Skip to content

v0.9.4 - MIN TL-XH Feature + Multiple fixes

Choose a tag to compare

View all tags
@0xAHA 0xAHA released this 21 Jun 06:19
v0.9.4
a53152a

Release Notes

Buy Me A Coffee

v0.9.4

Issues: #311, #323, #326, #327

  • Feature: MIN TL-XH priority mode control (Issue #311):
    Register 3018 (tl_xh_priority_mode) hardware-confirmed on MIN 4200TL-XH: write 0 → Load First, 2 → Battery First, 3 → Grid First (default). Appears as a select entity under the Battery device on MIN_TL_XH_3000_10000_V201 profiles. Note: value 1 is not a valid priority mode on this hardware (V1.39 maps it as a system topology setting).

  • Fix: WIT battery_voltage_bms reads 1/10th of actual (Issue #323):
    Register 8095 (battery_voltage_bms) on WIT inverters with JK BMS returns whole volts (e.g., raw 54 at 54.0 V). The previous scale: 0.1 was producing readings of 5.4 V. Scale corrected to 1.

  • Fix: WIT solar_total_power spikes to 429 MW (Issue #323):
    The WIT pv_total_power 32-bit register pair (regs 1–2) was missing signed: True. When the
    inverter sends a small negative value (e.g. at night or during certain grid conditions), it was
    read as an unsigned 32-bit integer — 0xFFFFFFFF × 0.1 = 429,496,729.5 W. Now correctly
    resolves to approximately −0.1 W.

  • Fix: WIT vpp_export_limit_w entity always shows Unknown (Issue #323):
    Holding register 203 (export_limit_w) was defined in the WIT profile but never polled or
    stored in the data object. The number entity read from a field that was never populated.
    Register 203 is now read each poll cycle and stored in data.export_limit_w.

  • Fix: Grid import/export and load sensors missing on SPH 3/6kW and 7/10kW (Issue #326):
    Power-flow registers 1015/1016 (power_to_user), 1021/1022 (power_to_load),
    1029/1030 (power_to_grid), and 1037/1038 (self_consumption_power) were present in
    SPH_8000_10000_HU and the V2.01 profiles but missing from the two base profiles
    SPH_3000_6000 and SPH_7000_10000. Without them, grid sensors fell back to a derived
    value that reported ~2.6 kW export when the inverter was actually importing. Both base
    profiles now include the full power-flow register set.

  • Fix: Spurious WARNING log before every control write (Issue #327):
    The integration intentionally closes the Modbus socket after each read cycle. When a
    write is requested, is_socket_open() correctly returns False and the code reconnects
    before writing — this is by design, not a fault. The "Socket not open, attempting
    reconnect" message has been downgraded from WARNING to DEBUG.

Assets 2
Loading