You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Hello,
currently, there's no simple way (without headaches) to get the plain PV power/energy values (without battery) when you have a battery installed.
This is because the modbus only reports the values of the "inverter" and "battery cable" but not the values of the "pv cable".
Following entities are currently available (and are correct):
ac_power
meterX_ac_power (Grid meter)
batteryX_power
ac_energy
meterX_exported/meterX_imported (Grid meter)
batteryX_charged/batteryX_discharged
For using the native energy dashboard or doing energy optimizations, the PV-energy/PV-power values (without battery included) are required.
Currently, i ended in calculating these 'virtual' values myself using a sensor template to calculate the PV-power value.
The sun state in the condition (and setting pv power to 0) is used to avoid short 'spikes' at night, when you have abrupt load changes.
The same is for the load_power (which can't be negative if you have no generators installed except the pv)
The export_power/load_power are for doing energy optimizations (switch on devices if enough energy is available).
To get the pv_energy, i used a riemann integration helper and a utility meter.
The utility meter is necessary to syncronize/set the inital value of the pv_energy.
The pv_energy should then run more or less sync with the ac_energy value over a whole day, because the battery enables you a 'shift in time'. (assuming the soc hasn't changed and when neglecting discharging losses)
Problematic at this solution is that i currently 'lose' pv_energy when home-assistent doesn't run. For such cases, it would be necessary to 'resync' the value to the 'ac_energy' value (but this only works simple when the soc hasn't changed meanwhile).
The same problems arise for the initial value as mentioned before.
A further problem is that the pv_energy/pv_power logically have no historic values until they have been created (but the solaredge values may already exist for a longer time) - so dropping the old data is necessary.
I know - it would be better if solaredge simply would add this 2 modbus registers before doing this junk - but this will not happen soon.
Maybe they don't want because then it would be easy to integrate, but this is one of the most important values (compared to shit like 'solaredge_m1_import_varh_q2_c' - does anybody know what this is in reality?)
The text was updated successfully, but these errors were encountered:
Hello,
currently, there's no simple way (without headaches) to get the plain PV power/energy values (without battery) when you have a battery installed.
This is because the modbus only reports the values of the "inverter" and "battery cable" but not the values of the "pv cable".
Following entities are currently available (and are correct):
For using the native energy dashboard or doing energy optimizations, the PV-energy/PV-power values (without battery included) are required.
Currently, i ended in calculating these 'virtual' values myself using a sensor template to calculate the PV-power value.
The sun state in the condition (and setting pv power to 0) is used to avoid short 'spikes' at night, when you have abrupt load changes.
The same is for the load_power (which can't be negative if you have no generators installed except the pv)
The export_power/load_power are for doing energy optimizations (switch on devices if enough energy is available).
To get the pv_energy, i used a riemann integration helper and a utility meter.
The utility meter is necessary to syncronize/set the inital value of the pv_energy.
The pv_energy should then run more or less sync with the ac_energy value over a whole day, because the battery enables you a 'shift in time'. (assuming the soc hasn't changed and when neglecting discharging losses)
Problematic at this solution is that i currently 'lose' pv_energy when home-assistent doesn't run. For such cases, it would be necessary to 'resync' the value to the 'ac_energy' value (but this only works simple when the soc hasn't changed meanwhile).
The same problems arise for the initial value as mentioned before.
A further problem is that the pv_energy/pv_power logically have no historic values until they have been created (but the solaredge values may already exist for a longer time) - so dropping the old data is necessary.
I know - it would be better if solaredge simply would add this 2 modbus registers before doing this junk - but this will not happen soon.
Maybe they don't want because then it would be easy to integrate, but this is one of the most important values (compared to shit like 'solaredge_m1_import_varh_q2_c' - does anybody know what this is in reality?)
The text was updated successfully, but these errors were encountered: