This site is now in read-only archive mode. Please move all discussion, and create a new account at the new Victron Community site.
Dear All / Dear Victron Team,
I´m using a the EVCS NS in ver.1.26 in combination with a ESS System v3.10-27.
I know that the EVCS is not a pure measurement deveice but i have do modify the calibration factor from 0,8422 / 5A ; 0,895 / 6A ... up to 1,023 / 16A.
The measurement was done with a shelly 3EM meter.
Is this normal?
Have anyone else this issue?
Regards Christian
Hi, for this hardware version, we are only measuring the current, not the voltage. So the power measurement is an approximation. Especially if the voltage is dropping as the vehicle consumes more.
@FChristian , @Lucian Popescu (Victron Energy Staff),
I have the same experiences. Maybe there should be some polynomial function for power calibration in the EVCS setting instead of simple coefficient...?
For power calculation, nominal 230 V value is used? Maybe EVCS should use phase voltages measured by the Multis and getting from the GX device. Even if the voltage drop on cable will be 10 V (already very high value), error in power calculation will be 100 W@10 A charging current, but now it should be more than 10 times higher (> 1 kW)...
@Lucian Popescu (Victron Energy Staff) , same experiences with previous Renault Zoe ZE50 R135 and with actual VW ID.3 1st Max Pro during 1p and also 3p charging. Zoe was unable to charging with current lower than 10 A with Victron EVCS, ID.3 is able to charging from 6 A without problems.
Problem is, that in EVCS setting there is only simple coefficient for power calibration. At lower charging currents error in power calculation is higher than at higher charging currents (in comparison with indicated power by car, measured power by Multis or measured power by grid operator's energy meter). For example at 10 A charging current calibration coef. must be set to approx. 0.91, while at 16 A to approx. 1.05 for proper power calculation. With one simple coefficient error in power calculation can be higher than 1 kW, so when you are using Auto mode with charging current in range 6 to 20 A, then charging statistics are not correct and are unusable...
If the reports are important, you can leave the internal coefficient at 0 and apply an external, for example using node-red.
@Lucian Popescu (Victron Energy Staff) The problem is not caused by voltage drop on the supply cable. I use 10 meters of CYKY-J 5x4 mm2 cable, so during charging voltage drop is less than 2 V. Even if voltage drop will be 2 V (between Multis and EVCS) and charging current will be measured 100% correctly, it will cause power error of 2 V x 20 A = 40 W. Problem is, that for different charging current must be used different calibration coefficient, which is not possible in actual FW.
If I remember correctly you wrote, that to VRM will be added more statistic data regarding to EVCS. But if power measuring is wrong, it will be completely unusable. I can't figure out, why you don't make calibration of current measuring on each EVCS during manufacturing stage, because this must be absolutely done with your great inverters/chargers... You only added simple calibration coefficient (with a pretty large range) to FW and essentially transferred the responsibility for the correct power measuring to us (installers), but customers want correct data, if they paid not exactly a little money for "clever switch" box on the wall.
I was not our intention to have this charging station used as a measuring device. Same as our inverters, If you want precise measurements, an extra meter is needed.
But I can give you an inside info, on the next version, a metering IC was added, so all the parameters will be measured. So with that one, all the reports should be very precise.
For me also 50w differents is not a problem, but when charging with 6A and with 0.95 coeficent it shows 800W more that measured with amper meter and voltmeter then it not ok. And with 14A and with 0.95 coeficent it is almoust correct.
Hi,
Thanks for the info on the new version but do you have a release date?
I'm planning to install an EVCS soon!
Hard to say, it depends, probably this year
I have no idea if this thread is still active, but I'll give it a try.
I'm as well annoyed by the 'creative' power meassuring of the EVCS. I have mounted a 3-phase meter in the distribution panel at the garage where the EVCS is connected so I have valid data to compare with.
While VRM and re-imbursing charging fees are valid reasons to expect better consumption data, my issue is that I just modified the automatic 1P/3P changeover (link see here), which basically works very well.
However it is triggered by the same problematic EVCS actual power reading and by using a fixed calibration co-efficient either I can not reach enough power @ 1P/20A to trigger the switchover from 1P to 3P or if I increase the CP the power will not drop low enough @3P/6A to switch back to 1P.
So simliar to @FChristian in the opening post I established the required co-efficient for 6A,8A,10A, etc and wrote a little NodeRed flow that sends the correct co-efficient to the EVCS via ModBus each time the charging current is changed (either by manual or auto mode).
I tested it yesterday for the first time and it worked pretty well during a first test. Unfortunately today it's quite cloudy, so no extended testing possible, especially 1P/3P.
So if anybody really wants to have this feature now, maybe give this workaround a try...
Best Regards, Jochen
2 things are coming
- a variable calibration will be added, here is a preview:
- the new hardware will have a dedicated meter inside.
But, most of these issues are coming because the power factor of the vehicle charger is not 1 and/or is not constant.
Did the same but as I do not have deticated meter to see actual power then I just used the graph Christian provided in first post. Seems correct.
Used this in fuction node.
var current = msg.payload
var output = 85
switch (current)
{
case 6:
output = 89;
break;
case 7:
output = 92;
break;
case 8:
output = 93;
break;
case 9:
output = 94;
break;
case 10:
output = 95;
break;
case 11:
output = 96;
break;
case 12:
output = 97;
break;
case 13:
output = 98;
break;
case 14:
output = 99;
break;
case 15:
output = 100;
break;
case 16:
output = 102;
break;
default:
output = 85
}
msg.payload = output;
return msg;
