Thermal Diagnosis of Shorts and Leakage

Section Overview

An ohmmeter says a rail is shorted but not where, and this closing thermal section solves that by forcing a current through the rail so the fault warms and a camera finds it (thermal-and-injection-methods-for-shorts). A soft short can be as important as a hard one. Leakage is current flowing where it should not, through a failing dielectric or a degraded junction, so a leaky part passes current it should not and warms as it does, a fault a meter may see only as a rail slightly low. Current is forced through the fault to develop heat. Driving a constant current into the shorted rail from a limited source gives a low-resistance fault a known current up to the supply's set voltage, so it dissipates a controlled, safe amount of heat exactly where it lives (heat-as-a-diagnostic-signal). The heat is then read to the one part. Thermal short-localization is finding the fault by that injected current — the shorted or leaky component warms, and a thermal camera or a careful finger pinpoints it, turning a whole-rail short into a single part (thermal-imaging-and-camera-technique). One limit is respected honestly. A near-zero-ohm dead short drops almost no voltage, so it dissipates little at itself and heats the current path instead, meaning the method finds a resistive short or leak best and a trace on a hard short (thermal-and-injection-methods-for-shorts). Follow the heat that current makes, and a short that was somewhere on a whole rail becomes a single warm part you can point to.

Why This Matters

A short an ohmmeter can only place on a whole rail is localised in seconds by the heat an injected current makes, so thermal methods turn a shorted net into a single guilty part (thermal-and-injection-methods-for-shorts). This matters because heat pinpoints where a meter cannot: an ohmmeter reads the same few ohms everywhere on a shorted net, but the injected current warms only the fault, so heat gives the location a resistance reading cannot (heat-as-a-diagnostic-signal). This matters because leakage hides from a meter: a soft partial short passes current without collapsing the rail, so it barely shows on an ohmmeter yet warms plainly under injection, making thermal the way to find it. It matters because injection is controllable: a constant current develops a known, safe heat at the fault, so the fault is provoked without damaging the board. It matters because the method has an honest limit: a perfect dead short dissipates little at itself, so knowing it heats the trace instead keeps you from chasing the wrong hot spot (thermal-and-injection-methods-for-shorts). And it matters because thermal and electrical agree: the injected current's voltage gives the fault's resistance while its heat gives the location, so the two pictures together confirm the fault (finding-the-overheating-component). Force current through the short and read the heat, and the guilty component names itself.

Required Prerequisites

• Heat as a Diagnostic Signal — Section 6.1 established that dissipated power becomes heat; this section forces current through a short so that heat appears at the fault and localises it.
• Thermal and Injection Methods for Shorts — Section 4.4 introduced injecting current and reading heat to localise a short; this section deepens it with thermal-camera technique and the leakage case.

Recommended Consumables

• A current-limited bench supply — to inject a controlled current into the shorted rail (thermal-and-injection-methods-for-shorts)
• A thermal camera with a tuned span — to find the component that warms under injection (thermal-imaging-and-camera-technique)
• Fine injection leads — to drive current in at the right point on the rail
• Isopropyl alcohol — to wet the rail and watch where it dries first, a low-tech heat finder
• A notebook for the readings — to log the injected current, the voltage, and the part that warmed

Recommended Practice Hardware

• A board with a shorted rail — to inject current and find the warming part (thermal-and-injection-methods-for-shorts)
• A board with a leaky capacitor — to see leakage warm a part a meter barely flags
• A current-limited supply and a thermal camera — to practise injection and reading the heat (thermal-imaging-and-camera-technique)
• A board with a near-zero-ohm dead short — to see the heat appear on the trace, not the short
• A multimeter — to read the fault's resistance from the injected voltage and current
• A schematic of the rail — to know which components sit on the shorted net

Real-World Applications

Thermal localization is how a technician turns a shorted rail into one warm part. A repairer with a rail shorted to ground injects a controlled current and sweeps a thermal camera until one capacitor glows, localising a short an ohmmeter placed only on the whole net (thermal-and-injection-methods-for-shorts). A technician chasing a rail that is slightly low injects current and finds a leaky part warming, catching leakage a meter barely showed (heat-as-a-diagnostic-signal). Someone localising a short on a dense board watches which part warms first and fastest, pinning the fault among forty candidates. A repairer facing a hard dead short sees the trace warm rather than a part and follows the warming copper toward the short, using the method's limit to advantage (thermal-and-injection-methods-for-shorts). And a technician confirming a suspect reads the injected voltage for the fault's resistance and the heat for its location, agreeing the two before replacing it. The failures this prevents: lifting forty parts to find one short, missing leakage that a meter dismissed, and chasing a hot trace as if it were the shorted component.

