![[ZIGBEE] 20A Zigbee socket with energy metering (Tuya TS011F_plug_3)](https://obrazki.elektroda.pl/1000454900_1749493372_thumb.jpg)
Hello.
I have created a separate topic separating from the thread: https://www.elektroda.pl/rtvforum/topic4095968.html
topic PV Inverter Broker .
Together with @Daro1003 we are testing the program inverterbroker .
This is my own program written under Linux, initially running on a router running OpenWrt, then on a laptop running Debian to finally sit on a Raspberry Pi.
It communicates with the Sofar inverter, passing data from it via the Wifi logger LSW-3 to the mqtt server, mosquitto. Further this data is at your discretion, I use node-red.
The source code and binaries for Raspberry processors are here: https://github.com/michalzd/pvInverterBroker/tree/mqtt.
For this you need the library: https://github.com/michalzd/MQTT_LinuxLib
I've had the program running on a RaspberryPi 1B+ for over a year now, and the information from it is used to turn on the DHW cylinder heat pump and additionally the heaters when the voltage exceeds 252V without letting it exceed the magic 253V.
I have an on-grid Sofar 15KTL-X G3.
Colleague @Daro1003 has installed at his place and is testing on a Sofar HYD 15KTL hybrid inverter.
It looks like the modbus data is identical across the KTL family[/i].
Added after 10 [minutes]:
I separated for one reason.
@Daro1003's tests show an anomaly:
as long as the panels are giving some energy, the power transmitted by the inverter is correct, but once it is dark and there is no energy from them, the power value :
is from the hours when it is still light, the power sent to the grid P 310 or 31W (the power the inverter returns in decaWatts) :
.
In the moment of darkness it is 65531
Today it occurred to me that this is a number with a sign. As the modbus register data is 16-bit, I converted 65536 - 65531 = -5 W.
Daro wrote back to me on pw: "This may be correct because I have 6W per phase on the PZEM 004T meter."
As I also have phase voltages and phase currents among the information from the inverter, I calculated this as well.
It comes out for the selected phase: 241.4 V * 0.02 A = 4.828 W. A little less, but let's say that's it.
The question is why the power shown in the ActivePowerOutputTotal register is not the sum of all phases?
This is only the case when there is no power from the panels, if they are working the ActivePowerOutputTotal power is correct.
At my place I am not able to check this, I do not have a hybrid.
I have created a separate topic separating from the thread: https://www.elektroda.pl/rtvforum/topic4095968.html
topic PV Inverter Broker .
Together with @Daro1003 we are testing the program inverterbroker .
This is my own program written under Linux, initially running on a router running OpenWrt, then on a laptop running Debian to finally sit on a Raspberry Pi.
It communicates with the Sofar inverter, passing data from it via the Wifi logger LSW-3 to the mqtt server, mosquitto. Further this data is at your discretion, I use node-red.
The source code and binaries for Raspberry processors are here: https://github.com/michalzd/pvInverterBroker/tree/mqtt.
For this you need the library: https://github.com/michalzd/MQTT_LinuxLib
I've had the program running on a RaspberryPi 1B+ for over a year now, and the information from it is used to turn on the DHW cylinder heat pump and additionally the heaters when the voltage exceeds 252V without letting it exceed the magic 253V.
I have an on-grid Sofar 15KTL-X G3.
Colleague @Daro1003 has installed at his place and is testing on a Sofar HYD 15KTL hybrid inverter.
It looks like the modbus data is identical across the KTL family[/i].
Added after 10 [minutes]:
I separated for one reason.
@Daro1003's tests show an anomaly:
as long as the panels are giving some energy, the power transmitted by the inverter is correct, but once it is dark and there is no energy from them, the power value :
is from the hours when it is still light, the power sent to the grid P 310 or 31W (the power the inverter returns in decaWatts) :
.
In the moment of darkness it is 65531
Today it occurred to me that this is a number with a sign. As the modbus register data is 16-bit, I converted 65536 - 65531 = -5 W.
Daro wrote back to me on pw: "This may be correct because I have 6W per phase on the PZEM 004T meter."
As I also have phase voltages and phase currents among the information from the inverter, I calculated this as well.
It comes out for the selected phase: 241.4 V * 0.02 A = 4.828 W. A little less, but let's say that's it.
The question is why the power shown in the ActivePowerOutputTotal register is not the sum of all phases?
This is only the case when there is no power from the panels, if they are working the ActivePowerOutputTotal power is correct.
At my place I am not able to check this, I do not have a hybrid.
.
