L1 L2 issue in graphs

[Invy55]

I hope this is in the right place. I have some questions regarding voltages across the dashboard expecially in the vitals and in the voltages panel. This is currently what I can see:

First of all I'm having trouble understanding what L1 and L2 should be, since my home in europe is on a single phase (230 V) and that's the usual naming for the three phases (as a disclaimer, I have two powerwall on the same phase L1). As you can see the y axis for the light blue and green series has something wrong. I've looked on the github for similar issues and the only thing I found is this comment from the owner that mentions a regex to fix some issue with the american system (that I can't figure out). While this might work in USA, removing the regex fixes the issue for me. I'm now wondering if this is an intended behaviour or some sort of bug. Also from what I can understand from the quoted post I should not be seeing the second phase.

The second strange thing that makes me wonder if there is some glitch in my configuration is, as I was saying, the voltages in the vitals. As you can see they are somewhat wrong (160 V ??):

Thanks in advance

[jasonacox]

Wow! that is interesting.

The data for the voltages is coming from this API:

http://localhost:8675/freq

And you can see the different phase leg to leg (L1L2) and leg to neutral (L1N) values that is used to graph the data:

{
  "PW1_name": "x--",
  "PW1_PINV_Fout": 60.004000000000005,
  "PW1_PINV_VSplit1": 120.30000000000001,
  "PW1_PINV_VSplit2": 119.9,
  "PW1_PackagePartNumber": "2012170-25-E",
  "PW1_PackageSerialNumber": "x",
  "PW1_p_out": 710,
  "PW1_q_out": 20,
  "PW1_v_out": 240,
  "PW1_f_out": 60.004000000000005,
  "PW1_i_out": -16.7,
  "PW2_name": "TEPINV--x--x",
  "PW2_PINV_Fout": 60.004000000000005,
  "PW2_PINV_VSplit1": 120.4,
  "PW2_PINV_VSplit2": 120.10000000000001,
  "PW2_PackagePartNumber": "3012170-05-B",
  "PW2_PackageSerialNumber": "x",
  "PW2_p_out": 670,
  "PW2_q_out": 10,
  "PW2_v_out": 240.20000000000002,
  "PW2_f_out": 60.004000000000005,
  "PW2_i_out": -15.9,
  "ISLAND_FreqL1_Load": 60,
  "ISLAND_FreqL1_Main": 60,
  "ISLAND_FreqL2_Load": 60,
  "ISLAND_FreqL2_Main": 60,
  "ISLAND_FreqL3_Load": 0,
  "ISLAND_FreqL3_Main": 0,
  "ISLAND_GridConnected": "ISLAND_GridConnected_Connected",
  "ISLAND_GridState": "ISLAND_GridState_Grid_Compliant",
  "ISLAND_L1L2PhaseDelta": null,
  "ISLAND_L1L3PhaseDelta": null,
  "ISLAND_L1MicrogridOk": null,
  "ISLAND_L2L3PhaseDelta": null,
  "ISLAND_L2MicrogridOk": null,
  "ISLAND_L3MicrogridOk": null,
  "ISLAND_PhaseL1_Main_Load": null,
  "ISLAND_PhaseL2_Main_Load": null,
  "ISLAND_PhaseL3_Main_Load": null,
  "ISLAND_ReadyForSynchronization": null,
  "ISLAND_VL1N_Load": 120.5,
  "ISLAND_VL1N_Main": 120.5,
  "ISLAND_VL2N_Load": 121.5,
  "ISLAND_VL2N_Main": 121.5,
  "ISLAND_VL3N_Load": 0,
  "ISLAND_VL3N_Main": 0,
  "METER_X_CTA_I": 1.94,
  "METER_X_CTA_InstReactivePower": -24,
  "METER_X_CTA_InstRealPower": 95,
  "METER_X_CTB_I": 1.8985,
  "METER_X_CTB_InstReactivePower": -85,
  "METER_X_CTB_InstRealPower": -112,
  "METER_X_CTC_I": 0,
  "METER_X_CTC_InstReactivePower": 0,
  "METER_X_CTC_InstRealPower": 0,
  "METER_X_LifetimeEnergyExport": null,
  "METER_X_LifetimeEnergyImport": null,
  "METER_X_VL1N": 119.65,
  "METER_X_VL2N": 120.29,
  "METER_X_VL3N": 0,
  "METER_Y_CTA_I": 0,
  "METER_Y_CTA_InstReactivePower": 0,
  "METER_Y_CTA_InstRealPower": 0,
  "METER_Y_CTB_I": 0,
  "METER_Y_CTB_InstReactivePower": 0,
  "METER_Y_CTB_InstRealPower": 0,
  "METER_Y_CTC_I": 0,
  "METER_Y_CTC_InstReactivePower": 0,
  "METER_Y_CTC_InstRealPower": 0,
  "METER_Y_LifetimeEnergyExport": null,
  "METER_Y_LifetimeEnergyImport": null,
  "METER_Y_VL1N": 119.65,
  "METER_Y_VL2N": 120.28,
  "METER_Y_VL3N": 0,
  "grid_status": 1
}

