5 Reasons Your Circuit Board Fails (And How We Fix It)

A circuit board can look clean and still fail during testing, installation, or daily use. A small layout issue, a weak solder joint, or a part placed under the wrong load may stay hidden until the board reaches a customer’s product.

When a board stops working, I do not start by replacing random parts. I trace the fault from the symptoms, check the design and production records, then test the board under controlled conditions. This helps separate a design problem from a process problem.

Here are five common causes I see and the steps I use to address them.

1. Poor solder joints

A weak solder joint can create an open circuit, unstable signal, or intermittent fault. The board may pass one test and fail during vibration, heat, or repeated use.

Common causes include:

• Too little or too much solder
• Incorrect reflow temperature
• Contamination on the pad
• Poor component placement
• Oxidized leads or pads

I inspect the joints under magnification and compare suspicious areas with the assembly drawing. For hidden joints, such as those under BGA packages, X-ray inspection can help reveal voids, bridges, and uneven solder coverage.

A practical example is a power board that worked on the bench but restarted when the enclosure was moved. The cause was a weak joint on a connector pin. The repair included reworking the joint and checking the connector’s mechanical support.

2. Incorrect component values or placement

A resistor with the wrong value can change a current level, timing signal, or sensor reading. A capacitor with the wrong voltage rating may fail after the board has been used for some time. A diode installed in the wrong direction can stop an entire section from operating.

I check the bill of materials, schematic, PCB layout, and assembly data as one set. Looking at only one file can leave gaps. I also compare the markings on the physical part with the approved component list.

For a control board that showed an unstable sensor output, the measured resistor value did not match the schematic. The part number looked similar to the approved item, but its resistance was different. Replacing it restored the expected signal.

3. Power supply stress

Many board failures begin in the power section. The input voltage may be too high, the regulator may run too hot, or a load may draw more current than the design allows.

I measure:

• Input voltage during startup
• Output voltage under load
• Current draw at idle and peak demand
• Heat around regulators and power components
• Ripple on sensitive supply lines

I also check whether protection parts are present and correctly rated. A fuse, TVS diode, or reverse-polarity circuit cannot help if it is missing, poorly placed, or not matched to the system.

A board that works from a laboratory supply may behave differently when connected to a long cable or a motor. Cable resistance, switching noise, and startup current can change the conditions at the board input.

4. Heat and poor thermal design

Excess heat can shorten component life and create faults that appear only after extended operation. Heat may come from a regulator, processor, power transistor, LED driver, or nearby high-current trace.

I use thermal measurements while the board runs under its normal load. I check copper area, thermal vias, airflow, enclosure space, and the distance between hot and temperature-sensitive parts.

A useful repair may involve more than adding a heatsink. The board may need a wider copper area, a different regulator, a lower switching frequency, or a change in the enclosure.

I also test the board after it reaches a stable operating temperature. A short bench test may miss a fault that appears after thirty minutes of use.

5. Layout and signal interference

A board may fail even when every component has the correct value. Poor grounding, long signal paths, weak decoupling, and high-current traces placed near sensitive lines can create noise or communication errors.

I review:

• Ground return paths
• Placement of bypass capacitors
• Separation between power and signal traces
• Differential pair routing
• Clock and high-speed line paths
• Connector pin assignments
• Trace width for expected current

For a communication board with random data errors, the issue was not the main processor. The signal path ran close to a switching node, and the return path was incomplete. Adjusting the routing and improving local decoupling reduced the noise during testing.

How I investigate a failed board

I use a clear fault-check process:

1. Record the failure condition
   I note when the board fails, what the user sees, and whether the fault appears at startup, under load, or after heating.

2. Check for visible damage
   I inspect burnt areas, cracked parts, lifted pads, corrosion, loose connectors, and signs of mechanical stress.

3. Measure power rails
   Incorrect voltage or excessive current can point to a short, damaged component, or poor power design.

4. Compare the board with its documents
   I review the schematic, layout, bill of materials, assembly files, and test limits.

