EmonTX v2.2 / phase swinging from positive to negative consumption

Submitted by mehale on Mon, 10/11/2014 - 20:05

Hi folks,

I have this odd behaviour. Phase #2 is swinging from positve to negative watts.

I have an emonTX v2.2 loaded with emonTx_CT123_3Phase_Voltage. Three 127v phases, the 9vac reference is plugged to phase1 + neutral.

My settings are:

double Vcal = 119.180;//228.268;
double Ical1 = 90.9; //˜0.67A                           // Calibration constant for current transformer 1
double Ical2 = 90.9; //˜0.64A                           // Calibration constant for current transformer 2
double Ical3 = 90.9; //˜0.60A                           // Calibration constant for current transformer 3
double Phasecal1 = 1.00;                       // Calibration constant for phase shift L1
double Phasecal2 = 1.35;                       // Calibration constant for phase shift L2
double Phasecal3 = 1.37;                       // Calibration constant for phase shift L3

Please find attached a screenshot of my dashboard and the .ino file.

Re: EmonTX v2.2 / phase swinging from positive to negative consumption

Submitted by Robert Wall on Mon, 10/11/2014 - 20:46.

Would I be correct to think that the graphs are Top: L1, Centre: L2, Bottom: L3?

I can see a 1100 W load on L1 & L2, followed by a low power 'generation' on L2 of a few watts and a switched 'generation' of about 100 W.

What do you think is the actual load on L2? If there is a highly reactive load and the phase calibration is not exact, then because of the approximations of this sketch, it may well be that a low power factor load could appear as generation, which is what the negative power means.
You also need to remember that the emonTx V2 does not give reliable results below about 250 W (1% of its maximum reading) even on a single phase supply, due to the way the ADC works and noise pick-up.

Re: EmonTX v2.2 / phase swinging from positive to negative consumption

Submitted by mehale on Mon, 10/11/2014 - 21:24.

Hi Robert,

Yes, the graphs, top to bottom are L1, L2, L3. However, I just realized that I failed to check if the code represents the markings on the PCB. Probably yes, and I havent changed the code regarding to this.

I haven't measured the actual load at L2 with a clamp meter. I will do it tomorrow and compare the results.

Do you suggest a way to calibrate the phases?

Thanks,
Felipe

Re: EmonTX v2.2 / phase swinging from positive to negative consumption

Submitted by Robert Wall on Mon, 10/11/2014 - 22:27.

All the calibration instructions are in the comments at the beginning of the sketch.

Normally to set the value of Phasecal, we say use a pure resistive load and set for a power factor of 1.0. That coincides with maximum real power. The problem with that is it is hard to find the exact maximum as the peak is very broad. If you have a large capacitor that is suitably rated for both voltage and current, there is a more precise way to set the value of Phasecal. If you have a pure reactance (a capacitor), then you adjust for zero real power, and it is easier to find a zero than it is to find a maximum.
There is one big problem with this method, which is why it's not mentioned: few people have a suitable large capacitor available. I use 3 μF and that is a lot smaller than I would like, so to multiply the current I put the wire through the c.t. many times to give me a 10-turn primary winding (equivalent to 30 μF and showing a current of about 2.2 A @ 240 V, giving me a load of about 550 var).

When I asked about the load on L2, I was hoping to be able to estimate the power factor. I want to know what sort of load is it? Which appliances?

Archived Forum