Board 820-00163 — MacBook Air 13" 2015 (A1466)

The 820-00163 is the main logic board for the MacBook Air 13" Mid 2015 (A1466), featuring an Intel Broadwell Core i5 or i7 processor and integrated graphics. This board is distinguished by its compact form factor, reliance on the ISL6259 charger controller for power delivery, and the ITE IT8913E embedded controller that handles keyboard, touchpad, and thermal management—making it a common point of failure for both charging and input device issues.

Key Test Points & Voltages

Net / Test Point	Power State	Expected Voltage	Notes
`PPVRTC_G3H`	Off (S5)	3.16–3.32 V	RTC rail, powered by coin cell + main battery. Persistent across sleep. Critical for wake signals. Measure at `C4138`.
`PP3V42_G3H`	Off (S5)	3.38–3.46 V	GPU/display aux rail. Always-on when battery present. Key for lid wake. Test at `C4232`.
`PPBUS_G3H`	Off (S5)	~11.9–12.6 V	Main unregulated bus from charger or battery. Primary indicator of charging circuit health. Measure before ISL6259 input `VBAT_CHRG`.
`PP5V_S5`	Off (S5)	4.8–5.1 V	Secondary 5 V rail (always-on when PPBUS present). Powers charger IC logic and auxiliary circuits. Derived from buck converter; check at `C4145`.
`PP3V3_S5`	Off (S5)	3.24–3.36 V	Auxiliary 3.3 V always-on. Powers EC standby logic and sensor circuits. Measure at `C4179`.
`PP1V8_S3`	Off (S5), present in S3	1.76–1.84 V	Core I/O rail for EC and memory. Only active when CPU domain alive. Disappears in deep sleep.
`PP3V3_S3`	Off (S5), present in S3	3.24–3.36 V	Broadwell I/O supply, keyboard controller rail. Dies when EC enters S5. Check for shorts here if keyboard is dead.
`PPVCORE_S0`	S0 only	0.9–1.05 V (varies with load)	CPU core voltage under Broadwell DVFS. Only valid during OS boot and runtime. Ramps from 0.75 V during P-state transitions.
`PP1V05_S0`	S0 only	1.02–1.08 V	Uncore/GPU IO rail. Present only with CPU active. Test near graphics output.
`PP5V_ISENSE`	Always (when USB-C power connected)	4.9–5.0 V	Charger feedback for ISL6259. Falls below 4.8 V indicates bad charger IC or input protection diode `D4100`.

Diode Mode Reference

Use a multimeter set to diode mode (200 mV or equivalent). Place black probe to GND (chassis ground, e.g., GND via). Red probe touches each test point. Readings should be forward-biased (0.4–0.7 V typical for silicon; 0.2–0.4 V for Schottky). A reading of 0 V or > 0.8 V indicates potential shorts or open junctions.

Test Point / Net	Red (−) to GND	Expected Range	Common Bad Reading	Notes
`VBAT_CHRG` (ISL6259 pin 1)	Forward junction	0.48–0.58 V	0.0 V (open) / 1.2+ V (shorted)	Direct input to charger. Bad reading almost always means dead ISL6259. Measure both battery and external charger paths.
`D4100` (anode, charger input protection)	Forward bias	0.32–0.42 V	0.0 V (burnt out) / Open (delaminated)	Schottky diode. Failure here blocks external charger input. Board will not charge on AC, only battery.
`PP5V_S5` node at output of buck converter	Forward bias (cap leg)	0.50–0.62 V	0.0 V (shorted output) / 0.95+ V (missing regulation)	If 0 V, buck converter is shorted; unplug charger immediately. If > 0.65 V, converter controller or coil failure.
`PP3V3_S5` (3.3 V always-on rail)	Forward bias	0.54–0.64 V	0.0 V (short) / 0.8+ V (open converter)	Shorts on this rail cause ISL6259 to brown-out and restart. Measure at multiple cap pads to isolate shorted region.
`PP3V3_S3` (EC I/O rail)	Forward bias	0.54–0.64 V	0.0 V (severe short) / 0.75+ V (regulator dead)	If dead with `PP3V3_S5` alive, regulator `U4200` (often PWM controller) has failed. Keyboard dead fault often starts here.
`PP1V8_S3` (memory/core I/O)	Forward bias	0.50–0.60 V	0.0 V (shorted rail) / 1.0+ V (regulator open)	Faults here cause boot hangs or immediate shutdown. Common point for liquid damage ingress near RAM slots.
`PPVCORE_S0` (CPU core supply)	Forward bias (off state only)	0.48–0.56 V	0.0 V (core short / VRM dead) / > 1.5 V (open circuit)	Only test with system off. If shorted, CPU or power delivery module is defective. Do not power on.
`IT8913E` pin 1 (`VDD`)	Forward bias	0.55–0.65 V	0.0 V (EC short) / 0.9+ V (missing supply)	EC chip supply. 0 V reading = IC failure or shorted decap. This will kill keyboard, trackpad, and thermal control.
`CS42L83` pin (codec supply, typically `AVDD`)	Forward bias	0.52–0.62 V	0.0 V (codec short) / Open (missing supply trace)	Audio codec. Failure here kills all audio output. Low impedance reading suggests capacitor leakage or IC delamination.
`PPVRTC_G3H` (RTC battery supply)	Forward bias	0.50–0.58 V	0.0 V (shorted by CMOS battery) / 1.8+ V (open regulator)	RTC rail often fails from aged coin cell or regulator leakage. If 0 V, suspect CMOS backup diode shorted.
`PWR_BTN` to GND (power button sense)	No diode effect expected	Open or 4.7 kΩ pull-up resistance	0 Ω (button stuck pressed) / Short to GND	If stuck 0 Ω, button latch or EC input is shorted. Board will not stay off.
`SMC_RESET` signal (EC watchdog pin)	Float / open	No diode effect (logic signal)	Stuck low / stuck high	Use logic probe or scope. Stuck low = EC reset loop. Stuck high = EC not initializing. Critical for diagnosing hang states.

