Common Faults and Symptoms
Every technique in this module — systematic reasoning, half-split testing, injection, tracing, substitution, in-circuit measurement and scope work — needs a starting hypothesis to work from. This lesson is that starting point: a practical, symptom-organized catalog of the failure patterns met most often in power supplies, amplifiers, microcontroller projects and everyday electronic equipment, ranked by likelihood within each category so you know where to look first.
Use this map to jump directly to the symptom category that matches what you are observing.
View LargerNo Power At All
A completely dead unit — no LEDs, no display, no fan, no click — is statistically one of the easiest faults to find, because the list of things that can interrupt power entirely is short and mostly involves simple mechanical or protective parts.
| Likely cause (most probable first) | Quick test |
|---|---|
| Power outlet, adapter, cable or connector fault | Measure the voltage at the device end of the cable while plugged in |
| Blown fuse (external or internal) | Visual inspection and continuity check; find out why it blew before replacing it |
| Reversed polarity on a DC input (some equipment protects itself by not powering on) | Check the connector polarity against the device's marking before reapplying power |
| Power switch or button failure | Continuity check across the switch contacts |
| Failed reverse-polarity protection diode or input MOSFET (often fails after a polarity mistake) | In-circuit voltage check just after the protection device |
| Failed internal regulator | Voltage check on each internal supply rail |
Power Supply Faults
Power supply problems cause more equipment faults than anything else, and often masquerade as faults elsewhere. Always check the supply rails — voltage, ripple and behavior under load — early in any diagnosis.
| Likely cause (most probable first) | Quick test |
|---|---|
| Dried-out or bulging electrolytic capacitors (high ripple, hum, unstable output) | Visual check for bulging tops; measure ripple with a scope; check ESR with an ESR meter |
| Output correct with no load but sagging under load | Measure under load; check the regulator, pass transistor, and current-limit circuit |
| Switching supply clicking, ticking or restarting (hiccup mode) | Disconnect the load: if it then runs, look for a short or excessive load; if not, check the output diodes and capacitors |
| Failed rectifier diode (one half-cycle missing, high ripple at mains frequency instead of double) | Scope the ripple: 50/60 Hz ripple on a full-wave supply means a rectifier fault |
| Failed feedback component (output too high or wildly wrong) | Check the feedback divider resistors and reference; switch off immediately if the output is too high |
| Shorted output component pulling the rail down | Current-limited bench supply on the rail and a thermal camera or finger test for the hot part |
Powers On But No Output
When the equipment lights up and the controls respond but the main function does nothing — no sound from an amplifier, no movement from a motor driver, no reading from a sensor — the fault is in the signal or output path, or in a control function holding it off.
| Likely cause (most probable first) | Quick test |
|---|---|
| Mute, standby, enable or protection mode active | Check indicators and the enable or mute line; look for a protection LED |
| Input, output or speaker cable fault, or wrong input selected | Substitute known-good cables; check the input selection |
| Headphone or output jack switch stuck (cuts the main output) | Insert and remove a plug a few times; check the switch contacts |
| Failed output stage or blown output fuse | Signal injection backward from the output; check output fuses |
| Protection circuit triggered by a DC offset or overload at the output | Measure DC at the output with no signal; disconnect the load and retry |
| Missing supply rail to one section | Voltage check on each rail at the dead section |
Weak or Distorted Output
Output that is present but clearly wrong — too weak, distorted, or noisy — points toward a degraded stage rather than a completely failed one, since a complete failure usually produces nothing at all.
| Likely cause (most probable first) | Quick test |
|---|---|
| Gain or level set too high, causing clipping | Reduce the input level and observe the output on a scope |
| Stage bias drift (resistor drift, aging transistor) | In-circuit voltage check against a healthy channel or service data |
| Supply voltage sagging under load | Measure the rails while the output is at full level |
| Dried-out coupling or decoupling capacitor | Signal tracing; check the capacitor out of circuit or by substitution |
| Damaged speaker, motor or other load | Substitute a known-good load or a load resistor |
| Oscillation riding on the signal | Scope the output on a fast timebase |
Microcontroller and Digital Faults
In projects and modern equipment, many faults are in the digital side. They are often caused by power or wiring problems rather than by the microcontroller itself.
| Likely cause (most probable first) | Quick test |
|---|---|
| Brown-out resets: supply dips when a motor, relay, LED strip or radio module switches on | Scope the supply rail at the microcontroller while the load switches; add bulk capacitance and separate the load supply |
| Missing or poor decoupling capacitors | Check for a 100 nF capacitor at every supply pin, close to the pin |
| Floating input pins causing random behavior | Check that unused inputs and buttons have pull-up or pull-down resistors |
| I²C or SPI bus hangs (no pull-ups, wrong address, bus too long) | Logic analyzer on the bus; check pull-up values and device addresses |
| Will not program or boot (wrong boot pin state, crystal not starting, reset held low) | Check reset and boot pins with a meter; scope the crystal pins with a 10x probe |
| Output pin overloaded or damaged by driving a load directly | Measure the pin voltage with the load disconnected; use a transistor or driver for loads |
| Firmware fault | Load a known-good test program, such as a blinking LED, to separate hardware from software |
Intermittent Faults
Faults that come and go are the most frustrating category, because a single static test often shows nothing wrong. The common thread across most intermittent faults is a marginal mechanical connection — something that is “almost” making good contact — or a component on the edge of failing.
