Archived Forum

Did you ever get this working?  It seems like one problem would be the difference in phase, As the one leg rises, the other drops and I think this would cancel out a lot of the "signal".  You kind of get the same thing when you put a CT clamp over a pair of wires.

To (belatedly) answer ddcarter's question, I haven't tested either of those, so that's why I did not comment at the time he posed the question. As a rider, I wouldn't worry too much about getting exactly the correct burden value. Certainly, the correct value is best but I find it hard to decide what is "correct", as you don't know the shape of the current waveform, so you can't calculate the peak value. I usually suggest never more than 1.66 V rms for an emonTx Shield with a 5 V supply, and if you are likely to be working near maximum current, I'd aim somewhat lower unless you're confident that you have a good sine wave for your current.

There's absolutely no reason why paralleling two CTs should not work, provided of course that both are identical (otherwise, it will be impossible to apply the correct calibration and phase compensation).
However, officeboy, your description of a possible problem ignores the fact that the two should face opposite directions - obvious when you think of a 240 V load which takes its current from one leg and returns it to the other leg, you measure the current twice but only half the voltage and so the power comes out correct.

You don't get the same thing with a 2-core cable because in this case, the ampere-turns cancel in the CT core. But you CAN use 1 CT if you pass the cable for one leg backwards through the CT (think about the 240 V load again and twice the current × half the voltage).

Is this the same for North American power with two 120V legs (240 combined) and a neutral?  Many household circuits will only use one leg of power and the return will be via the neutral cable. (Kind of a center tap on our 240V transformers)   

I do understand and agree that measuring with two CT's on a balanced 240 V load can be done as described, like on a water heater. 

(Although I don't understand why you wouldn't just double the voltage or wattage with math instead of using two CT's)

"you wouldn't just double the voltage or wattage with math [sic] instead of using two CT's" because, if you're measuring the power draw of the whole installation, there will also be 120 V loads that are not necessarily - in fact almost certainly will not be - balanced, therefore the two line currents will be unequal and there will be a neutral current. If you're only measuring a single 240 V load, then you can of course measure the current in one leg only, because given no third connection the current has nowhere to go and must return in the other leg.

Yes, I should have been more clear, my comment about using math to do the work was only for a single balanced load (like the water heater).  

Back to ddcarters issue, on measuring incoming mains.  Can you use two CT's connected in parallel, or do they need to be measured individually?  

See above.