energy monitoring with a current transformer

[TrystanLea]

Hello Ladyada and anyone else interested. I have recently been building various energy monitoring / power metering bits and pieces both for mains AC energy use and wind turbine monitoring. I have so far been concentrating on the measurement side: Building circuits to do the job of the killawatt meter in your tweet-a-watt project. Ive had a some luck with a simple circuit using a clip on CT sensor connected up to an Arduino that may be of use to you in the tweet-a-watt project.

I'm measuring current using an efergy elite CT sensor and then "synthesising" the voltage waveform using the current waveform measurements. You just enter the rms Voltage and frequency, the arduino program creates a sine wave from these values that is aligned with the current waveform: the peak of the synthesised voltage waveform is made to align with the peak of the current waveform. Im not sure if this is the best way to do this but I seem to be getting good results the real power, and power factor measurements using the CT method seem to agree quite well with measurements from a plug-in power meter that measures both the voltage and current directly.

I have written a page about it including circuit schematics and arduino sketch here: http://openenergymonitor.org/emon/node/28

Here's a picture of the electronics:



A picture of the output graphed in KST.
http://openenergymonitor.org/emon/sites/default/files/Output.png

Results comparison between the CT method and a plug in power meter.
http://openenergymonitor.org/emon/sites/default/files/Results.ods

My next step is to follow your guide on the xbee's and posting to twitter, I have bought two xbees from you and built your adapters. Thanks for the fast delivery and great build documentation!

Hope the above can be of use

Trystan

[adafruit]

Thanks for posting this up! Its a common question we get here

[TrystanLea]

Thanks for the reply!

[richms]

Don't write off the changes that line voltage has on the power used - I see between 217 and 242 here all the time, and to a heater thats the difference between 1900 watts and 2400 - a substantial change in the cost to run the heater, and in the case of the samwige press the temperature it gets to.

Easiest case you could get a cheap 9VAC wall adapter and a resistor divider into the arduino.

[TrystanLea]

richms, your quite right, variations in voltage is the one of the largest sources of error with this method and the 9V adapter is a good solution to get around this problem.

My intention with only using the current transformer sensor and the preset voltage was to try and build an energy monitor that works in the same way as most CT commercial energy monitors you can buy. The efergy energy monitor I bought only uses a CT sensor and requires you to enter a value for voltage and so if its good enough for the commercial guys then its good enough for me.

If anyone wants more accurate measurements, I could write a tutorial on using a 9V ac adapter if anyone's interested?

[rik]

hi,

I'm new to arduino (received it last week). I bought it in order to build a home powerusage monitor just like you did.

I am certainly interested in the Voltage part via a 9V AC adapter. I've read your project, and will certainly try to make it by myself. Thanks for sharing the information.

Last week I "found" 2 efergy CT's in the shop. I do not find any information about the current reduction factor for those units. Do you have more information?


thanks in advance,

Rik

[TrystanLea]

Hey Rik
Great to hear that your going to try to build it!
The current reduction factor on the new efergy elite sensor seems to be 1500, the old efergy sensor (white housing rather than grey) seems to be 1000.
Im away at the moment but will be back at the weekend, will have a look at the voltage sensing early next week and let you know.
Ive been thinking of hopefully using the voltage adapter both as a power supply for the arduino so that it can be used in standalone mode with a 7segment display and for the voltage sensing, could be quite neat...will see how it goes.
Trystan

[SiliconFarmer]

John has nice writeup of his power meter project using Arduino + CT sensors at http://jarv.org/pwrmon.shtml. Includes theory, source code, even calibration info. He talks about power factor, peak vs RMS current, and how all that affects the power measurement.

I particularly like his use of Open Flash Charts. I plan to use them myself.

[rik]

Hi Lea,

I did some testing with your setup. It seems that the new efergy currentransformers ( distributed as "chacon" in Belgium) have a reduction factor of 3000. This weekend I will do some more testing with different kind of loads.

rik

[TrystanLea]

Hey Rik

Good to hear that you've got it up and running. Interesting that they all have different reduction factors.

Ive been doing a bit of work on the voltage measurement with the voltage adapter, Ive been logging the voltage since this morning and this is the plot I have so far:

Ignore the spikes, I was moving things about and accidentally touched the wires.
Im not sure if the rise at the start is due to the grid or the adapter warming up.
I think if I leave it for a few more ours, see if there is a drop towards the late evening.

Trystan

[TrystanLea]

Rik, do you by any chance have a link to the Belgium supplier? I cant seem to find it.

And also does anyone know of a supplier for these current transformers in the US?

Thanks

[rik]

The Belgian brand is CHACON. The product is known as ecowatt and is compatible with the 700, 800 and 850 (elite, eone and e2 from efergy). The website can be found at www.chacon.be But I have to say that the site is bad. The ecowatt product line can not be found. The only thing they do is repackaging the efergy product and distribute it to the different shops. (that's my opinion)

regards,

rik

[TrystanLea]

Thanks Rik

[TrystanLea]

Im sorry I haven't updated on the power adapter situation sooner. Im pretty sure the results shown in the graph above are due to the heating of the transformer, I mean grid voltage doesn't vary by 40V right? I think maybe the way to do this is to monitor the voltage on the high voltage side using maybe a voltage divider and optoisolator similar to this: http://openenergymonitor.org/emon/node/2 But this adds complication and safety concerns and so Im not sure if its worth pursuing this method... does anyone have any thoughts on this?

[diogenes]

A large number of the industrial sensors aren't isolated from the mains at all, which does introduce safety issues. The power supplies tend to be transformerless and use the neutral wire as ground throughout the board. It gives an easy to use ground reference for the ADCs and lets you run everything off a single PSU. The voltage input is then as simple as a resistor divider straight into an ADC, but the ADC needs to be capable of measuring negative voltage portions of the AC wave.

It also means that there is full mains voltage potential all over that board. Solid enclosures are the only way to keep something like that safe. I wouldn't recommend it for home use, but it eliminates some of the complexity for power supply and input circuits.

If you really want to get true voltage and current measurements going for the highest accuracy, I'd recommend using a dedicated energy measurement IC. Analog Devices makes some good ones, as does SAMES. I'd also recommend a good life insurance policy before you start playing around with the non-isolated signals.