Thermal Cameras — Hotspot Detection in Repairs

Section Overview

Some faults are invisible to the meter's where — it tells you a board draws too much current, but not which of a hundred components is the short. A thermal camera answers that question directly: it images infrared radiation (heat) and shows the board's temperature distribution as a false-color image — a heat map where hot areas glow bright (red or white) and cool areas are dark (blue) — so you see the temperature of every part at once, without touching anything. The reason this is powerful in repair is simple physics: a shorted, overloaded, or failing component dissipates extra power and gets hot, so the camera pinpoints exactly which part is the fault — the one glowing hot in the image. That's the classic move: you measured excessive current draw with the meter (Chapter 6), you power the board, and the thermal camera reveals the single hot part — a shorted capacitor, a failing regulator, a shorted semiconductor. You'll learn the core technique: power the board carefully (see the safety note) and hunt the abnormal hotspot — the outlier that's hotter than its neighbors, than it should be, or than a known-good board — because a dead short heats a part within seconds, and because a stage that's cold when it should be warm means no power is reaching it. It's the relative hotspot (the outlier) that fingers the fault, more than the absolute temperature. You'll learn the crucial emissivity caveat — shiny/metallic surfaces read falsely cool, so matte them for a true reading — plus resolution (for small SMD parts), and that affordable smartphone-attachment thermal cameras make this accessible. And you'll learn to do it safely, because the camera is safe — the hazard is the powered, faulty board.

Why This Matters

Finding the culprit is often the hardest part of a repair, and thermal imaging is one of the fastest ways to do it for a whole class of faults. Consider the most common hard case: a board is dead or protecting, and the meter shows it's drawing far too much current — there's a short somewhere. Without a thermal camera you're in for a slow hunt: injecting a small current and measuring tiny voltage drops, removing components one at a time, or reasoning through the schematic. With a thermal camera, you power the board (carefully, current-limited) and look — and the shorted part heats up and lights up, often within seconds, telling you exactly where to work. Minutes instead of hours. And it's non-contact: you're not probing fragile traces or bridging pins, just looking at a heat map. Beyond shorts, it reveals a component running hotter than it should (a stressed regulator, an overloaded resistor) before it fails, and — read the other way — it shows which stages are warm (working, drawing power) versus cold (dead, or not receiving power), a quick way to localize where a signal or supply stops. The capability used to be expensive, but smartphone-attachment thermal cameras have made it affordable enough for any serious bench. The judgment this section builds is knowing what the heat map is telling you — that the outlier is the suspect, that a shiny part can lie about its temperature, and that a cold stage is as informative as a hot one — and knowing how to power a possibly-shorting board safely while you look. It turns "there's a short somewhere" into "there's the short."

Required Prerequisites

• Current Measurement — measuring a board's current draw is how you know there's a short or overload (excessive current); the thermal camera is how you find which component is causing it, so the two work together — the meter says a fault exists, the camera says where.

Recommended Consumables

• A little matte paint, masking tape, or thermal-imaging (emissivity) spray to give shiny/reflective parts a true-reading surface
• Nothing else is consumed. A current-limited bench supply (Section 8.5) is strongly recommended for powering a suspected-short board — treat it as essential kit here rather than a consumable

Recommended Practice Hardware

• A thermal camera — a standalone handheld unit, or (very accessible) a smartphone-attachment thermal camera such as a FLIR ONE or Seek Thermal that clips to a phone
• A current-limited bench supply (Section 8.5) to power boards safely, with current-limit set low for a suspected short
• A practice board — ideally one with a known short or overheating part (or a resistor you can deliberately overload) to see a hotspot form; and a known-good board to compare against

Real-World Applications

Thermal imaging is a first reach for a whole family of faults. A technician facing a device that's dead and drawing too much current sets a bench supply to the right voltage with the current limit set low, powers the board, and watches the thermal camera: within seconds a single component glows — the short — and they've localized in one step what could have been an hour of probing. Repairing a supply that runs hot and shuts down, they image it under load and find the one regulator or component running far hotter than its neighbors, the stressed part to replace or investigate. Diagnosing a stage that isn't working, they read the map the other way: the stages that should be warm (drawing power) are cold, showing where the power or signal stops. Chasing an intermittent thermal fault, they watch a part's temperature climb as it's exercised. In every case the workflow is look, don't probe — the heat map points to the part, and only then do they bring the meter or scope to confirm. The failures this prevents are the slow, invasive ones: hour-long component-by-component short hunts, or poking a live board and bridging something. The one discipline that makes it reliable is reading the outlier, not the number — and matting a shiny part that would otherwise lie about being cool. Fast, non-contact, and decisive on exactly the faults that are hardest to localize otherwise.

