1
\$\begingroup\$

I am currently looking at this inexpensive power meter modules with TTL interface.

It seems to use internally the same ASIC as many chinese wall-wart power meters: It comes with a current transformer for measuring the current through a live wire and has connections for sampling the voltage on the wire.

My question now is this: Assuming I have a 3-phase AC setup, can I thread all three live phases (i.e. shifted 120° relative to each other) through the current transformer to measure the sum of the POWER in all three phases?

My hunch is that this will not work. My primary concern is that the voltage on only a single phase will be sampled while the current transformer sees the current on three phases which will make calculating the power difficult. Also I believe on AC the direction of the current in the three phases will generally not be aligned leading to cancellations.

Is this correct?


Edit 1: Clarification. I am not attempting to measure a 3-phase load but multiple single-phase loads attached to the three phases. The return current goes through neutral.

\$\endgroup\$
  • \$\begingroup\$ "The return current goes through neutral." Generally it doesn't. If the phase loads are equal then the neutral current is zero. If current is coming in on one phase it is going out on at least one of the others. Neutral only carries the imbalance in phase currents. \$\endgroup\$ – Transistor Aug 30 '18 at 21:22
  • \$\begingroup\$ @Transistor "Neutral only carries the imbalance in phase currents." Indeed. Since a load has only connections to a single phase and neutral, they produce by definition 100% phase imbalance, no? \$\endgroup\$ – ARF Aug 31 '18 at 8:09
  • \$\begingroup\$ No. If there are three equal loads the net current on the neutral is zero and it's not required - even if the loads are connected L-N. However, for single-phase loads that would be bad practice because if one load switched off the other two wouldn't have the correct voltage. \$\endgroup\$ – Transistor Aug 31 '18 at 15:03
3
\$\begingroup\$

The current sum through all phases is zero, given you don't have neutral current. If you have, things are even more complicated.

You need one current transformer per phase and have to multiply it with the phase voltage before summing. By multiplying with the phase voltage, the phase difference vanishes.

\$\endgroup\$
  • \$\begingroup\$ Ok, my question was not clear. In Germany residential housing has 3 phase connections to the grid. The loads are distributed among the three phases. But each load is sill single phase. Therefore, the return current from all phases goes through neutral. \$\endgroup\$ – ARF Aug 30 '18 at 17:54
  • \$\begingroup\$ This makes things even more complicated. The current sum through all phases isn't zero but the negative value of the neutral then. But because you don't measure the neutral current, it's just an arbitrary value. And even if you measured it, it's still making up zero. That's worth nothing. You have to measure each phase by itself, then multiply with the corresponding voltage to get rid of the phase difference. Then, you can sum it. \$\endgroup\$ – Janka Aug 30 '18 at 17:59
  • \$\begingroup\$ Thanks for confirming this. I had a "hunch" that this was the case, but my thinking was not very clear. Thanks for explaining. \$\endgroup\$ – ARF Aug 30 '18 at 18:03
2
\$\begingroup\$

can I thread all three live phases (i.e. shifted 120° relative to each other) through the current transformer to measure the sum of the POWER in all three phases?

If you mean directly through the aperture of the CT well, it won't cause a safety issue but the magnetic fields will sum to zero on a balanced load/supply and you won't be able to use it to measure power.

Information: You can run two phases through the CT aperture to get a current reference that is in phase with the voltage reference. For example, if your three phases are red, blue and yellow you can route red forwards then blue in the reverse direction to produce a net current phasor that is in phase with your red-blue line voltage.

Maybe this is an alternative idea? I've seen 3 ph power meters use it but I can't vouch for its accuracy and it will need scaling appropriately.

\$\endgroup\$
  • \$\begingroup\$ The net current phasor idea is shrewd. Can you tell me if this only works for 3-phase loads or also whether it would also work for multiple single-phase attached to each phase? In Germany residential housing has 3 phase connections. The loads are distributed equally to the three phases. But each load is sill single phase. Therefore, the return current from all phases goes through neutral. My question was not clear. Sorry for that. \$\endgroup\$ – ARF Aug 30 '18 at 17:56
  • \$\begingroup\$ It only works for 3 phase because it shifts the current reference by 30 degrees and aligns it to the phase of line voltage for unity power factor load. Your question was clear enough for me. \$\endgroup\$ – Andy aka Aug 30 '18 at 20:21
1
\$\begingroup\$

The power meter interface is designed for single phase. It cannot have a slightest idea what is going on, if you put three wires to its current transformer, its continues to assume all output is caused by single current. Unfortunately that output from the transformer doesn't reflect at all the total 3 phase power. Besides you have no place for 2 of the phase voltages.

In theory you can get some statistical estimate of the total power if you have a multiplexer that connects one phase (voltage and current) at a time to the system and changes the phase under measurement ater every few AC cycles. But that kind of multiplexer surely costs 500% more than having one interface for each phase.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.