5. Reproduce the fault
   I use the same load, input condition, temperature range, and cable setup when possible.

6. Confirm the repair
   I test the board again under normal conditions and record the result. A repair is not complete when the board merely powers on.

The most useful lesson is simple: a failed circuit board usually gives clues. The failure may come from assembly, component selection, power input, heat, or layout. Careful testing helps identify the actual cause instead of treating only the visible symptom.

If you are dealing with intermittent resets, unstable signals, overheating, or failed production tests, share the board symptoms, input conditions, and available design files. That information gives the investigation a stronger starting point.

When a circuit board starts failing, the symptoms can look confusing. A device may restart without warning, lose power, show unstable readings, or stop responding after several minutes of use. The board itself may not be the only cause. A damaged connector, weak solder joint, failed component, moisture, heat, or a poor power supply can create similar signs.

I usually treat the fault as a chain of evidence. The goal is not to replace parts at random. It is to find where the electrical path stops working.

Start with the symptoms

Write down what the device does before opening it.

• Does it fail as soon as power is applied?
• Does it work when cold and stop after warming up?
• Does pressing the board or moving a cable change the result?
• Are only some functions affected?
• Do lights, displays, or motors behave differently from normal?
• Did the problem appear after a drop, liquid spill, repair, or voltage change?

These details can point toward different causes. A complete loss of power may involve the input circuit or a broken connection. A single failed function may relate to one component, signal path, or solder joint.

Check the power source

A circuit board cannot work correctly with unstable input power.

I begin by checking the adapter, battery, fuse, switch, and power cable. I use a multimeter to measure the voltage at the board’s input points, not only at the wall outlet or battery terminals. The reading should match the value shown on the device label or service information.

A power supply can show the correct voltage with no load and still drop when the board starts operating. If the voltage falls sharply during startup, the adapter, battery, cable, or input protection circuit may need attention.

Never test an unknown board with a higher-voltage supply as a trial. That can damage several parts at once and make the original fault harder to find.

Look for visible damage

With power removed, inspect the board under a bright light.

Look for:

• Burn marks or dark areas
• Cracked components
• Swollen capacitors
• Green or white corrosion
• Loose connectors
• Lifted copper tracks
• Broken solder joints
• Foreign objects or metal debris

A burnt area shows that heat was present, but it does not always identify the original cause. A failed regulator may burn because of a shorted component farther along the circuit. Replacing the visibly damaged part without checking the surrounding area can lead to another failure.

Liquid damage often leaves a thin residue that is easy to miss. Corrosion may continue under connectors and integrated circuits even after the board appears dry.

Test for short circuits

A short circuit can prevent startup or cause a power supply to shut down for protection.

With the board disconnected from power, I check resistance between the main power rail and ground. A very low reading may suggest a short, but resistance measurements need context. Some boards naturally show low resistance because capacitors charge during the test or because certain circuits have low-resistance paths.

A thermal camera can help locate a part that warms up when limited power is applied. A current-limited bench supply is also useful for controlled testing. This work requires care. Applying power without current limits can turn a small fault into a burnt trace or damaged chip.

Inspect solder joints and connectors

Mechanical stress often causes faults that appear random.

Connectors near cables, switches, relays, and heavy parts receive repeated movement. Their solder joints can crack over time. The board may work when the connector is pressed and stop when the cable moves.

I inspect these areas with magnification. A dull, ring-shaped, or cracked joint may need rework. The correct repair depends on the board material, pad condition, and component type. Excess heat can lift a pad or damage nearby parts.

A useful test is gentle movement while the circuit operates, but the probe must not bridge two contacts. A short made by a tool can create a new fault within seconds.

Check for heat-related failure

Some boards operate normally for a few minutes and then stop. Heat can reveal weak components, poor solder joints, or damaged power regulators.

I compare the board’s temperature during normal operation with the temperature near the fault. A component that becomes unusually hot deserves closer testing. A component that stays cold in a circuit where it should be active may also indicate an open connection or missing input.

For example, a small control board in an industrial fan may run normally at room temperature but stop after the motor has operated for ten minutes. A weak voltage regulator or cracked solder joint can expand with heat and interrupt the circuit. Cooling the area with controlled airflow may help confirm the pattern, though it should not replace proper testing.

