(Beginner) Three-phase inlet, two-phase loads

A little background first. Question at the bottom.

I just set up the emonPi with two CT’s, a pulse counter, and an AC-AC adapter. I might expand soon :slight_smile:

My goal is to monitor the total use for my home, and a few select circuits to see how much I’m using where (ie. Hotwater tank, charging of electric vehicle, heating.) .

As I now understand, the standard firmware is made for 2-phase systems. I have a three-phase inlet, and about 20 circuits spread evenly between the phases.
When I initially installed, I randomly connected the voltage reading to phases 1-2, one CT on 2-3- circuit, and another CT on 3-1-circuit. Then my power readings where totally off.
I moved my voltage reading to 3-1, then the second CT seemed to be correct.

Is it possible to have the power calculated as just Vrms * Irms, without considering the phase?
That way I would get a power reading regardless the phases of the circuit.
The voltage over the different phases seem to be pretty balanced, within 0-1V deviation. My loads are pure resistive.

How would I change the firmware and upload on the emonPi? The guides seem to be only for the EmonTx.


Sorry, no. All OEM equipment is designed for the UK single phase domestic supply system.

This suggests to me that you have a 3-wire supply, (3 phase conductors, no neutral conductor, and not counting the protective earth conductor) and you measure 230 V between any two phases, and phase-earth voltage is therefore 132 V.

Yes, the numbers are calculated but not used. But that (Vrms * Irms) is apparent power, not real power, which is the quantity you pay for.

The instructions I believe are in the emonPi section of Resources. I have a sketch for the emonTx and a 3-phase, 3-wire supply (not on GitHub though). In essence, it delays the voltage wave by 60°, which is the phase angle between two phases using the third phase as a reference, i.e. it becomes a 2-phase system as there are only 3 wires, and uses that to calculate the real power on the second input. It will give you the total real power across your 3 phases - only the total has a meaning, the “power” per phase is meaningless.

If you confirm your supply, I’ll search out that sketch and see what needs to be done to convert it for the emonPi.

Thanks for your reply!

You’re correct, I have a 3-wire supply, 230V between any phases. (Sorry for the vague description and confusion about two-phase vs single phase)

I will measure the total consumption through the supply with the pulse counter mounted on the domestic meter.
This is sufficient for me.

With the CT’s I’m only interested in measuring power on individual single phase loads.
Also, because the loads are purely resistive, I figured the real = apparent power, or close enough…

Can I extract the apparent power to a feed?


Checking the emonPi sketch, where there is now:


You need to replace 
```  ct2.realPower```
``` ct2.Irms * ct2.Vrms```
assuming that CT1 is the current associated with the voltage measurement, and CT2 is the one that is wrong.

That will mean you don't have to change emonHub: Power1 will be the real power, Power2 will be the apparent power (but provided your loads remain resistive, the same), and Power1+Power2 will be your total power.

If you want the incorrect power measurement, or if you want the currents independently, then you must add some variables to 
```typedef struct { 
int power1;
int power2;
int power1_plus_2;                                                    
int Vrms; 
int temp[MaxOnewire]; 
unsigned long pulseCount;  
} PayloadTX;     ```

assign values to them, and extend emonHub to accept the new extended message.