Home/ Symptoms/ no-power
Symptom Hub · All Boards

No Power Diagnosis

7 boards 32 diagnostic checks Updated March 15, 2026

Overview

A board that shows no signs of life — no fan spin, no LED, no response to power button — requires a systematic approach. The correct first measurement is always the always-on bus rail (PPBUS_G3H on Apple, or the primary VIN rail on GPUs and consoles). Working top-down through the ACPI power state chain (G3H → S5 → S0) eliminates entire fault zones with each measurement. Never shotgun-replace components without following the rail sequence.

Diagnostic Methodology

Follow these steps in sequence. Each step eliminates an entire fault zone — do not skip ahead.

1. Confirm power input
Verify the battery or adapter is delivering voltage at the board connector. Dead input = no board fault. Check fuse continuity before measuring ICs.
2. Measure always-on rail
PPBUS_G3H (Apple) or primary VIN (GPU/console). If absent with confirmed input: PMU/charger IC, short to ground, or blown fuse.
3. Check for shorts
Set DMM to diode mode. Measure PPBUS_G3H to GND. <0.3 V reading = shorted component on bus. Inject voltage and use thermal camera or freeze spray to locate.
4. Verify standby rails
PP3V3_S5, PP1V8_S5, PP5V_S5 must be present when machine is 'off'. Missing = PMU not sequencing out of deep-off state.
5. Check power-on signal
Apple: SMC_ONOFF_L must pulse low when power button is pressed. GPU: PWR_EN from riser. Console: standby MCU enable line.
6. Follow S0 sequence
Only after S5 rails confirmed present. S0 rails (PP1V0_S0, VCORE, etc.) come up sequentially. Missing S0 rail = sequencer or VRM fault.

Per-Board Fault Trees

Board-specific checks ordered by failure likelihood. Most common root cause listed first.