Check the clock and reset circuits

Digital boards can appear completely dead when their clock or reset circuit is not working.

If the main supply voltage is present, check whether the processor receives a stable reset signal. A damaged crystal, oscillator, reset supervisor, or related capacitor may stop the system from starting.

This type of fault often needs an oscilloscope or logic analyzer. A multimeter may show the correct supply voltage while missing a signal that is switching too quickly to measure accurately.

Review recent repairs or changes

A board that fails after maintenance may have a connection problem rather than a failed design component.

I check for:

• Reversed connectors
• Pinched cables
• Missing screws or grounds
• Solder bridges
• Incorrect replacement parts
• Damaged pads
• Components installed in the wrong direction

A common example appears in small audio equipment. After a connector replacement, the unit may power on but produce no sound because one ground or signal pin was placed in the wrong position. The visible board may look clean, yet the wiring order creates the fault.

Photographs taken before repair can help confirm the original connector position.

Use the circuit diagram when available

A schematic makes fault tracing more direct. It shows where power enters, how it is regulated, and which signals connect the major sections.

I follow the circuit from the input toward the failed function:

1. Confirm the input voltage.
2. Check protection parts such as fuses and diodes.
3. Measure each power rail.
4. Check the enable and reset signals.
5. Trace the signal path to the affected component.
6. Compare readings with a working board when one is available.

A board without a schematic can still be tested, but the process usually takes longer. Photos, board markings, component data sheets, and connector labels can provide useful clues.

Know when replacement makes more sense

Repair is not always the best choice. A board with damaged inner layers, burned processor pads, severe corrosion, or unavailable programmed components may cost more to restore than to replace.

I compare the board price, repair time, part availability, and risk of repeat failure. A replacement board should match the hardware version, firmware needs, connector layout, and voltage ratings. Two boards may look similar while using different software or pin assignments.

For equipment used in safety-sensitive settings, testing should follow the manufacturer’s service instructions and the applicable workplace procedures.

Protect the board during testing

Static electricity, loose tools, and incorrect probes can cause damage. I disconnect power before resistance or continuity tests, use suitable personal protection, and keep metal objects away from exposed contacts. Capacitors may hold charge after the device is unplugged.

If the board is connected to mains voltage, battery packs with high current, medical equipment, heating systems, or moving machinery, qualified service support may be needed.

A failing circuit board gives clues through timing, heat, voltage, movement, and visible changes. Careful testing turns those clues into a repair path. When I record each measurement and change only one condition at a time, I can separate a board fault from a power, wiring, or surrounding equipment fault without adding unnecessary damage.

A printed circuit board can fail for a simple reason: one solder joint is weak, one trace is damaged, or one component receives the wrong voltage. When a board stops working, replacing parts at random often creates more faults and makes the original problem harder to find.

I use a step-by-step check instead. I start with visible damage, review the power path, test for shorts, and then check signals. This method helps separate a design problem from a production defect, assembly error, or handling issue.

1. The board does not power on

When a PCB shows no LED activity and the main circuit does not respond, I check the power path before testing the processor or other complex parts.

I look at:

• The input connector
• Fuse or protection components
• Reverse-polarity parts
• Voltage regulators
• Ground connections
• Cracked solder joints
• Burn marks near the power section

A multimeter set to continuity mode can help confirm whether the ground path is complete. I then measure voltage at each stage of the power circuit. For example, a board designed for a 12 V input may use a regulator to create 5 V and 3.3 V rails. If 12 V reaches the input but the 5 V rail reads 0 V, the fault may be located around the regulator, its enable pin, or a short on the 5 V line.

I avoid applying power again and again when a short is present. A bench power supply with a suitable current limit can reduce the chance of extra heat damage during testing.

2. A short circuit appears between power and ground

A short between a supply rail and ground can stop an entire board. It may come from a solder bridge, a reversed capacitor, a damaged integrated circuit, or a copper fault.

I begin with the board unpowered. Resistance mode can show whether the rail has an unusually low reading, though some circuits naturally have low resistance because of coils, motors, or large capacitors.

