 at 17:21
  • Well, one trick is to not use relays and to use contactors whose contacts have features such as the appropriate plating and magnetic blowouts. But I agree with Rodo in that something is missing in your description about "bunch of relays and microcontroller" because that won't let you charge a battery which requires DC from an AC input. – DKNguyen Aug 07 '23 at 17:23
  • Are you asking about the particular unit you have in front of you (in which case you should post some pics & schematics), or are you asking about a more general case? – brhans Aug 07 '23 at 17:25
  • @brhans I am interested in general case. Any good practice is interesting. – TQQQ Aug 07 '23 at 17:30
  • @DKNguyen in the said situation there is an onboard charger on the vehicle, "my job" so to say is only to turn on the power after negotiating over the PP pin. – TQQQ Aug 07 '23 at 17:44
  • 1
    Consider zero-crossing TRIACs. Then there are no contacts to fuse (but do be mindful of the max current through TRIACs - too much inrush can fuse them closed just like relay contacts.) – rdtsc Aug 07 '23 at 18:23
  • @TQQQ In that case the onboard charger could have soft-starting features. – DKNguyen Aug 07 '23 at 19:48
  • @DKNguyen I guess so, but still could have input capacitors. One thing - maybe the 5m wire inductance is a good measure for inrush current control. But I am looking for other ideas too. – TQQQ Aug 07 '23 at 19:53
  • @TQQQ You don't need to have input capacitors *directly* on the input connector with nothing in between. – DKNguyen Aug 07 '23 at 20:12
  • What do they usually have on the OBC? Is there another contactor? I would expect a choke and a rectifier, of course, but still a potential of a few hundreds of amps on turn on. – TQQQ Aug 07 '23 at 21:08
  • 1
    @TQQQ Have you considered hybrids? These use semiconductors first (like a triac or double-scr) to make the connection before and for a short time while the relay is engaging. The relay then takes over, but the voltage difference is by then very small and welding far less likely. – periblepsis Aug 07 '23 at 23:20
  • 1
    @periblepsis honestly never heard of it. I'll go googling, sounds like a good idea. – TQQQ Aug 08 '23 at 01:09
  • 1
    @TQQQ Because the semiconductors are being used for such a short time, they don't absorb much energy in the short time before the relay is fully engaged. Which means (mostly) no heat sinking needed. Or, at least, not much. – periblepsis Aug 08 '23 at 05:35

1 Answers1

1

I don't have experience of EV chargers, but have dealt with switching large power levels to capacitive loads before.

Several approaches are commonly used:

  • NTC varistor to provide in-rush limiting. Predictability of load, ongoing power dissipation, and recovery time for cyclic switching may be an issue.
  • As above, but with a fixed resistor (for current limit) and a second relay to short after initial period. Both relays have to be rated for near full current.
  • As above, with resistor just sized for slow "pre-charge". Only main relay needs full current rating, but remote end needs to be compatible with pre-charge approach.
  • Inclusion of a charging inductor to limit initial inrush. Potential for spike on "off", but can be used with snubber, and potentially inductor may be part of PSU circuit (if DC)
  • With AC switching (as the question indicates), then a zero-crossing circuit will help with capacitive load (but not with inductive), by connecting at the point of minimum voltage.

Which approach is best/used will depend on a range of factors, including: cost, space, characteristics of supply & load.

Having said all of that, I would expect that the EV vehicle may well have its own active inrush limiter, so this may be a non-issue. I'd look at the detailed charging specs for confirmation.

colintd
  • 2,018
  • 2
  • 17