Fix 10x internal AC voltage/frequency on AC500

[viliusdidit]

Problem

DecimalField.parse() computes (raw / 10**scale) * multiplier. AC500 defines:

DecimalField(FieldName.INTERNAL_AC_VOLTAGE, 71, 1, 10),
DecimalField(FieldName.INTERNAL_AC_FREQUENCY, 74, 2, 10),

The AC500 reports these registers in tenths/hundredths (same as AC300), so the multiplier=10 cancels the scale and both values render 10x too high — 230.4 V shows as 2304 V, 50.0 Hz as 500 Hz.

Fixes Patrick762/hassio-bluetti-bt#248.

Verification

Measured on a real AC500 (grid-tied): with this fix internal_ac_voltage reads 233–238 V, tracking the same unit's ac_input_voltage (no multiplier, always correct) within 0.3 V, and internal_ac_frequency reads 50.0 Hz matching ac_input_frequency. Without it, the same polls render 2334–2380 V / 500 Hz.

Parse sanity check: raw 2334 → 233.4, raw 5001 → 50.01. Full test suite passes (73 passed).

Why AC500 only (and #145)

AC200L/M/PL share the same field pattern but there the multiplier=10 is intentional: that family reports integer volts, and the multiplier was added from device feedback to correct 23 V → 230 V (Patrick762/hassio-bluetti-bt#145, integration commit b37276e) — the exact opposite direction of the AC500 bug. That likely means #145 is already resolved on current 0.2.x and just needs reporter confirmation; if reports still show wrong values there, the fix would mirror this one on the AC200x definitions. I'm leaving those untouched — I have no AC200-series hardware to verify against, and this family has already been bitten once by cross-device assumptions.

🤖 Generated with Claude Code

[matthiasexl]

I do have two AC500 which report it wrong. But I also have an AC200M which reports it correctly. If you need more info, just let me know what I should do/which info you require.