0
\$\begingroup\$

I'm looking for a cheap way to monitor my electricity consumption across multiple power plugs, on a per-wire basis (ie. I'm not looking for total power consumption across all wires)

There are many commercial products I can use either at the power plug level or directly in the electrical cabinet, but they are designed to monitor a single wire, and cost usually around $15, so monitoring multiple wires skyrockets the global price quite fast, not to mention the space issue for fitting all those ammeters.

Also, I would have to use multiple clamps in a small space, and I'm not sure this wouldn't cause unwanted artifacts in the measurements.

Is there an easy way to achieve this? I was thinking it should be possible to monitor a single wire and build a circuit to switch a different power cord on that wire every second or so. Is this a stupid idea, or does it make sense?

Or are there ammeters so small and cheap that it makes sense to connect one in series with every single fuse in my electrical cabinet?

I was considering building a prototype over a 12vac supply. Am I going to run into scaling issues when upgrading to 110 or 220vac?

\$\endgroup\$
3
  • 1
    \$\begingroup\$ "I was thinking it should be possible to monitor a single wire and build a circuit to switch a different power cord on that wire every second or so" That indeed is not a good idea. You should not switch powerlines like that, electrical equipment that is using that line will not thank you (they will break). I suggest you get some $15 power monitors and use these on in a socket. I do not see how you could implement a cheaper solution. Any changes in your electrical cabinet will always cost more, are dangerous and must be done by a professional ! \$\endgroup\$ Commented Jan 4, 2016 at 11:24
  • \$\begingroup\$ @FakeMoustache, for a typical electrical cabinet with 40 fuses, it would double the cabinet size and cost 600$+. I'm not considering doing it tomorrow, I just want to work on this on 12vac current as an educational project, but I'm curious whether it could work in the real world as well - if done properly of course! \$\endgroup\$
    – Brann
    Commented Jan 4, 2016 at 11:27
  • \$\begingroup\$ Related question: electronics.stackexchange.com/questions/116195/… plus, you are way out on your pricing: alibaba.com/showroom/cheap-current-sensor.html \$\endgroup\$
    – Andy aka
    Commented Jan 4, 2016 at 11:54

3 Answers 3

3
\$\begingroup\$

Multiplexing the actual AC current is not a good idea. If you're going to multiplex anything, it should be the small current signals.

That means you are going to have a sensor per AC line. The simplest is probably to use dedicated current transformers. The clip on transformers are for convenience, but more expensive and larger than dedicated fixed-mounted current transformers. You'll have to do a little re-wiring, but only once. I would include the burden resistors with each transformer.

Once you have the two current transformer output wires for each AC line you want to measure, you have more options. One possibility is to have these all come to one board that does the absolute value and low pass filter functions on each signal, and has a microcontroller scan the results. You adjust the low pass filters so that the micro can scan each signal without aliasing. The micro then reports all the aggregated measurements to wherever you want them, or displays them, or whatever. You didn't say what you wanted to do with the results.

However, if the purpose is to measure power, then you really need to take the waveforms into account. Lots of devices don't look resistive and can have significant power factors. Put another way:

  Watts = ave(volts x amps)

The product of average volts and average amps is NOT power, and has little meaning. There is something called the "VA" (Volts x Amps) measurement:

  VA = rms(volts) x rms(amps)

but that's not real power either, and even this requires RMS volts and amps, not just average. The ratio of Watts to VA is the power factor. Only a pure resistive load has a power factor of 1 such that you can get away with computing VA and using it as real power.

If you are trying to measure power consumption, like what the power company bills you for, you want real power, not VA. This is not so simple to measure because you have to sample the voltage and current many times per power line cycle, take the product of each of the sample pairs, and low pass filter the result. Multiplexing all that would be difficult, so it effectively means a microcontroller per AC line. These can then all dump out their data over a CAN bus or something, which allows collecting it all in one place.

\$\endgroup\$
0
0
\$\begingroup\$

Trying to switch the current being measured will require heavy duty switches and careful coordination, and will add complexity aroud the safety equipemnt (fuses, current balances etc) so don't attempt to do that.

You shouldn't worry about multiple clamps in close proximity, the clamps don't radiate much nor will they pick up much interference.

You could have multiple clamps (or instead one piece current transformers) each with burden resistor connected full time and switch between them using small signal relays them taking readings from each periodically.

\$\endgroup\$
2
  • \$\begingroup\$ Thank you Jasen; I guess my question wasn't clear enough, I've just updated it. What I'm looking for is a way to monitor individual power consumption, so I would need one clamp per wire ! I'm looking for a way to monitor n different values with only one clamp \$\endgroup\$
    – Brann
    Commented Jan 4, 2016 at 11:37
  • \$\begingroup\$ ok. edited that solution out. \$\endgroup\$ Commented Jan 4, 2016 at 11:43
0
\$\begingroup\$

No you can't.

You will find, however, that the majority of devices use constant power when on (e.g., a heater, lamp, TV, etc.) and that if, instead, you measure their consumption once (how many watts they consume) and thereafter just record their 'on' time. From these two pieces of information you can calculate the kWh consumed by each item.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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