Common Challenges

• A dead short barely warms at itself. A near-zero-ohm short drops almost no voltage, so it dissipates little at the fault and the heat shows on the trace instead — this is physics, not a failure, and the warming copper still leads toward the short (thermal-and-injection-methods-for-shorts).
• The heat can spread along the copper. Injected current warms the whole path, and the plane conducts heat, so more than the fault can look warm — reading the hottest point and the earliest to warm still finds the source, but the spread must be allowed for.
• Too little current shows nothing, too much damages. A fault needs enough current to warm but not so much it harms the board — finding the current that provokes without damage is a judgement each board demands (heat-as-a-diagnostic-signal).

Safety Notes

Risk Level: Medium. Injecting current to warm a fault means driving real current into a board and reading parts that can get hot, so it carries burn and electrical hazards, and this section is Medium risk.

Professional Tips Before Starting

• Inject a limited current and raise it slowly. Too much current damages, too little shows nothing — start low and raise until the fault warms (thermal-and-injection-methods-for-shorts).
• Read the heat map, not one spot. The fault is the earliest and hottest to warm — watch the whole rail warm and find where it starts (thermal-imaging-and-camera-technique).
• Expect a dead short to heat the trace. A near-zero-ohm short barely warms itself — follow the warming copper toward it instead (heat-as-a-diagnostic-signal).

Finding a Short or Leak by the Heat It Makes

Recap and Frame

Section 4.4 injected current and read heat to localise a short; this closing thermal section deepens that with camera technique and the leakage case, and the frame is that current forced through a fault makes heat where the fault lives (thermal-and-injection-methods-for-shorts). Power at the fault becomes heat. An injected current times the voltage it develops across the fault is power dissipated at the fault, so the shorted or leaky part warms exactly where it is (heat-as-a-diagnostic-signal). Leakage is a fault heat finds well. A soft partial short — a failing dielectric, a leaky junction — passes current a meter barely flags yet dissipates heat under injection, so thermal reaches faults resistance struggles with. A constant current keeps it controlled. Driving a set current from a limited source gives the fault a known current and a safe, predictable heat, so the injection provokes without running away (thermal-and-injection-methods-for-shorts). The camera turns heat into a location. Reading the heat map for the earliest and hottest part localises the fault, so a whole-rail short becomes a single warm component (thermal-imaging-and-camera-technique). One honest limit is held. A near-zero-ohm dead short dissipates little at itself and heats the trace, so the method finds a resistive short or leak best and a trace on a hard short. Hold the frame — current makes heat at the fault, leakage warms, a constant current keeps it safe, and the camera localises — and a shorted rail gives up its guilty part.

Why a Short or Leak Makes Heat

The method rests on a simple fact: a fault carrying current dissipates power, and that power becomes heat at the fault, so pushing current through a short warms it (heat-as-a-diagnostic-signal). Know that a resistive short heats. A short with some resistance drops a voltage as current flows, so it dissipates that current times that voltage as heat, and a resistive short warms in proportion to the current forced through it (thermal-and-injection-methods-for-shorts). Know that leakage heats. A leaky part passes current through a resistance it should not have, so it too dissipates power and warms, which is why leakage a meter dismisses shows under injection. Understand the dead-short exception. A perfect near-zero-ohm short drops almost no voltage, so its power is tiny and it barely warms at itself — the current still flows, but the heat appears where there is resistance, along the trace and connectors carrying it (thermal-and-injection-methods-for-shorts). See the heat scale with current. Because heat is current times voltage, more injected current makes more heat at the fault, so raising the current makes a faint fault warm enough to find, within safe limits. Tie heat to the current path. The current flows from the injection point through the fault, so the whole path can warm, and the fault is the point where the heat concentrates most, the earliest and hottest. Relate it to over-dissipation. A short or leak warming under injection is the same over-dissipation seen elsewhere, current times voltage as heat, applied deliberately to provoke a fault (finding-the-overheating-component). The resistive short, the leakage, the dead-short exception, the scaling with current, the current path, and the tie to over-dissipation understood — and why a short or leak makes heat is clear. Force current through a fault with resistance, and it warms where it lives.

