Top-board standalone evidence
If a supported model can boot or progress farther with one layer separated, that is useful isolation evidence. It does not automatically prove every circuit on that layer is healthy.
IPHONE DIAGNOSTICS / NO POWER / CURRENT FORENSICS
Read the current.
Read the current.
Find the first missing stage.
Combine DC current shape with power rails, clock/reset, NAND access, USB state and panic/restore evidence. No single amp reading should be treated as a final component diagnosis.
SOURCE-DERIVED POWER-ON TREE
01BATT / MAIN
02AON
03KEY / 32K
04AP BUCKS
05NAND
06I²C / OS
Source V1 organizes the sequence from main battery/system rails through always-on, key/clock, AP bucks, NAND and I²C/OS stages.
00 — MODEL / BOARD CONTEXT
Set the architecture before judging a measurement
Board construction, connector routing, PMIC generations and normal low-resistance rails change by iPhone family. Use exact boardview/schematic data whenever available.
01 — IPHONE POWER SEQUENCE
Trace where the power-on chain stops
Click each stage to inspect source-derived rails, expected behavior and the safer interpretation of a missing stage.
02 — PSU CURRENT SIGNATURES
Current draw is a clue, not a verdict
The seven source signatures are preserved, but V2 converts each into a probability branch with explicit confirmation tests.
SELECTED CURRENT PATTERN
03 — MASTER POWER / GR DATABASE
Rail, voltage, GR, resistance and signal state
Values are retained as uploaded-source references. Exact model/board known-good data should take precedence.
| Stage | Rail / Signal | Operating Voltage | GR / Diode | Resistance to GND | Signal Type | What Abnormal Supports |
|---|
04 — LIVE BENCH EVALUATOR
Compare one reading without over-diagnosing it
V2 explicitly distinguishes hard short, low reading, open circuit and voltage dropout from the final failed component.
05 — BENCH DECISION ENGINE
Combine current, rail state, USB and thermal evidence
The engine ranks fault domains from combined evidence. It intentionally avoids one-number component verdicts.
06 — FORENSIC CASE DOSSIER
Preserve before / after proof
Record the measurements that justified the repair and the measurements that confirmed it afterwards.
07 — SPLIT-BOARD / SANDWICH FORENSICS
Isolation first, heat second
The source includes useful split-board ideas, but V2 removes universal temperature and “100% healthy” claims.
Main-rail short isolation
Measure the short on each separated layer and identify which side still carries the abnormal load before thermal injection or component removal.
CPU/RAM low resistance
Very low resistance can be normal on high-current core rails. Never inject main-rail voltage into a CPU-core rail; use the exact rail voltage/current limit appropriate to the board.
NAND / restore crossover
Restore failure plus NAND-rail collapse supports a storage-power branch. Error percentage alone should not be used as a universal NAND diagnosis.
08 — EVIDENCE STANDARD
What V2 refuses to assume
The source contains several direct one-to-one mappings. V2 keeps them as source hypotheses unless supported by a confirming measurement.
DFU is a boot-state clue
Check storage power, clocks, USB mode and restore behavior before NAND rework.
Restore errors are stage evidence
Use exact restore logs and board measurements; do not map one error number to one component universally.
Read the panic sensor name
Use the exact panic string to identify the missing sensor or bus before replacing the charging flex by assumption.
Current shape needs confirmation
Check panic/restore/storage/I²C and board-layer evidence before sandwich rework.