MacBook Air A2337 (M1) Start: PPBUS_G3H (8.6 V typ)
⊞ Repair guide ◈ Schematic hub
  • PPBUS_G3H absent
    T6022 BMIC not generating bus rail. Check battery voltage at connector (should be 10.8–12.6 V). Inspect T6022 region for liquid damage corrosion under IC.
  • PP3V3_S5 absent
    T6022 not transitioning to S5 standby state. Verify PPBUS_G3H is present first. Check SMC equivalent reset line from T6022.
  • Short to GND on PPBUS
    Diode mode: PPBUS_G3H pin to GND. <0.25 V = active short. Common culprits: C7135/C7136 area (bulk caps on PPBUS), liquid damage under T6022.
  • No response to USB-C
    CD3215 on port not asserting VBUS detection. Check VCONN. Try both ports. If only one side dead: replace CD3215 for that port.
  • Board powers on briefly
    Thermal shutdown or current fold-back. Load test PPBUS_G3H rail with dummy load. Check M1 VRM output caps for ESR.
MacBook Pro A1989 (Touch Bar) Start: PPBUS_G3H (12.6 V typ)
⊞ Repair guide ◈ Schematic hub
  • PPBUS_G3H absent
    ISL9239C charger not outputting. Check SMC_BC_ACOK signal — must be 3.3 V with adapter connected. If absent, ISL9239C or surrounding passives.
  • SMC_BC_ACOK absent
    ISL9239C not detecting adapter. Check D6900 (TVS diode), fuse F6900, and VBUS on USB-C port. Very common after surge damage on USB-C ports.
  • PP3V3_S5 absent, PPBUS OK
    T2 (T8012) not generating S5 standby. Check PP3V3_G3H to T2 power domain. Liquid under T2 is common after keyboard spills.
  • Short on PPBUS_G3H
    Check area around ISL9239C and bulk capacitors. Inject 4 V / 1 A through PPBUS rail to locate short thermally.
  • No power-on from button
    SMC_ONOFF_L not pulsing when button pressed. Verify PP3V3_S5 present first. Check power button cable and flex connector.
MacBook Air A2337 / Pro A1989 — Liquid Damage Protocol Start: None — clean board first
⊞ Repair guide
  • Step 1: Do NOT power on
    Powering a liquid-damaged board with residue present causes additional corrosion and shorts. Disassemble first.
  • Step 2: Ultrasonic clean
    Isopropyl 99%+ or ultrasonic bath with distilled water. Remove all visible corrosion. Pay attention to T6022/ISL9239C/T2 areas.
  • Step 3: Inspect under USB-C controllers
    CD3215 / CD3217 VCONN pins are highly vulnerable to liquid. Reflow or replace if corrosion is visible.
  • Step 4: Measure after cleaning
    Only after fully dry. Follow standard no-power sequence starting from PPBUS_G3H.
MacBook Pro 820-01949 (A2141 — Intel + AMD) Start: PPBUS_G3H (12.6 V typ)
⊞ Repair guide ◈ Schematic hub
  • PPBUS_G3H absent
    ISL9239C not outputting. Check D7000 fuse and VBUS on all 4 USB-C ports. Also check SMC_BC_ACOK with known-good Apple 96W adapter.
  • PP3V3_G3H present but dead
    T2 chip (T8012) power domain fault. Verify PP3V3_G3H → T2 pin delivery. T2 reset line (T2_RESET_L) should be high.
  • Short on PPBUS
    AMD Radeon GPU AVDD/MVDD or main bus caps. Inject voltage current-limited and use FLIR or freeze spray. Reflow rarely fixes — reballing required.
  • Draws <100 mA then dead
    SMC not booting. Check PP5V_S5 from T2. Inspect T2 solder balls — reflow under T2 may recover.
GTX 1060 (GP106) Start: PCIe slot 12 V pin A1–A3
⊞ Repair guide ◈ Schematic hub
  • No 12 V from PCIe slot
    Motherboard power or riser. Test with known-good GPU in same slot. Check slot fuse/PTC on motherboard.
  • 12 V present, no POST
    uP9512 VRM not generating VGPU. Check PWR_EN from PCIe (must be high). Measure VGPU output at GPU core caps.
  • uP9512 not starting
    Check EN signal, VIN, and PGOOD feedback. Measure driver output (PHASE node) — should be switching at 300–500 kHz.
  • GDDR5 VDDQ absent
    VDDQ regulator (separate from core VRM). Check enable signal from GPU power sequencer. Measure GDDR5 ball array voltage.
  • Short on 12 V
    Remove PCIe power connectors (if 6/8-pin present). If short clears: check MOSFETs in VRM. If persists: GDDR5 or GP106 die.
GTX 1080 Ti (GP102) Start: PCIe slot 12 V + 8-pin aux
  • No 12 V aux (8-pin)
    Check 8-pin connector continuity and PSU rail. GP102 requires 75 W from slot + 150 W from 8-pin (dual 8-pin on some SKUs).
  • VGPU absent
    PWM controller (uP9512 / equivalent). Check ENOP (enable) signal from GP102 power sequencer. Measure PHASE switching.
  • GDDR5X rails absent
    GP102 uses GDDR5X memory (higher bandwidth, higher power). VDDQ on GDDR5X is generated separately. Trace from PWM to GDDR5X.
  • Short on VGPU rail
    GP102 die short is common after thermal stress (thin die). Inject 1.0 V / 3 A current-limited. Use thermal camera to locate.
  • Card detected, no display
    VGA arbiter fault — this is not a power fault. See /symptom/no-display/ for GPU display faults.
MacBook Pro A1278 (Intel only) Start: PPBUS_G3H (8.6–12.6 V)
  • PPBUS_G3H absent
    ISL6259 charger IC on A1278. Check SMC_ADAPTER_EN, SMC_BC_ACOK. Inspect MagSafe connector for bent pins and corrosion.
  • PP5V_S5 absent
    SMC not generating standby. Check SMC crystal (25 MHz) and SMC supply rails. Dead SMC = PP5V_S5 absent.
  • Short on PPBUS
    Common culprit: ISL6259 or nearby capacitors. A1278 liquid damage is very common around DC-in board connector area.
  • No power-on
    SMC_ONOFF_L not detected. Check keyboard/power button connector. Verify PP5V_S5 present — SMC must be alive to respond.

Frequently Asked Questions

What does PPBUS_G3H mean?
PPBUS_G3H is Apple's notation for the always-on bus rail — the primary voltage rail generated from either the battery or the charger adapter. G3H = ACPI G3 state (hard off), meaning this rail is present even when the machine is 'off'. It's the starting point for any no-power diagnosis on Apple logic boards.
The board draws current but doesn't boot. What is this?
This typically indicates the board is reaching a partial power state (S5 standby) but failing to enter S0 (active). Measure PP3V3_S5, PP5V_S5, and PP1V8_S5 — all three must be present. If they are, measure the power-on signal (SMC_ONOFF_L pulse on Apple). If SMC_ONOFF_L is present but nothing happens, the issue is in S0 bring-up (CPU VCore, PCH, etc.).
Is thermal camera required for this diagnosis?
No, but it significantly speeds up short-circuit location. A DMM in diode mode is the minimum required tool. Inject current-limited voltage (4 V / 500 mA) through a bench power supply into the suspected rail and use freeze spray to locate cold spots, or touch-to-warm if safe. FLIR / thermal camera is optional but recommended for GDDR5/GPU die shorts.