Wrong (lower) current values with SCT-013-030?

[dotto59]

I have an SCT-013-030 (30A, 1V peak signal, internal burden resistor 62 Ohm) connected to a UNO as "regular" schema (voltage divider with 2x10k from +5V to GND, and a 100nF condenser to GND).

My target is to monitor my Photovoltaic production, the same way I did before for years with an Efergy e2 power monitor (you can see it on the enclosed picture) that uses its own similar clip current sensor but without sensor specs I opted for that SCT-013. I'm aware that the measure can't be exact, as can vary depending on the load characteristics, but I can say Efergy always gave me not more than +/-5% difference from the inverter readings, so I expected the same range with SCT, Emonlib and Arduino UNO.

So, I made this simple sketch to perform some tests and calibration:
`#include "EmonLib.h"
EnergyMonitor emon1;
//Pin del sensore SCT su A1
#define SENS_CORRENTE A1
void setup()
{
Serial.begin(115200); //Apro la comunicazione seriale
Serial.print("Calibrating...");
// Just a bit of void readings to stabilize...
for (int i=1; i<=4; ++i)
analogRead(SENS_CORRENTE);
delay(500);
// Then calibrate (const= Ratio/Res. Burden 1800/62 = 29).
emon1.current(SENS_CORRENTE, 29);
Serial.println();
}

void loop()
{
// Read current
double Irms = emon1.calcIrms(1480);
// Just to ignore small fluctuations
if ( Irms <= 0.13 ) Irms = 0;
// Power ad current x 230V line voltage
int watt = (Irms*230);
watt = watt * 10;
// Show values
Serial.print(Irms,2);
Serial.print(" ");
Serial.println(watt);
delay(1000);
}`
The problem is I always get a power reading LOWER than real. For example, if I clip the sensor over the active wire of a power cable and connect a plug-in energy monitor and then a hairdryer, the energy monitor gives me the correct 1.7kW value, but that Arduino sketch gives me around 950W!
I then connected the sensor to the PV line out (the same wire I used with Efergy...) and here again I must multiply the resulting power by 1.7 to have a "similar" reading than the PV inverter says, but this 1.7 factor seems to be somehow "variable", depending on the PV output power, sometimes lower and sometimes higher for a bunch of hundred of watts, with an estimated factor ranging from 1.6 to 2.1 to get almost "exact" values!

I can accept a +/-5% precision like Efergy did, even a bit more, but an almost constant +70% is unacceptable. It seems something is wrong either with my circuit (but it's pretty simple..), and/or my implementation (but the measure sketch is like the example one) and/or the sensor (could be defective?), and/or the library (but I don't thik so), and/or what?
Thanks!

[bottosac]

I'm in your same situation. Need to replace Efergy with Arduino energy monitor for my PV.
I'm more near real values but with a burden resistor of 75ohm and a calibration value of 180. But in some ranges I'm away from real value of 200-300W

[beckmx]

@dotto59 how did you get to this value?
Then calibrate (const= Ratio/Res. Burden 1800/62 = 29).
1800 I suppose is the average value at 0 volts consumption?

[wangeris]

did you ever figure it out?