Hi,
Just got my 25K-ET installed a few days ago and still experimenting with everything.
Noticed there's a large discrepancy in the power reported by SEMS/SolarGo and also the ModBus outputs.
When reading out the registers via modbus as configured right now, you get the following:
vpv1: PV1 Voltage = 696.9 V
ipv1: PV1 Current = 3.3 A
ppv1: PV1 Power = 1729 W
vpv2: PV2 Voltage = 696.9 V
ipv2: PV2 Current = 0.0 A
ppv2: PV2 Power = 2922 W
vpv3: PV3 Voltage = 647.4 V
ipv3: PV3 Current = 5.3 A
ppv3: PV3 Power = 0 W
ppv: PV Power = 4651 W
For context, my configuration population is 1 string each on both MPPT1 and MPPT2, with MPPT3 unpopulated.
The problem as you see in the above breakdown, is that the power calculation as described by the power registers isn't accurate.
In comparison, this is what SolarGo is reporting, which seems to be accurate as far as the population config:
<style>
</style>
| |
V |
A |
P calculated |
P modbus |
| PV1 |
696 |
3.3 |
2296.8 |
1729 |
| PV2 |
696 |
0 |
0 |
2922 |
| PV3 |
647 |
5.3 |
3429.1 |
0 |
| PV4 |
647 |
0 |
0 |
n/a |
| PV5 |
0 |
0 |
0 |
n/a |
| PV6 |
0 |
0 |
0 |
n/a |
| |
|
Total P |
5725.9 |
4651 |
In the SolarGo flow overview, the app is showcasing the 5.7kW figure as generation coming from the panels.
When I go into SEMS, that is reporting the 4.6kW figure, which seems to match the modbus power total.
What's actually even more odd, is that although the calculated figure seems more accurate, the breakdown of where the power is flowing afterwards seems to want to indicate that the ModBus/SEMS figure to be the accurate ones, as it's showing ~2.3kW into the backup/grid output and 2.3kW into the battery and 0W into the grid meter.
However, SolarGo is showing 3.5kW into the house, which I know to be inaccurate, and I also noted that SolarGo always overestimates the house power by a large amount which seems to correlate with the calculated vs mdobus power sum difference.
To sum it up, it's one giant mess across all the apps and reporting figures.
At the very least, what is certain, is that the modbus ppv2/ppv3 figures are not showcasing what they're labelled as, assuming the values they are showing are even accurate. Instead what they seem to be showing is per-MPPT string-aggregate power figures, and the V/A registers are showing per-string data, hence the mismatch in the labelling.
Unfortunately I don't have MPPT3 populated as it would be interesting to see what the modbus output even shows when power would be flowing through that, given that PV4/PV5/PV6 data doesn't exist in the current known mappings (I saw the mention that one might attempt reverse-engineer things from the app).
Another FYI: I have the whole house exclusively connected to the backup output, however one will note that the modbus pgrid/2/3 figures will be +- mirroring the backup load outputs even though they should be zero (as shown by the meter itself), so that data is also relatively untrustworthy and useless and forces you to refer to the meter data and calculations to do any possible grid-tied loads.
Hi,
Just got my 25K-ET installed a few days ago and still experimenting with everything.
Noticed there's a large discrepancy in the power reported by SEMS/SolarGo and also the ModBus outputs.
When reading out the registers via modbus as configured right now, you get the following:
For context, my configuration population is 1 string each on both MPPT1 and MPPT2, with MPPT3 unpopulated.
The problem as you see in the above breakdown, is that the power calculation as described by the power registers isn't accurate.
In comparison, this is what SolarGo is reporting, which seems to be accurate as far as the population config:
<style> </style>In the SolarGo flow overview, the app is showcasing the 5.7kW figure as generation coming from the panels.
When I go into SEMS, that is reporting the 4.6kW figure, which seems to match the modbus power total.
What's actually even more odd, is that although the calculated figure seems more accurate, the breakdown of where the power is flowing afterwards seems to want to indicate that the ModBus/SEMS figure to be the accurate ones, as it's showing ~2.3kW into the backup/grid output and 2.3kW into the battery and 0W into the grid meter.
However, SolarGo is showing 3.5kW into the house, which I know to be inaccurate, and I also noted that SolarGo always overestimates the house power by a large amount which seems to correlate with the calculated vs mdobus power sum difference.
To sum it up, it's one giant mess across all the apps and reporting figures.
At the very least, what is certain, is that the modbus ppv2/ppv3 figures are not showcasing what they're labelled as, assuming the values they are showing are even accurate. Instead what they seem to be showing is per-MPPT string-aggregate power figures, and the V/A registers are showing per-string data, hence the mismatch in the labelling.
Unfortunately I don't have MPPT3 populated as it would be interesting to see what the modbus output even shows when power would be flowing through that, given that PV4/PV5/PV6 data doesn't exist in the current known mappings (I saw the mention that one might attempt reverse-engineer things from the app).
Another FYI: I have the whole house exclusively connected to the backup output, however one will note that the modbus pgrid/2/3 figures will be +- mirroring the backup load outputs even though they should be zero (as shown by the meter itself), so that data is also relatively untrustworthy and useless and forces you to refer to the meter data and calculations to do any possible grid-tied loads.