Common Challenges

• Reading absolute temperature instead of the outlier. The fault is the relative hotspot — the part hotter than its neighbors or a good board — not a specific temperature number; hunt the outlier.
• A shiny part that reads falsely cool. Reflective/metallic surfaces have low emissivity and read too cool, hiding a hot part; matte the surface (paint, tape, spray) for a true reading.
• Powering a shorting board unsafely. Applying full, unlimited power to a suspected short risks damage or a violent failure; power through a current-limited supply and watch briefly.

Safety Notes

Risk Level: Medium. The camera is non-contact and safe — the hazard is that you must power a possibly-faulty board, where a fault can get very hot or worse.

Professional Tips Before Starting

• Power through a current-limited supply. Set the current limit low on a bench supply (Section 8.5) before powering a suspected short — it caps the fault energy and still lets the hotspot show.
• Hunt the outlier, not a number. Look for the part hotter than its neighbors, than expected, or than a good board — the relative hotspot is the fault, not a specific temperature.
• Matte the shiny parts. A reflective/metallic surface reads falsely cool; a dab of matte paint or tape (or emissivity spray) gives a true reading before you trust it.

Finding Hotspots with a Thermal Camera

What a Thermal Camera Shows: A Heat Map of the Board

A thermal camera is a camera that images heat. Every object emits infrared radiation in proportion to its temperature — the hotter it is, the more it emits — and a thermal camera's sensor images that infrared and maps temperature to color, producing a false-color heat map: hot areas rendered bright (typically red, orange, or white), cool areas rendered dark (typically blue or black). Pointed at a circuit board, it shows you the temperature of every part at once, live, without touching anything. That's the whole capability: see, at a glance, what's hot and what's cool across the board. Where a meter gives you a number at one point and your finger gives you a crude, risky sense of "warm," the thermal camera gives you the whole board's thermal picture instantly and safely from a distance. And because heat is where wasted power goes, that picture is often a direct map of where a fault is.

The Key Technique: Power Carefully and Hunt the Abnormal Hotspot

The core repair use is finding the abnormal hotspot. The logic: a healthy component dissipates little power and stays near ambient or mildly warm; a shorted, overloaded, or failing one dissipates extra power and gets hot. So you power the board (carefully — see the safety note) and look for the outlier: the component far hotter than it should be, hotter than its neighbors, or hotter than the same spot on a known-good board. That outlier is your suspect. A dead short is especially dramatic — the shorted part or the trace feeding it often heats visibly within seconds of applying power, lighting up the map and localizing the fault in one step. Read the map the other way too: a stage that's cold when it should be warm (a section that normally draws power sitting at ambient) tells you no power is reaching it — a dead stage or a break upstream — which is just as diagnostic as a hot spot. The key mindset: it's the relative hotspot — the outlier against its surroundings — that fingers the fault, more than the absolute temperature. You're not asking "how many degrees?"; you're asking "which part doesn't belong?"

Emissivity and the Shiny-Surface Caveat

There's one physics caveat you must respect, or the camera will lie to you: emissivity. Emissivity is how well a surface emits infrared — and it varies with the surface. Matte, dark surfaces (most component bodies, plastic packages, dark PCB) emit well and read close to true. But shiny, metallic, reflective surfaces (bare metal cans, polished heatsinks, some connectors) emit poorly — they read too cool, or reflect the temperature of other things — so a genuinely hot shiny part can appear deceptively cool and hide from you. The fix is simple: give the shiny surface a high-emissivity coating — a dab of matte paint, a piece of masking tape, or a shot of thermal-imaging (emissivity) spray — and it will read true. In practice, most components read reasonably well as-is (their dark plastic is fine), so you mainly watch for bare-metal parts. But the rule to carry is: if a suspiciously-implicated part looks cool and it's shiny, don't trust it — matte it and look again.

Resolution, Sensitivity, and Reading Relative vs Absolute

A few camera qualities shape what you can see. Spatial resolution — how many pixels the sensor has — sets how small a part you can resolve: a low-resolution camera shows a blob of heat, while a higher-resolution one can pinpoint which tiny SMD component in a dense cluster is the hot one, so resolution matters most for fine SMD work. Thermal sensitivity is how small a temperature difference the camera can distinguish — useful for spotting a subtle outlier that's only slightly warmer than its surroundings. Temperature range sets the span it can measure. But here's the practical freedom: for fault-finding you're hunting the relative outlier, so you don't need a lab-calibrated absolute temperature — you need to see which part is hottest. That means even a modest, uncalibrated camera is genuinely useful, because "this part is much brighter than everything around it" is exactly the read you want. Resolution (to resolve the part) tends to matter more than absolute accuracy for repair.