Common Faults & Symptoms

No Charge — ISL6259 Failure

Symptom: USB-C charger connected; LED does not illuminate; system runs on battery but will not charge. LED remains dark even after 30 seconds.

Likely Cause: ISL6259 charger controller IC (U4101) has failed, or input protection diode D4100 is open. This is the single most common fault on 820-00163 boards. The ISL6259 manages input current limiting, voltage regulation, and battery charging state. Failure typically follows electrolyte leakage or thermal cycling.

Diagnostic Step: (1) Measure PPBUS_G3H with USB-C charger connected. If <6 V, input path is compromised—check D4100 with diode mode. If D4100 reads 0.0 V or open, replace it first. (2) If PPBUS_G3H is 12+ V but PP5V_S5 is missing, the ISL6259 internal buck converter has failed. (3) Measure PP3V3_S5 at ISL6259 feedback pin. If 0 V, the chip is offline. (4) Look for charger LED control—typically CHRG_LED_N pulled low by ISL6259. If LED is pulled high (open-drain driver dead), ISL6259 is not responding.

Fix: Replace U4101 (ISL6259, BGA 49-pin package) with a known-good part. If diode D4100 tested bad, replace it first (SOD323 Schottky, typical part: MBR120T1G). Test secondary rails (PP5V_S5, PP3V3_S5) after ISL6259 replacement. Note: Do not substitute charger ICs—this board requires ISL6259 specifically for USB-C power delivery negotiation.

Keyboard Dead — ITE EC Failure

Symptom: System boots normally; display works; trackpad may or may not respond; all keyboard keys produce no input. No lights on keyboard when powered on. Fn + brightness keys also dead.

Likely Cause: ITE IT8913E embedded controller (U3800) has crashed, failed to initialize, or lost firmware. The IT8913E communicates with the keyboard matrix via a GPIO/I2C protocol and also manages thermal sensors, battery gauge, and SMC functions. Failure is often triggered by liquid damage near the EC (keyboard area), corrupted firmware from a bad flash, or EC supply rail collapse.

Diagnostic Step: (1) Power on system and measure PP3V3_S3 and PP1V8_S3 at EC supply pins. Both should be stable and within spec. If either is 0 V or missing, the EC power tree has failed—check for shorts on decoupling caps near U3800 or regulator U4200. (2) Probe SMC_RESET (EC watchdog signal, typically a GPIO to PMU or PMU to EC). If stuck low, EC is in reset loop; if stuck high, EC did not boot. Use an oscilloscope to verify reset pulse timing. (3) Check for water damage or corrosion on EC pads and surrounding area, especially keyboard connector. (4) Attempt to recover EC firmware: connect USB-C power only (no battery), wait 30 seconds, then press power button. If LED blinks slowly, EC has entered DFU mode and may be recoverable. (5) If available, use a Sunrisepoint PMU monitor or BDI debugger to verify EC clock and reset signals.

Fix: If EC firmware is corrupted (determined by DFU-mode detection), attempt reflash via external programmer if you have EC firmware image. For hardware failure: replace U3800 (BGA 88-pin IT8913E). More commonly, the fault is a shorted cap or regulator failure in the EC power supply—systematically replace each decoupling capacitor on PP3V3_S3 and PP1V8_S3 near the EC, then test. For persistent cold-solder on EC BGA, reflow the entire EC footprint.