[jasonacox]

That's good, that's the feeder for Voltages. For voltage it uses: http://localhost:8675/aggregates

[Invy55]

That's odd. On the new api endpoint the voltages are actually 160V. But that makes 0 sense, how could I have anything at that voltage in a 220v system

[jasonacox]

There must be something off in the logic or in the TEDAPI data...

Code that creates aggregates for TEDAPI connected systems is here: https://github.com/jasonacox/pypowerwall/blob/832dec530fe2c02ccfde5b3c393a65532568fcbf/pypowerwall/tedapi/pypowerwall_tedapi.py#L316-L509

Pulling raw data:

• http://localhost:8675/tedapi/status
• http://localhost:8675/redapi/config

[Invy55]

I think I've figured something out in the pypowerwall lib. Let me know if I should open an issue to your other github to post this. Anyway the issue is the following:
My system is a mono phase, 2 powerwalls (GW 2) european installation. My three phase readings are:

L1: ~240
L2: 7
L3: 0.5

This messes the compute_LL_voltage function that doesn't account for mono phase systems and doesn't account for slight fluctuation in voltages. I don't think my apis are the issue here, maybe there should be some rounding to avoid it mistakenly considering the system three phase.

For reference:

[Invy55]

If this is the only issue another fix could be using config -> site_info -> phase_usage (as noted by @martinmanyhats) to determine wich phases to consider in the logic.

[martinmanyhats]

FYI, seeing very similar numbers from tedapi/status for L1, L2, L3. Single phase UK, PW3. tedapi/config does include: "phase_usages": { "phase1": "Backup", "phase2": "NotConfigured", "phase3": "NotConfigured" }, martin

…

On Mon, 14 Jul 2025 at 11:18, Marco ***@***.***> wrote: I think I've figured something out in the pypowerwall lib. Let me know if I should open an issue to your other github to post this. Anyway the issue is the following: My system is a mono phase, 2 powerwalls (GW 2) european installation. My three phase readings are: L1: ~240 L2: 7 L3: 0.5 This messes the compute_LL_voltage function <https://github.com/jasonacox/pypowerwall/blob/832dec530fe2c02ccfde5b3c393a65532568fcbf/pypowerwall/tedapi/pypowerwall_tedapi.py#L29-L43> that doesn't account for mono phase systems and doesn't account for slight fluctuation in voltages. I don't think my apis are the issue here, maybe there should be some rounding to avoid it mistakenly considering the system three phase. — Reply to this email directly, view it on GitHub <#660 (reply in thread)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAI23OLVZDTBUQMMJJQOHV33IN7XFAVCNFSM6AAAAACBILPOQCVHI2DSMVQWIX3LMV43URDJONRXK43TNFXW4Q3PNVWWK3TUHMYTGNZVGA3DKMA> . You are receiving this because you are subscribed to this thread.Message ID: <jasonacox/Powerwall-Dashboard/repo-discussions/660/comments/13750650@ github.com>

[Invy55]

Very interesting, can you share the voltages you read in http://localhost:8675/tedapi/status :

I wanna confirm if my theory is correct

[jasonacox]

Excellent! You are right, it is not a configuration accounted for in the voltage calcs for pypowerwall. I've created an issue to track: jasonacox/pypowerwall#195