The voltage between the two mains conductors is nominally 240V. That is broken down into two 120V potentials between each of the Mains and the neutral, but unless they are in perfect balance, there will always be a difference in voltage between the two phases. That said, the difference in most situations is practically negligible and the sum of the two will always equal the voltage between the two Mains.
This is an important point, because in order to measure real-power, what your meter measures and what you pay for, it’s necessary to also have a reference voltage on that circuit.
IoTaWatt supports multiple voltage references, but as a practical matter, you only need one to measure any of the three circuits described above.
The simplest and most effective reference is a 120V wall transformer connected to an ordinary 120V plug as close to the load center as possible. That will provide a reference for the phase it is plugged into, and a reference that is the exact opposite of the other phase, or to put it another way, the reverse of the other phase.
To add to this, the reference transformer is AC because the actual reference it needs is the sign wave of the phase to be able to calculate real-power. IoTaWatt could have just used the AC signal and have it rectified to get the DC it needs for the circuit inside, probably what Aeotec does or at least I hope, because if it uses a DC power supply, it is not giving you real-power readings and is probably trying to read off the current sensors or something and that's not very accurate IMHO.
But I guess they decided to go with 2 different transformers so that the rectifier circuit would not introduce noise or other things on the reference signal. This makes me believe that it must be very accurate and you will probably have basically the same readings as your power supplier.
I am finding mine to be very accurate. Typically within a few hundred watts on a monthly basis and I attribute that to what time of day the meter is read or polled.
When clamping CT's (current transformers) around your wires in the box, do you have to take extra care to run them at a 90 degree angle through the CT? E.g. if one of your clamps sags and the wire goes through at an angle will it affect the accuracy of the reading?
I haven't found that to be an issue. I have a 200 amp box using almost all half height breakers so wiring is really packed in there. Comparing them my multimeter it seems pretty accurate.