Diagnostic Workflow (No Power)

Use this procedure when the board does not turn on at all: no fan spin, no LED, no chime, no display backlight, even with AC charger connected.

Visual Inspection: Examine ISL6259 (U4101) for visible damage, lifted balls, corrosion, or heat damage. Check for liquid residue around charger input connector. Inspect all large capacitors in the ISL6259 output stage for bulging or leakage.
Measure PPBUS_G3H (Charger or Battery Input): Connect USB-C charger. Probe PPBUS_G3H at capacitor C4100 anode (charger-side input). Expected: 12.0–13.2 V if charger is present, or 10.5–12.6 V if running on battery. If 0 V, the input path is open or shorted. If present, advance to step 3. If absent with charger plugged in, check diode D4100 next.
Test Input Protection Diode D4100: Disconnect charger. Set multimeter to diode mode. Red probe to D4100 anode (charger side); black to ground. Expected: 0.32–0.42 V (Schottky forward bias). If 0.0 V (open) or > 0.5 V (shorted), replace D4100 immediately. If D4100 is good, reconnect charger and remeasure PPBUS_G3H; if still 0 V, the ISL6259 is offline or short-circuiting the input.
Measure Secondary Rails PP5V_S5 and PP3V3_S5: With charger connected, probe the output of the ISL6259 buck converter. Expected: PP5V_S5 = 4.8–5.1 V; PP3V3_S5 = 3.24–3.36 V. If both are 0 V and PPBUS_G3H is present, the ISL6259 converter is not running. Check if PP5V_S5 feedback is shorted (measure with diode mode: should be 0.50–0.62 V at capacitor). If feedback appears shorted (0.0 V), there is a short on PP5V_S5 rail—check all major capacitors and power-tree ICs on this rail.
Check ISL6259 Control Signals: Look for charger LED control line (typically CHRG_LED_N) at ISL6259 open-drain output. Probe with logic analyzer or oscilloscope. If LED line is stuck high (not pulled to GND), ISL6259 is not asserting control. Measure ISL6259 supply pins (pins 2, 48—VDD_CHRG) with diode mode. Expected: 0.54–0.62 V. If 0.0 V, the ISL6259 is dead or severely shorted.
Measure Always-On EC Rails: Probe PP3V3_S5 at EC decoupling capacitors near U3800. Expected: 3.24–3.36 V. This rail is independent of ISL6259 and should come from a separate regulator. If missing, the EC power tree regulator (U4200 or similar) is dead. Measure the input to this regulator (typically 5 V or 12 V) and the feedback network.
Probe EC Reset and Power-Good Signals: Connect oscilloscope to SMC_RESET and PWR_BTN_SENSE (power button). When plugging in charger, PWR_BTN_SENSE should see a clean pull-up to 3.3 V. SMC_RESET should see a short low pulse (typically 100–500 ms) during startup. If SMC_RESET is continuously low, the EC is in reset loop—check for missing clocks to EC (32 kHz oscillator) or a shorted VDD.
Attempt Forced Power-On: If all rails are present and EC appears to initialize, press the power button firmly for 3 seconds. If the board remains inert, measure PPVCORE_S0 and PP1V05_S0—these should ramp up to 0.9+ V and 1.05 V respectively if the CPU is booting. If they remain at 0 V, the CPU power domain is stuck off—investigate the PMU (power management unit) inputs and PSYS rail.
Thermal Monitoring: Feel the ISL6259 thermal footprint after 30 seconds with charger connected. It should be warm but not painful to touch. If scalding hot, the ISL6259 is short-circuiting—immediately unplug charger and check for shorted power planes or output cap failures.
If Still No Power After Step 8: The fault is likely in the CPU power delivery or EC firmware. Proceed to reflash EC firmware (if DFU mode is accessible) or replace the ISL6259 and primary power regulators. If you have already replaced ISL6259 and secondary rails are stable but the board still will not boot, suspect CPU or memory corruption—the board may require micro-soldering or return to depot level service.

Related Boards

The 820-00163 shares the Broadwell power delivery architecture with 820-00164 (MacBook Air 13" 2015, same generation but different logic revision) and 820-00138 (MacBook Pro 13" Retina, 2015 model with similar ISL6259 charger stage). However, the 820-00163 is a smaller form factor and has a more compact layout—decoupling capacitors are closer together, increasing risk of liquid damage propagation. The 820-00164 introduced a revised ISL6259 gate driver stage that improved reliability under high-current charging. The main schematic differences are in the keyboard matrix routing and EC GPIO assignment, but the power tree is nearly identical, allowing some component cross-reference. Do not confuse 820-00163 with 820-00165 (MacBook Air 13" 2017)—the 2017 model uses a different EC (IT8987E) and does not share ISL6259 charger topology.