Injecting Current to Develop the Heat

Heat is developed at the fault by injecting a controlled current into the shorted rail, so the injection is set up to provoke the fault safely and reveal it (thermal-and-injection-methods-for-shorts). Confirm the short first. The rail is confirmed genuinely shorted with a resistance reading before injecting, so current is driven into a real fault and not a good rail (thermal-and-injection-methods-for-shorts). Discharge and unpower the board. The board is unpowered and its bulk capacitors discharged before injecting, so an external current is driven into a safe board and not fighting its own supply. Use a constant, limited current. A current-limited bench supply drives a constant current into the rail, giving a low-resistance fault a known current up to the supply's set voltage and a controlled heat that cannot run away, the safe way to inject. Start low and raise slowly. The current is started low and raised only until the fault warms enough to find, so a trace is not overheated and a part is not lifted by too much current (heat-as-a-diagnostic-signal). Inject across the short. The current is injected so it flows through the fault — from the rail to its return — and not through sensitive parts that a wrong injection point would stress, so the leads are placed to drive the fault. Read the voltage as you go. The voltage the constant current develops across the rail gives the fault's resistance, so the injection measures the fault electrically while it heats it, tying the two methods together. The short confirmed, the board discharged, a constant limited current used, started low, injected across the fault, and the voltage read — and current is injected to develop the heat. Drive a controlled current into the confirmed short, and the fault begins to warm.

Finding the Warming Component with the Camera

With current flowing, the fault warms, and a thermal camera or a careful touch reads the heat map to the one component that heats — the localization itself (thermal-imaging-and-camera-technique). Sweep the rail with the camera. A thermal camera swept over the injected rail shows which part warms, so the fault is spotted as the hot point among the cool components of the net (heat-as-a-diagnostic-signal). Find the earliest and hottest. The fault warms first and most, so watching which part heats earliest as the current rises, and which is hottest, points at the shorted or leaky component rather than a warmed neighbour. Tune the span to a small rise. Narrowing the temperature span makes a faintly warming part bloom, so a leak that warms only slightly under a safe current is still seen (thermal-imaging-and-camera-technique). Use touch or alcohol where no camera is at hand. The back of a finger, briefly, or a film of alcohol that dries first at the warm spot, finds the heating part without a camera, so the method works with simple tools too. Separate the fault from the warmed path. The current warms the whole path, so the fault is told from the merely warmed trace by being the concentrated peak, and by cooling suspects to see which drops the heat. Mind the dead-short trace. If no part warms but a trace does, the short is a near-zero-ohm one heating its copper, so the warming trace is followed toward the short instead of hunting a hot part (thermal-and-injection-methods-for-shorts). The rail swept, the earliest and hottest found, the span tuned, touch and alcohol used, the fault told from the path, and the dead-short trace minded — and the warming component is found. Read the heat to its concentrated peak, and the short or leak names its part.

Distinguishing Leakage from a Hard Short

Not every low rail is a hard short, and telling leakage and a resistive short from a dead short changes what the thermal method shows and how it is read (heat-as-a-diagnostic-signal). Read the resistance to classify. The rail's resistance sorts the fault — a near-zero reading is a hard short, a few ohms to hundreds is a resistive short or a strong leak, and a high but not open reading is leakage — so the ohmmeter frames what to expect thermally (low-ohms-and-voltage-drop-short-localization). Expect a resistive short to warm well. A resistive short has enough voltage across it under injection to dissipate real heat, so it warms clearly and localises well thermally, the method's best case. Expect leakage to warm faintly. A leak's higher resistance would need more than the supply's set voltage to reach the full injected current, so it passes only a smaller current and warms less, and finding it needs a tuned span and patience, but it still concentrates heat at the leaky part (thermal-imaging-and-camera-technique). Expect a dead short to heat the trace. A near-zero-ohm short develops almost no voltage, so it barely warms itself and the heat is on the trace, meaning a hard short is localised by following the copper, not by a hot part. Use the injected voltage as the tell. The voltage the constant current develops reveals the class directly — near zero for a dead short, a clear drop for a resistive one — so the electrical reading predicts the thermal picture. Match the method to the fault. A resistive short or leak is chased by the part that warms, a dead short by the trace that warms, so recognising the class first aims the search correctly (thermal-and-injection-methods-for-shorts). The resistance classified, a resistive short warming well, leakage warming faintly, a dead short heating the trace, the injected voltage as tell, and the method matched — and leakage is told from a hard short. Classify the fault by its resistance, and the thermal picture you expect follows.