A practical test looks like this:

1. Disconnect external modules and cables.
2. Remove the battery or power source.
3. Check the resistance between the suspected rail and ground.
4. Inspect fine-pitch components under magnification.
5. Use a current-limited supply at a safe voltage.
6. Look for the area that warms up first.

A thermal camera can help, but it is not required for every repair. A careful visual inspection, alcohol evaporation test, or touch-free temperature check may reveal the faulty area. I never touch an energized board to locate heat.

When the short is caused by excess solder, removing the bridge may solve the issue. When a chip becomes hot, I check its supply pins and nearby capacitors before replacing it. The chip may be damaged, but it may also be heating because another part is forcing the rail into an unsafe state.

3. Components are placed in the wrong direction

Diodes, electrolytic capacitors, LEDs, connectors, and integrated circuits can fail when their orientation does not match the board design.

I compare the part marking with:

• The PCB silkscreen
• The schematic
• The assembly drawing
• The component data sheet

The stripe on a diode, the negative mark on an electrolytic capacitor, and the pin-one indicator on an IC should match the intended footprint. A missing or unclear silkscreen can create assembly errors even when the circuit design is correct.

I once reviewed a small controller board where the LED did not light after assembly. The resistor value was correct, and the firmware worked on another board. The LED had been rotated because the polarity mark on the footprint was difficult to see. The fix was simple: update the silkscreen and add a clearer assembly note.

A board should not rely on memory or guesswork. Clear polarity marks reduce repair time and prevent repeated assembly mistakes.

4. Solder joints look dull, cracked, or uneven

A weak solder joint can create an intermittent fault. The board may work on a test bench and fail after vibration, temperature changes, or normal handling.

I inspect joints for:

• Cracks around the component lead
• Solder that has not fully covered the pad
• Excess solder between nearby pins
• Lifted pads
• Voids around large connectors
• Movement when the part is gently tested without power

A clean, properly heated joint usually has a smooth connection between the lead and pad. A dull appearance alone does not prove a fault, so I combine visual inspection with continuity and mechanical checks.

For a repair, I clean the area, add suitable flux, heat the joint evenly, and allow it to cool without movement. Large connectors often need support from both the solder joints and the mechanical structure of the product. If the connector receives repeated force, solder alone may not provide enough support.

5. The board works only sometimes

Intermittent failures are among the hardest PCB problems because the fault can disappear during testing.

I ask when the failure occurs:

• After the board warms up
• When a cable moves
• During motor startup
• At a certain input voltage
• After a software reset
• When several outputs operate together

This information helps narrow the search. I may use a gentle cable movement test, thermal cycling within a safe range, or an oscilloscope to observe the supply rail during startup.

A common example is a cracked joint beneath a large connector. The board works when the cable is still and stops when the cable is moved. Reworking the joint may restore operation, but I also check the connector mounting and enclosure design. If the cable pulls on the connector, the same failure may return.

6. Signals are missing or unstable

A board may power on while data, clock, reset, or sensor signals remain incorrect. I check the signal path from the source to the destination.

A useful order is:

1. Confirm the supply voltage at both devices.
2. Check that both devices share the correct ground reference.
3. Inspect pull-up or pull-down resistors.
4. Verify connector pin mapping.
5. Measure the signal with an oscilloscope or logic analyzer.
6. Compare the waveform with the interface requirements.

For an I²C bus, both SDA and SCL lines usually need suitable pull-up resistors. If one line stays low, the cause may be a short, a device holding the bus, or an incorrect voltage level. For SPI, I check clock, chip-select, data direction, and ground connections.

A logic analyzer can show whether the controller is sending data, but it cannot confirm that the receiving device is electrically healthy. I combine waveform checks with voltage and continuity tests.

7. The PCB becomes too hot

Heat can come from a wrong component value, excessive current, poor cooling, a shorted load, or an unsuitable power regulator.