Types and Affordable Smartphone Cameras

Thermal cameras come in a few forms. Standalone handheld thermal cameras are self-contained units with their own screen — capable, and the traditional choice. But the accessible revolution for repair is the smartphone-attachment thermal camera — a small module (such as a FLIR ONE or a Seek Thermal) that clips onto a phone and uses the phone's screen and an app — bringing thermal imaging to a modest budget. For most repair work, one of these is entirely sufficient to find the hot part. The main thing that separates a cheap unit from a better one, for electronics, is resolution: a higher-resolution camera resolves smaller, denser SMD parts, which matters on tightly-packed boards, while a low-resolution one may only show a general hot region. So if you work on fine SMD, favor resolution; for general board-level fault-finding, even an entry-level smartphone attachment earns its keep. The capability — see the heat map, find the outlier — is now within easy reach.

When to Use It in Repair

Reach for a thermal camera when the question is "where is the heat?" Specifically: locating a short or excessive-current-draw culprit (power the board and find the part that gets hot — the flagship use); finding an overheating or failing component (the part running hotter than it should, before or as it fails); checking whether a stage is powered (warm = drawing power / working, cold = dead or no power reaching it); and spotting thermal design or stress problems (a part running hot under load). When the question is instead "is this signal clean / what value is on this line?" — that's the scope or logic analyzer (Section 8.1), not the thermal camera. The thermal camera is your tool for the thermal signature of a fault, which — because faults waste power as heat — is a remarkably common and fast way to localize a problem. Look first, then confirm with the meter or scope.

Common Mistakes

• Chasing absolute temperature. The fault is the relative outlier (hotter than neighbors/expected/a good board), not a specific degree reading; hunt the part that doesn't belong.
• Trusting a shiny part's reading. Low-emissivity (shiny/metallic) surfaces read falsely cool; matte them (paint, tape, spray) before believing a hot suspect looks cool.
• Powering a suspected short at full power. Use a current-limited supply with the limit set low to cap the fault energy; don't apply unlimited power to a short.
• Touching the hotspot. The part you're hunting is the hot one — a burn risk; read it non-contact and let it cool (or cut power) before handling.
• Ignoring a cold stage. A section that should be warm but is cold means no power is reaching it — that's a diagnostic clue, not a non-result.

Troubleshooting Guidance

Reading a heat map is interpretation, and a few situations recur. If no obvious hotspot appears: the fault may be low-power (spread out or only mildly warm) — raise the current limit slightly (safely) so the fault dissipates enough to show, give it a few seconds, and check emissivity on any shiny suspect that might be reading falsely cool. If a part you suspect looks cool but it's shiny: that's the emissivity trap — matte the surface (paint, tape, or emissivity spray) and look again; a genuinely hot metal can will now read true. If you're fixated on a temperature number: step back and look for the outlier — the part hotter than its surroundings — which is what actually points to the fault. If a whole stage reads cold that should be warm: that's no power reaching it — trace upstream for the break or dead supply. If nothing gets hot at all and current draw is normal: the fault may not be thermal (a signal or logic problem) — switch to the scope or logic analyzer. And if a part heats alarmingly fast or smokes: cut power immediately — you've found the short, but don't let it cook. The throughline: power carefully and briefly, hunt the outlier (not a number), matte shiny parts, and read a cold stage as "no power here."

Verification & Testing Methods

Use this as a thermal-hunt checklist — confirm these when hunting a hotspot:

• [ ] I power a suspected-short board through a current-limited supply (Section 8.5) with the limit set low, and watch briefly rather than leaving it powered.
• [ ] I hunt the abnormal/relative hotspot — the part hotter than its neighbors, than expected, or than a known-good board — not a specific absolute temperature.
• [ ] I matte any shiny/metallic suspect (paint, tape, or emissivity spray) so low emissivity doesn't make a hot part read falsely cool.
• [ ] I read a cold stage that should be warm as no power reaching it (a diagnostic clue).
• [ ] I do not touch the hotspot — I read it non-contact and let it cool or cut power before handling (burn risk), and I wear eye protection with my face out of the line of the board (a capacitor into a short can vent or burst).
• [ ] I respect any mains/live-circuit hazard on the board (Sections 3.1, 3.2), since I must power it to look.