Confirming and Isolating the Fault

A warming part is a strong suspect, and confirming and isolating it before replacement is what turns a thermal hit into a proven repair (finding-the-overheating-component). Confirm the part is the peak. The warming component is confirmed as the concentrated hottest point, not a warmed neighbour, by reading the peak and cooling suspects to see which drops the heat (heat-as-a-diagnostic-signal). Cross-check with resistance. An unpowered resistance reading on the suspect part — a low-ohms reading to ground, a leaky junction — confirms the thermal hit electrically, so heat and resistance agree on the fault (low-ohms-and-voltage-drop-short-localization). Isolate by lifting or removing. Lifting a leg or removing the suspect and seeing the short or leak clear on the rail proves it was the fault, so isolation confirms the culprit rather than a coincidence (isolating-the-shorted-component). Beware more than one fault. A rail can have more than one leaky or shorted part, so after removing one, the rail is re-checked and re-injected in case another remains, not assumed clear. Trace leakage to its cause. A confirmed leak is understood — a failed dielectric, a stressed junction, moisture or contamination — so the repair addresses why it leaked, cleaning contamination as readily as replacing a part (thermal-and-injection-methods-for-shorts). Re-inject to confirm the fix. After the repair, the rail is re-injected and read to confirm no part warms and the short or leak is gone, closing the loop from a shorted rail to a clear one. The peak confirmed, cross-checked with resistance, isolated by lifting, a second fault watched for, leakage traced, and re-injected to confirm — and the fault is confirmed and isolated. Confirm the warm part electrically and isolate it, and the short or leak is proven and cleared.

Common Mistakes

• Injecting too much current. Too much current overheats traces and lifts parts — start low and raise only until the fault warms (thermal-and-injection-methods-for-shorts).
• Chasing a hot trace as the fault. A dead short heats its trace, not itself — follow the warming copper toward the short rather than replacing the trace (heat-as-a-diagnostic-signal).
• Missing faint leakage. A leak warms only slightly — tune the span and be patient rather than concluding the rail is clear.
• Replacing on heat alone. A warm part is a suspect, not proof — cross-check with resistance and isolate before replacing (isolating-the-shorted-component).
• Assuming one fault. A rail can have several leaky or shorted parts — re-inject after each removal to catch another (low-ohms-and-voltage-drop-short-localization).

Troubleshooting Guidance

Thermal short work comes down to confirm, inject, read, classify, and isolate. If a rail is shorted but you cannot place it: inject a limited constant current and sweep a thermal camera for the part that warms (thermal-and-injection-methods-for-shorts). If nothing warms but a trace does: the short is a near-zero-ohm one heating its copper — follow the warming trace toward it (heat-as-a-diagnostic-signal). If a rail is only slightly low: suspect leakage and inject current, watching for a part that warms faintly under a tuned span (thermal-imaging-and-camera-technique). If too much of the rail looks warm: lower the current and read the earliest and hottest point, and cool suspects to find the source. If you are unsure of the fault class: read the resistance and the injected voltage — near zero is a dead short, a clear drop is resistive or a strong leak. If a suspect warms: cross-check its resistance and lift or remove it to confirm the short or leak clears (isolating-the-shorted-component). If the rail is still shorted after a repair: re-inject — another leaky or shorted part may remain. The throughline: confirm the short, inject a safe constant current, read the heat to the peak, classify by resistance, and isolate the guilty part.

Verification & Testing Methods

Confirm you localised the short or leak by its heat:

• [ ] I confirmed the rail was shorted, discharged the board, and injected a constant current from a current-limited supply, started low and raised slowly (thermal-and-injection-methods-for-shorts).
• [ ] I practised thermal short-localization — sweeping the injected rail and reading the heat to the earliest and hottest component rather than a warmed neighbour.
• [ ] I recognised leakage — a part warming faintly under injection where a meter saw only a slightly low rail — and did not dismiss it as clear (thermal-imaging-and-camera-technique).
• [ ] I distinguished a resistive short or leak, which warmed at the part, from a near-zero-ohm dead short, which warmed the trace instead (heat-as-a-diagnostic-signal).
• [ ] I cross-checked the warm part with a resistance reading and isolated it by lifting or removing it, then re-injected to confirm the fault was gone (isolating-the-shorted-component).

Then try the practice exercises below — thermal short-localization practice on unpowered boards with injected current; scenarios differ from the quiz.

