Charger startup sequence
We’ve had somewhat of a blank spot when it comes to the start of a charging session. Understanding it well can help diagnosing why the car might refuse. SpeakEV user ElectricBeagle bit the bullet and started measuring currents in all the leads to find out what is going on. For brevity, I assume that the cable is in and locked, connecting protective earth [PE] to the car’s chassis, and everything to activate the chargepoint is done. The main contactor in the chargepoint is about to close.
- The main contactor is closed in the chargepoint. This usually creates a significant current spike in the lead(s) as all L lines have a 100 uF capacitor to N, see the blue box in the third picture of this post. Correction: pretty sure now the capacitors are in a delta configuration, not a star.
- Car allows for dampening out ringing for a second or two. Note that because of the capacitors, a significant though reactive current (6A per lead) is drawn. This is, ignoring some losses, not a power uptake.
- The car injects two 20 mA pulses between PE and N in both polarities to measure earth resistance. If the resulting voltage exceeds 4 V, meaning an earth resistance of more than 200 ohms, the charge point is rejected. *)
- The car ramps up power uptake from 0 to the requested power in about one second. It’s very gradual, which is also notable in the ramp up of the “whine”.
Maybe you remember fairy tales about > 500 A startup currents by the ZOE, usually told by chargepoint mechanics. Well, this is what it is, a 100 uF capacitive load. Because of lead resistance and filter coils, inrush is actually limited and the spikes are 30 – 50 amps and last about one ms. This is no big deal whatsoever and any notion that this is problematic can be dismissed.
Now there are other tests done such as lead voltage and earth leak current, but these test can not be detected, though they should be pretty straight forward. I read of one driver seeing it’s car still charge on 180 V from an off grid PV system. I don’t know if the car would actually start a session this low.
I am now tempted to build myself an Arduino based earth tester using the same principle as have an extra diagnostic tool. For the Netherlands though, this is hardly worth the effort as most, if not all chargepoints are connected using decent earthing.
I’ll write in another post about the currents while charging.
*) Entire books have been written about safe earthing of power distribution networks. For the technically inclined Dutchies (or the ones who can bear Google Translate), this is such a book, link to the relevant part. I will only say here that ZOE of course is unable to have a true earth reference. So she assumes a network where N and PE are actually connected. In modern TN-C-S networks this should always be the case. In TT networks, where earth is provided with a local earth rod, things might not work out that way, let alone floating or midpoint-earthed networks. These are dying breeds but still all over Europe, especially in the more rural areas.