I measure the temperature of the suspected part and compare the reading with the data sheet limits. I also check:

• Input and output voltage
• Load current
• Heat spreader area
• Airflow
• Copper width around high-current paths
• Component spacing
• Shutdown or protection behavior

A linear regulator can become hot when the difference between input and output voltage is large. For example, dropping 12 V to 5 V at a high load creates more heat than dropping 7 V to 5 V at the same load. The repair may involve reducing the load, improving heat transfer, or selecting a suitable power circuit during the design stage.

I do not treat heat as a component-only issue. The load and the power path must be checked together.

8. Traces or pads are damaged

A lifted pad or broken trace often appears after excessive heat, repeated connector stress, or a short circuit repair.

I trace the connection with a multimeter and compare it with the schematic. If a trace is open, a small insulated wire can sometimes restore the connection. The wire should be secured so that vibration does not pull on the repaired pad.

For high-current paths, a thin jumper wire may not be suitable. I check the expected current, wire size, contact area, and heat rise before choosing a repair.

A damaged pad under a surface-mount part may need a small pad repair or a connection to another exposed point on the same net. The repair should be inspected under magnification and tested for both continuity and unwanted contact with nearby nets.

9. The board fails after moisture or contamination

Dust, flux residue, metal particles, and moisture can create leakage paths between conductors. This is common around connectors, high-impedance sensor inputs, and fine-pitch components.

I remove power and inspect the board for:

• White or green corrosion
• Sticky residue
• Water marks
• Darkened copper
• Dirt near fine-pitch pins
• Damaged conformal coating

Cleaning should match the board materials and component limits. After cleaning, I allow the board to dry fully before testing. Corrosion can continue under a component, so a visual cleaning may not repair the electrical damage.

If the board is used in a humid or dusty environment, the design may need better enclosure protection, drainage, connector selection, or coating. A repair can restore one board, but the surrounding conditions may still create repeat failures.

10. The fault comes from the design, not the assembly

A board can be assembled correctly and still fail because the circuit does not meet its operating conditions.

I review:

• Maximum current
• Voltage tolerance
• Startup sequence
• Connector ratings
• Trace width
• Ground return paths
• Decoupling capacitor placement
• Electromagnetic noise sources
• Clearance around high-voltage sections

For example, a motor driver may reset a microcontroller when the motor starts. The cause may be a temporary supply drop or electrical noise rather than a software defect. Measuring the supply rail during motor startup can reveal a short voltage dip that is invisible during idle testing.

When I find a design issue, I record the measured condition, the expected condition, and the proposed change. That record helps the design and production teams work from the same information.

A practical PCB troubleshooting routine

I use this order when a board arrives for repair:

• Record the board version and visible condition.
• Ask what changed before the failure appeared.
• Disconnect external loads.
• Inspect the board under good lighting.
• Check resistance between power rails and ground.
• Apply power with current control.
• Measure each supply rail.
• Test reset, clock, and communication signals.
• Inspect suspected joints and traces.
• Repair one issue at a time.
• Test the board under the same conditions that caused the fault.

Random part replacement can hide useful evidence. A measured approach gives me a better chance of finding the original cause.

PCB troubleshooting becomes easier when I separate symptoms from causes. A dark LED may point to a bad LED, but it may also indicate a missing supply rail. A hot chip may be damaged, or it may be reacting to a shorted load. Clear measurements help distinguish these cases.

Good board design also reduces repair work. Clear polarity marks, accessible test points, suitable protection parts, strong connector support, and documented revisions give technicians useful information before they remove a single component.

Interested in learning more about industry trends and solutions? Contact lingchao: mr.xu@lingchaopcb.com/WhatsApp +8613780181891.

1. IPC — 2022 — IPC-A-610H Acceptability of Electronic Assemblies

2. IPC — 2022 — J-STD-001H Requirements for Soldered Electrical and Electronic Assemblies

3. Paul Scherz and Simon Monk — 2016 — Practical Electronics for Inventors

4. Eric Bogatin — 2018 — Signal and Power Integrity Simplified

5. Henry W Ott — 2009 — Electromagnetic Compatibility Engineering

6. John H Davies — 2017 — The Printed Circuit Designer’s Guide to Fundamentals of PCB Design