Some charging faults are obvious. A shorted MOSFET, an open fuse, a damaged DC jack. This HP 14-dy0505 was none of those.

What initially appeared to be a failed charger IC evolved into a complex power sequencing fault where a collapsed internal bias rail prevented the entire charging architecture from starting. The diagnosis involved power rail analysis, battery recovery, IC replacement, and ultimately identifying a low-impedance capacitor on a critical internal supply rail.

Customer Complaint

The customer presented an HP 14-dy0505 with a charging fault.

The laptop would operate normally from a partially charged battery, but the battery itself was not charging. Eventually the battery became completely discharged, leaving the system unable to power up.

No obvious physical damage was present, and the charger was known-good.

Verifying Adapter Input

The first step was confirming that adapter voltage was reaching the motherboard correctly.

Using a genuine HP blue-pin adapter connected through a bench power supply:

• DC-IN measured 19.4V
• HP adapter ID line measured approximately 0.5V
• No abnormal current draw was observed

This immediately ruled out:

• Faulty power adapter
• Damaged DC jack
• Open DC-IN fuse
• Adapter identification failure

The motherboard was correctly receiving the adapter voltage.

Analysis of the Main Charging Circuit

This platform uses an ISL95338-controlled buck-boost charging topology.

The power stage consists of:

• Four switching MOSFETs
• A central inductor
• Input filtering network
• Battery interface circuitry
• ISL95338 charger controller

Architecturally, the design is very similar to modern USB-C and MacBook buck-boost power systems where the charger IC dynamically manages both adapter and battery power paths.

Initial measurements revealed:

This was a significant finding.

The board had valid DC-IN voltage, but the buck-boost converter was not switching at all.

The charger controller was clearly not enabling the power stage.

Focusing on the ISL95338

Attention shifted toward the charger controller.

After removing the motherboard, the charger IC was identified as an ISL95338 located on the reverse side of the board.

The investigation focused on startup and feedback signals.

One measurement appeared suspicious:

ADPS Pin Investigation

The ADPS pin (Adapter Voltage Sense) measured approximately 18.5V.

This raised questions because the pin normally receives a scaled representation of adapter voltage through an external resistor divider network.

Without a boardview available, tracing the divider network proved difficult.

Further probing produced inconsistent behavior, leading to the suspicion that the IC itself had suffered an internal failure.

Replacing the Charger IC

At this stage, the logical conclusion seemed straightforward.

A replacement ISL95338 was installed.

After reassembly and power application:

• DC-IN remained present
• Main rail remained at 0V
• Inductor remained inactive
• No charging current was observed

The replacement produced absolutely no change.

This is usually the point where a technician realizes the diagnosis has gone wrong.

The controller IC was not the root cause.

Battery Investigation

During further testing another abnormal condition was discovered.

The battery pack measured:

0V output

The battery protection circuitry had completely shut down.

Using an NLBA battery analyzer, the pack was evaluated externally.

The results showed:

• Cells remained healthy
• Pack protection was locked
• Deep-discharge condition detected

The cells were carefully recovered and balanced.

After recovery:

• Pack voltage returned to approximately 11.1V
• Protection lockout was removed

With the battery restored, expectations were that the charger circuit would resume operation.

Instead:

• No charging activity occurred
• No power rail generation occurred
• Adapter current remained near zero

The fault remained unchanged.

A Strange Discovery

Returning to the charger IC, a peculiar observation was made.

With the ISL95338 removed from circuit:

• ADPS measured 0V

This was unexpected.

Previously the pin measured approximately 18.5V.

The finding suggested that the voltage was not originating from the expected external divider network but was somehow being coupled internally through the IC itself.

Without boardview documentation, significant time was spent tracing surrounding circuitry and validating assumptions.

As an experiment, a temporary jumper was installed to feed ADPS through an alternative divider path.

The result:

Nothing changed.

No rail generation.

No startup activity.

No charging.

Investigating Internal Bias Power

The breakthrough came after revisiting the ISL95338 datasheet and focusing on startup requirements rather than feedback signals.

One rail immediately stood out.

VDD Measurement

The ISL95338 contains an internal LDO regulator that generates a 5V VDD rail used to power internal control logic and gate driver circuitry.

Expected measurement:

5V

Actual measurement:

1.8V

This was the first genuinely abnormal signal encountered during the repair.

The charger controller cannot execute startup logic with its primary internal bias rail collapsed.

Effectively, the IC was alive enough to show activity but not healthy enough to initialize the converter.

The Critical Experiment

To verify the theory, a current-limited bench power supply was connected directly to the VDD rail.

A regulated 5V was injected.

The response was immediate.

Within seconds:

• Status LED illuminated
• Charger IC initialized
• Main buck-boost converter started switching
• System rail rose to approximately 14.2V
• Standby current stabilized around 10-20mA

The power section had suddenly come back to life.

The strange part came next.

When the external 5V injection was removed, the system continued operating normally.

Even after removing a previously installed diagnostic jumper, the converter remained functional.

Root Cause Analysis

At this point the evidence pointed toward a fault on the VDD rail rather than the charger IC itself.

Further inspection identified a filter capacitor connected to the 5V LDO output.

The capacitor exhibited characteristics consistent with:

• Leakage
• Partial short circuit
• Low-impedance loading of the rail

The defective capacitor was dragging the VDD rail down from its intended 5V operating level to approximately 1.8V.

Because VDD powers the ISL95338's internal logic, the controller never completed initialization, preventing:

• MOSFET gate drive generation
• Buck-boost switching
• Main rail creation
• Battery charging

The charger IC itself was perfectly functional.

The entire motherboard was disabled by a single capacitor loading a bias rail.

Permanent Repair

To prevent recurrence, the suspect capacitor was removed and replaced.

Final verification showed:

The laptop booted successfully, displayed video, and charged the battery normally.

Technical Takeaways

This repair highlights several important diagnostic lessons:

1. Never assume the controller IC is faulty simply because the converter is inactive.
2. Always verify internal bias rails before replacing power-management ICs.
3. A partially shorted ceramic capacitor may not show as a dead short yet can completely disable a power subsystem.
4. Buck-boost charger controllers depend heavily on healthy internal LDO supplies for startup sequencing.
5. Battery lockout conditions can create distracting symptoms that mask the true underlying fault.

Conclusion

The HP 14-dy0505 ultimately suffered from a deceptively simple fault: a defective capacitor loading the ISL95338's 5V VDD rail.

That single component pulled the internal bias voltage down to 1.8V, preventing the charger controller from initializing and leaving the entire buck-boost power stage dormant.

The repair serves as a reminder that some of the most frustrating motherboard faults are not caused by expensive ICs or power MOSFETs, but by tiny passive components hidden on seemingly insignificant support rails. Sometimes the difference between a dead motherboard and a fully functional one is just a few microamps leaking through the wrong capacitor.