| Likely cause (most probable first) | Quick test |
|---|---|
| Cold or cracked solder joint | Visual inspection under magnification; gentle wiggle test while monitoring the symptom |
| Marginal connector or cable | Substitution with a known-good cable |
| Cracked PCB trace | Continuity check while gently flexing the board |
| Thermally dependent component failure | Gentle controlled heat (a hair dryer) or cold spray on a suspect area while monitoring the symptom |
| Supply rail on the edge of dropping out | Monitor the rail on a scope in normal (triggered) mode set to catch a dip |
| Oscillation that depends on hand or cable position | Observe whether the fault changes when you move your hand near the circuit without touching it |
Overheating and Burning Smells
Heat is a symptom with a cause. Find the hot part first — with a thermal camera, an infrared thermometer, or carefully, with the power off and capacitors discharged, by touch — then work out why it is hot.
| Likely cause (most probable first) | Quick test |
|---|---|
| Shorted capacitor or semiconductor loading a rail | Current-limited bench supply on the rail; find the hot spot |
| Excessive idle current in a power amplifier stage | Measure the current through the emitter or source resistors; check the bias network |
| Parasitic oscillation | Scope the output and supply rails on a fast timebase |
| Blocked ventilation, failed fan, or dried thermal paste | Inspect the airflow and fan; check the heat sink contact |
| Load drawing more than the design allows | Measure the load current and compare with the rating |
Oscillator, Clock and Timing Faults
A frequency, clock or timer that is wrong or wanders over time usually points toward the frequency-determining components or their supply, rather than a simple failure.
| Likely cause (most probable first) | Quick test |
|---|---|
| Normal oscillator warm-up drift (within specification) | Compare the drift with the manufacturer's specification |
| Wrong crystal load capacitors, so the crystal runs off frequency or fails to start | Check the capacitor values against the crystal's load capacitance |
| Loose or marginal crystal, trimmer or inductor | Gentle wiggle test while monitoring the frequency |
| Real-time clock gaining or losing time | Check the crystal and load capacitors; check the backup battery |
| Supply noise modulating the oscillator | Check the supply rail for ripple and noise with a scope |
Noise, Hum and Interference
Noise complaints split into two categories that are easy to confuse: noise generated inside the equipment or its connections (often solvable with grounding, filtering and layout, as covered in the Interference and Noise module) and noise arriving from outside sources.
| Likely cause (most probable first) | Quick test |
|---|---|
| Ground loop hum between connected equipment | Disconnect cables one at a time; try a ground-loop isolator in the signal path |
| Switching power supply or charger noise | Run from a battery or linear supply and compare |
| Failed filter capacitor (hum at twice the mains frequency) | Scope the supply ripple |
| Common-mode noise on cables | Clip ferrites onto cables near the equipment and observe the change |
| RF pickup in audio or sensor inputs (buzz from phones, radio breakthrough) | Move a phone near the equipment; add RC filters and ferrites at the inputs |
| External source (power line arcing, a neighbor's equipment) | Use a battery-powered portable radio to locate the source |
Display, Control and Settings Faults
Modern equipment is part electronics and part embedded computer, and faults on the control side can mimic or hide faults elsewhere. A frozen display, a stuck control, or settings that do not stick are common and usually have simple causes.
| Likely cause (most probable first) | Quick test |
|---|---|
| Processor lockup (often cleared by a power cycle or documented reset procedure) | Power cycle; consult the manual for a reset procedure |
| Stuck button, encoder, or sticky keypad contact | Visual inspection; operate each control individually and observe the response |
| Corrupted memory or settings that reset unexpectedly (often a failing backup battery) | Reload defaults; check the backup battery or supercapacitor |
| Backlight failure (separate from the display itself) | Shine a torch on the display to see whether the segments or image are still there |
| Firmware bug (a known issue with a specific version) | Check the manufacturer's support pages for an update |
Frequently Asked Questions
My symptom doesn't fit cleanly into one category. What should I do?
Many real faults span two categories — for example, distorted output that only appears at high volume may be a weak-output problem or a power supply problem. Read the likely causes in every category that fits and start with the cheapest, fastest tests. Power supply checks are almost always worth doing first, because supply problems cause symptoms in every other category.
Is this catalog a substitute for learning the techniques earlier in this module?
No — this catalog tells you where to start looking; the earlier lessons (systematic method, divide and conquer, injection, tracing, substitution, in-circuit measurement and scope use) tell you how to confirm the cause once you have a short list of suspects. The catalog speeds up the start of the process; the techniques finish it.
Why do power supply checks appear in so many categories?
Because the supply feeds everything. A sagging, noisy or dropping rail can produce no output, weak output, distortion, resets, intermittent faults and noise — and electrolytic capacitors in supplies are among the components most likely to age and fail. Checking the rails takes a minute with a meter and a scope, and it either finds the fault or eliminates a large part of the equipment from suspicion.
Should I always work through the causes in the exact order listed?
The order reflects general probability across many kinds of equipment, but your own context can override it — recent history, a known weak point in a particular product, or a recent event (a storm, a drop, a new accessory, a firmware update) should move a cause to the top of the list.
Test Your Knowledge
Answer the questions below to check your understanding of this lesson. Every answer can be found in the lesson above.