uzbek-plus/logs/b23/B23-RESULT.md

31 lines
2.5 KiB
Markdown
Raw Blame History

This file contains ambiguous Unicode characters

This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.

# B23 — faster battery charging: LBC fast-charge current 450 → 720 mA (2026-10-07) — **PASS, persistent**
## Change
One device-tree property only: `&pm8916_charger { qcom,fast-charge-safe-current = <720000>; }`
(`linux/dts/msm8916-jz08-aurora-b23.dts`, includes the B18 DTB). The pm8916_lbc supervisor (v4, B10B-5) programs
IBAT_SAFE and IBAT_MAX from it: code = current / 90 mA − 1 → **0x04 (450 mA) → 0x07 (720 mA)**. The simple-battery node
has no constant-charge-current property, so the safe current is also the charge current. Kernel (7.2.7-aurora-b18),
initramfs (B12Q) and cmdline are unchanged; the DTB differs from B18 only in this property and the model string.
Why not more: the LBC is a linear charger — PMIC dissipation is (VBUS − VBAT) · I; it has no input current limit,
so a weak source sags below VIN_MIN (4.308 V) instead of limiting. 720 mA was the first step (hardware max 1440 mA).
## RAM test (ram2, `b23/b23-boot.sh`, observer `b23/b23-mon.sh` every 30 s; nothing flashed)
- Probe: `IBAT_MAX 07`, supervisor INIT → CHARGING, CHG_STATUS 05 (CC), FAST_CHG_ON + USB path 02, VDD_MAX 08 (4.20 V).
- Source = laptop USB port: USB meter **0.724 A**; USBIN 4.67–4.78 V, no sag.
- CC held for 55+ min without restart/latch/FAULT (VADC VBAT under charge 3.98 → 4.10 V).
- PMIC temperature: max **55.7 °C** in the first 5 min, then 43–47 °C (part of the drop is the known ~7 °C
backlight-off artefact of the PMIC sensor). Alert threshold 60 °C never reached.
- CV (CHG_STATUS 03) from VADC ≈ 4.10 V; current then tapers (0.4 A on the meter) — normal, current does not speed up CV.
## Persistent
- Cache candidate `b23/cache/mkfs-cache-b23.sh` (B18 recipe, DTB b23) **01bfe522**; lk2nd extlinux emulator
(`b23/cache/emu/run-emu-b23.sh`): B18 regression identical, B23 boots label `b23` with DTB 89eade0f, fault variants revert.
- Written by `b23/B23-write.sh` (HW EDL): **B23_CACHE_WRITE_VERIFIED** — readback == image, all other partitions unchanged,
pre-write cache readback == B18 ac82811a kept. Rollback: `b23/B23-rollback.sh` → B18 `ac82811a`.
- Cold boot from eMMC (full power-off, no RESET): model "JZ08AU Aurora (B23 charge 720 mA)", `IBAT_MAX 07`, CHARGING.
## Effect / limits
CC phase (empty → ≈ 4.1 V) ≈ 1.6× faster; the CV tail is unchanged. Board load (CPU, backlight) comes out of the same
input current (no power path). Worst case (long CC from ≈ 3.7 V with CPU load) not yet measured thermally; still no
thermal derating in the supervisor.