Then try the practice exercises below — thermal-hunt reasoning; scenarios differ from the quiz.

Practice Exercises

1. Find the short (5 minutes, applied). Describe how you'd locate a short on a board drawing excess current: setting a current-limited supply, powering the board, and identifying the part that heats — and what you'd do the moment it gets hot.
2. The shiny-part trap (5 minutes, reasoning). A metal-can component you suspect reads cool on the camera, yet the board still draws too much current. Explain what emissivity has to do with it and how you'd get a true reading.
3. Read the cold stage (5 minutes, reasoning). A section of the board that should be warm under power is at room temperature on the heat map. Explain what that tells you and where you'd look next.
4. Relative, not absolute (5 minutes, reasoning). Explain why, for fault-finding, "which part is hottest relative to its surroundings" matters more than the exact temperature, and why even a modest camera is useful.

These core ideas — what a thermal camera shows (a heat map), why it finds shorts/overloads (the hot part), the find-the-abnormal-hotspot technique (and the cold-stage read), emissivity and the shiny-surface caveat, and the current-limited/hands-off safety — are tested in the Chapter Quiz at the end of this chapter, where a score of 80% is required to continue.

Key Takeaways

• A thermal camera images infrared radiation (heat) and shows the board's temperature as a false-color heat map — hot bright, cool dark — letting you see every part's temperature at once, non-contact.
• It's powerful for shorts and overloads because a faulty component dissipates extra power and gets hot: the meter tells you a short exists (excess current), the camera tells you which part — often within seconds.
• The technique is to power carefully and hunt the abnormal hotspot — the outlier hotter than its neighbors, expectation, or a good board — and to read a cold stage as no power reaching it. The relative hotspot beats the absolute temperature.
• Emissivity is the key caveat: shiny/metallic surfaces emit IR poorly and read falsely cool, so matte them (paint, tape, spray) for a true reading; most dark component bodies read fine as-is.
• Resolution (to resolve small SMD parts) matters more than absolute accuracy for repair; affordable smartphone-attachment thermal cameras (FLIR ONE, Seek) make the capability accessible.
• Safety: you must power a faulty board — do it through a current-limited supply with the limit low, watch briefly, don't touch the hotspot (burn risk), and respect mains hazards (Sections 3.1, 3.2).

Skills Learned

• You can now explain what a thermal camera images and displays.
• You can now locate a short or overload by finding the abnormal hotspot.
• You can now account for emissivity when a shiny part reads falsely cool.
• You can now power a suspected-short board safely through a current-limited supply.
• You can now read a cold stage as a sign that no power is reaching it.

Glossary Additions

• thermal camera — a camera that images infrared (heat) radiation and displays the temperature distribution of its subject as a false-color image (a heat map), with hot areas shown bright (red, orange, or white) and cool areas dark (blue or black); in repair it lets you see, non-contact, which part of a powered board is hot, making it a fast way to locate a shorted or overloaded component that dissipates extra power. Also called a thermal imaging camera.
• emissivity — a measure of how well a surface emits infrared radiation, which determines how accurately a thermal camera reads its temperature: matte, dark surfaces have high emissivity and read close to true, while shiny, metallic, or reflective surfaces have low emissivity and read too cool (or reflect other temperatures), so they can hide a genuinely hot part. Coating a shiny surface with matte paint, tape, or emissivity spray raises its effective emissivity for a true reading.
• hotspot — a localized area of abnormally high temperature on a powered board, seen on a thermal camera as a bright spot; because a shorted, overloaded, or failing component dissipates extra power as heat, the hotspot often marks the faulty part directly. In fault-finding the diagnostic target is the relative hotspot — the outlier hotter than its neighbors, than expected, or than a known-good board — rather than a specific absolute temperature.
• infrared radiation — the invisible heat radiation that every object emits in proportion to its temperature; a thermal camera's sensor images this infrared and maps it to color to produce a temperature picture. Because hotter objects emit more infrared, the camera can render a board's temperature distribution and reveal where power is being dissipated.

Suggested Next Sections

Must read next:

• ESR Meters — Testing Capacitors In-Circuit — the next advanced tool: an ESR meter that finds a failing electrolytic capacitor by its equivalent series resistance, often without unsoldering it — catching the bad cap that measures fine on capacitance and that a thermal or ripple check only hints at.

Recommended:

• Current Measurement — measuring excess current draw is how you know a short exists; thermal imaging is how you find which part it is.
• Logic Analyzers — When and How — the sibling advanced tool for the other kind of question: many digital lines, rather than where the heat is.