3 Phase current meter with SCT013 sensors

---

I think that adding a voltage Sensor for each phase would be as invasive as adding a 3 phase smart meter.

Hm… yes I guess so

on openEnergyMonitor i found a good way to measure my 3 Phase energy.
The EmonTX Shield is a cheap way to measure CT and actual AC with a Wall Plug. So i only need Plug on my DIN Rail an transformation to Power the Arduino and maybe the Raspberry.
The Shield is capable of sending data over 433Mhz or directly via serial to a Raspberry

https://openenergymonitor.org/emon/emontxshield
https://wiki.openenergymonitor.org/index.php/EmonTx_Arduino_Shield#emonTx_Arduino_Shield
here you can find some informations on this shield. now i’m thinking if i should give it a try to receive the date via 433mhz with my raspberry for pimatic, if i connect a raspberry zero via serial to push the data to somewhere or if i try to connect the arduino uno via WiFi to read out the Data.

so now some of my energy monitor hardware has arrived.
At the moment i have an Arduino Nano with an ESP8266 connected via SoftwareSerial and an EmonTX Shield on the Arduino Nano to monitor Voltage and Temperature. My Current Sensor have not arrived yet.
But the system is working and every 5 seconds a post with the actual data is send via the ESP to my personal server hosting an emoncms instance.
Next Step is to get the current sensors and the to get the calibration done to measure real power usage with actual Voltage.

I think the hardest thing will be the installation in the creepy stuff around my DIN Rails. Someone really build some crazy stuff there -.-

Wow! Very nice! How is the esp connected? Directly mounted on the board or sepatated? Ah ok now i see, thats supposed to be an esp01 on the right below the arduino. Maybe I should consider to move my setup to that solution one day.

Well yes, at my electronic enclosure it’s really narrow between the cables. The sct clamps are very huge and only fitted with some pressure

First had to figure out where the main energy cables are running. since I am having a photovoltaic on my roof there is a lot of stuff in the enclosure ^^

Well this is a bloody test situation. The esp runs directly on 5v which is not pretty good but works for testing.
In the final setup I want to use an arduino uno (that’s why the emon is build as a shield) and then I think the esp also will be fitted as a shield.
I really hope that the clamps will fit into my enclosure otherwise I have a little problem
When I’m finished I will show up the code and the build up

So i have my 3 Phase Energy Monitoring running for 1 month now. In comparison with my ferraris counter i have an accuracy of +/- 0,1kwh that is very well i think.
at the moment i still have the esp01, the arduino nano, and the shield hanging around on the clamps on the wall. The ESP is still powered over 5V an is sending every 10s to my EmonCMS instance. I have no problems with this setup at the moment. Next step would be to get some values from emonCMS into pimatic to have it displayed in one place.
I’m measuring the current on all 3 phases and the voltage on phase 1 and using this value for calculating the watts on all 3 phases.
I think that and also the building of a propper housing for the hardware i will wait till i moved into my new flat.
On software side i’m using at the moment only the openenergymonitor.org software

sounds very good and with your accuracy of +/- 0.1 kw/h it is a very reliable value!!
very nice!!
did you change something on the sketch for the emontxshield to reach this accuracy?

to reach the accuracy i did some calibration. at first i had to adjust the voltage, therefor i plugin an energy meter and changed the voltage calibration factor till it fits to the value of the energy meter, then i plug in a water boiler to get a pure resistive load an measured this again with the energy meter and adjusted the phase shift in the sketch till it fits an i had a power factor nearby 1.00. I repeater this step with the boiler for all 3 phases.
Sometime i have negativ values on some phases with a small consumption during the day but the overall value is within the accuracy.

the calibration way is described on the openenergymonitor site in the building blocks area

ah ok, i see! thx for the explanation!