Practice Exercises

1. Inject and localise a short (5 minutes, hands-on). On a board with a shorted rail, confirm the short, discharge the board, inject a low constant current, and sweep a thermal camera for the part that warms, raising the current slowly (thermal-and-injection-methods-for-shorts).
2. Find faint leakage (5 minutes, hands-on). On a board with a slightly low rail, inject current and tune the span narrow to catch a part that warms only faintly (thermal-imaging-and-camera-technique).
3. Recognise a dead short (5 minutes, hands-on). On a board with a near-zero-ohm short, inject current and observe the trace warming rather than a part, and follow the copper toward the short.
4. Confirm and isolate (3 minutes, reasoning). For a warm suspect, state how you would cross-check its resistance, lift or remove it, and re-inject to confirm the fault cleared.

These core steps — why a short or leak makes heat, injecting current, finding the warming part, distinguishing leakage from a hard short, and confirming and isolating — are tested in the Chapter Quiz at the end of this chapter, where a score of 80% is required to continue.

Key Takeaways

• Forcing a current through a shorted rail makes the fault dissipate power and warm where it lives, so a thermal camera turns a whole-rail short into a single warm component to point to (thermal-and-injection-methods-for-shorts).
• Leakage — a soft partial short through a failing dielectric or a degraded junction — passes current a meter barely flags yet warms plainly under injection, so thermal reaches faults resistance struggles with.
• Driving a constant current from a limited source gives a low-resistance fault a known current up to the supply's set voltage, so it develops a controlled, safe heat and the injection cannot run away — while a higher-resistance leak passes only the smaller current that voltage allows, and so warms faintly.
• Thermal short-localization reads the injected rail's heat to the earliest and hottest part, but a near-zero-ohm dead short dissipates little at itself and heats the trace, so a hard short is followed along the warming copper instead (heat-as-a-diagnostic-signal).
• A warm part is a suspect, not proof, so it is cross-checked with a resistance reading and isolated by lifting or removal, and the rail is re-injected after the repair to catch a second fault (isolating-the-shorted-component).

Skills Learned

• You can now explain why a short or leak develops heat where it lives.
• You can now inject a controlled current to develop heat at a fault.
• You can now find the warming component with a thermal camera or touch.
• You can now tell leakage and a resistive short from a hard dead short.
• You can now confirm the guilty component before replacing it.

Glossary Additions

• leakage — current flowing where it should not, through imperfect insulation or a degraded semiconductor junction, rather than through a hard metallic short. Leakage is a soft or partial short: a path that passes some current but not the near-infinite current of a dead short, so it dissipates power as heat at the leaky part without collapsing the rail to zero. Because a leaky component develops heat under the current it passes, leakage is often best found thermally — injecting current into the affected rail and watching for the part that warms — where a meter may only show a rail that is low or slightly loaded rather than dead shorted. A capacitor with a failing dielectric, a semiconductor with rising junction leakage, and moisture or contamination across a board are all common sources of leakage.
• thermal short-localization — the technique of finding a short or leak by injecting current into the affected rail so the fault dissipates power and warms, then using a thermal camera or a careful touch to find the component that heats. It turns a fault a meter can only place on a whole rail into a pinpoint, because the current forced through the fault develops heat exactly at the shorted or leaky part. Its strength is on resistive shorts and leakage, which have enough resistance to dissipate heat at the fault; a near-zero-ohm dead short dissipates little at itself, so the heat appears in the current path rather than at the short, and the method then finds the trace carrying the current instead. Driving a controlled current and watching the heat map is what makes thermal short-localization a fast route from a shorted rail to the guilty component.
• constant current — a controlled current driven into a rail from a current-limited source, held at a set value up to the supply's compliance (set) voltage, used to develop a predictable and safe amount of heat at a short or leak. Injecting a constant current, rather than a voltage, means a low-resistance fault such as a milliohm dead short or a resistive short gets a known current, so the heat it develops is controlled and the injection cannot run away; a higher-resistance leak that would need more than the set voltage to reach that current simply passes the smaller current its resistance allows, which is why a leak warms only faintly. Setting the current high enough to warm the fault but low enough not to damage the board or the trace carrying it is the balance that makes injection safe, and reading the voltage the current develops across the rail also gives the fault's resistance, tying the thermal method to the electrical one.

Suggested Next Sections

Must read next:

• The Signal Chain as a Diagnostic Path — Chapter 7 opens a new diagnostic approach for boards whose power is good but whose signals are wrong: injecting a known signal at one point and tracing it through a chain to find where it is lost, distorted, or stops — following a signal the way this volume has followed power and heat.

Recommended:

• Heat as a Diagnostic Signal — the principle that dissipated power becomes heat, which injection uses to make a short reveal itself.
• Thermal and Injection Methods for Shorts — the Chapter 4 introduction to injecting current and reading heat that this section extends with camera technique and leakage.