A switching power supply looks like a box of unrelated parts until you see the jobs. Five component types — the switch, the transformer, the rectifier diode, the output capacitor, and the controller — carry the whole conversion, and each one is a worker with a specific job whose failure has a specific signature. When an SMPS stops delivering power or efficiency, the symptom points back to one of the five workers. This article maps the five roles, names what each must do, shows the failure signature each leaves, and closes with the board-reading habit that turns a dead or inefficient supply into a diagnosis.
The Five Workers and Their Jobs
The switching supply converts AC line power to regulated DC through a chain of five jobs, and the chain is worth stating before any part is discussed. The controller decides, the switch chops, the transformer steps, the rectifier converts, and the capacitor smooths. Each worker’s job defines how the part is rated and how it fails.
The chain runs in order. The controller is the brain: it samples the output, compares it to the target, and decides when the switch turns on and off. The switch is the muscle: a MOSFET (or in some designs another fast-switching device) that chops the input current at the switching frequency, creating the pulse train that the rest of the chain converts. The transformer steps the pulsed energy between voltage levels via its leakage and magnetizing paths. The rectifier converts the stepped AC-like pulses back to DC. And the output capacitor smooths the rectified pulses into a usable rail. The SMPS components article and the SMPS guide define the parts; this article adds the failure-signature view that turns the block diagram into a diagnostic.
The five-job map is the whole value of the article, because a supply with a dead output can have a failed switch, a failed rectifier, or a failed capacitor — and the fix is different for each. Reading the board by the roles, rather than by the parts list, is the habit the final section builds.
The Switch: Rating and Loss at the Heart
The switch is the component that makes the SMPS an SMPS, and its ratings and losses set the supply’s ceiling. The switch does the chopping, and everything downstream is built to convert what the switch produces.
The switch’s first rating is voltage: it must block the peak of the line or the stepped bus with margin, and hold it through the switching events. Its second is current: the switch carries the input-side current in pulses, and its continuous and surge ratings must cover the worst duty. Its third is thermal: the switch dissipates conduction and switching loss, and the junction must stay under the limit at the load’s duty cycle. The switch is also where the switching loss lives — the per-event energy that multiplies by the frequency and shows up as heat at the heart of the supply. The MOSFET selection guide and the MOSFET for SMPS article cover the switch rating and selection.
The failure signature of the switch is the supply that stops switching. A supply that outputs nothing, has no switching activity, and runs cool at the switch is the classic failed-switch signature; a supply that heats at the switch while the output sags is the switch running over its ratings. The signature is the first diagnostic, and it separates the switch’s job from the rectifier’s.
The transformer deserves its own note because it is the worker that makes the SMPS small and is also the one most often forgotten in a first diagnosis. Its job is to step the pulsed energy between voltage levels: the primary winding sees the switch’s square-wave pulses, and the secondary delivers a stepped version to the rectifier. Because it runs at the switching frequency rather than the line frequency, the transformer core can be a fraction of the size of a line-frequency transformer — that is the size advantage of the SMPS made physical. The transformer’s ratings are the voltage stress of the insulation, the current of the windings, and the leakage inductance and magnetizing current that store the energy the switch and rectifier must handle. Its failure signature is the supply that hums, heats at the core, or leaks the switching energy as losses the rectifier and switch then carry. The SMPS guide and the isolated power supply article cover the transformer stages of the conversion chain.
The transformer also defines the boundary between the roles. The leakage energy it stores is what the snubber and the rectifier handle at every switch edge; the magnetizing current it draws is a loss term the designer budgets; and the insulation it provides sets the isolation safety of the whole supply. A first-read technician who skips the transformer will mis-attribute its signature — hum and core heat — to the switch or the rectifier. The five-worker map keeps the transformer in the read.
The Output Rectifier: Fastness and Loss
The rectifier is the worker that converts the transformer’s pulses back to DC, and its two properties — fastness and loss — decide how much of the conversion is wasted.
The rectifier’s speed matters because the transformer output is a high-frequency pulse train. Every turn-off of the rectifier lets the junction discharge stored charge, and that reverse-recovery event is loss and stress at the switching frequency. A standard-recovery rectifier at high frequency dissipates a measurable recovery loss; a fast-recovery or Schottky part cuts it. The rectifier’s conduction loss, the VF times the average current, is the second term. The rectifier selection for SMPS article and the fast recovery diode guide name the three roles and the enemies of each.
The failure signature of the rectifier is efficiency loss and heat at the output stage. A supply that works but runs hot at the rectifier, or one whose efficiency has dropped over time, points at the rectifier’s loss terms — a part running over its rating, an aging part, or a wrong recovery speed. The output rectifier is also where the synchronous-rectification upgrade lives, replacing the diode with a controlled MOSFET to cut the conduction loss; the synchronous rectifier article covers that role.
Capacitors and Controller: The Quiet Assumptions
The capacitor and the controller are the quiet workers: they do not produce the visible loss, but they carry the assumptions the whole design leans on, and their failures produce the most confusing symptoms.
The output capacitor’s job is to smooth the rectified pulses into a regulated rail, and its assumptions are the ripple budget and the hold-up energy. An undersized or aged capacitor passes more ripple to the load, and a capacitor with failing dielectric or ESR shows as ripple, noise, or a supply that cannot hold the rail under load steps. The controller’s job is to sample and decide, and its assumptions are the control-loop stability and the sensing accuracy. A controller that mis-samples, loses gain, or runs unstable produces a supply that hunts, oscillates, or fails to regulate — symptoms that look like the switch or rectifier until the controller is checked. The ripple diagnosis guide and the EMI filter article cover the capacitor’s ripple role and the filtering that follows.
The failure signatures are distinctive once known. Ripple or noise on the output under a stable load points at the capacitor or the filtering; a supply that regulates slowly, overshoots, or hunts points at the controller. The quiet workers fail quietly — the symptom takes the blame from the loud workers unless the board is read by role.
The component-to-symptom map is compact enough for a table, and the table is the diagnostic tool:
| Symptom | First suspect | Why | Confirm with |
|---|---|---|---|
| No output, no switching | Switch | Nothing chops the input | Gate waveform at the switch |
| No output, switch runs cool | Controller | Nothing commands the switch | Controller output and supply |
| Low output, hot rectifier | Rectifier | Conduction or recovery loss | Case temperature, output voltage |
| Ripple or noise on rail | Output capacitor | Filtering degraded or undersized | Scope on output, ESR check |
| Hunt, overshoot, instability | Controller | Loop unstable or mis-sampling | Output under load step |
The table is the board-reading habit in one view. Each symptom names the worker most likely responsible, the reason, and the confirming measurement — which is exactly the order a technician should follow instead of replacing parts by guess. The power supply repair guide and the ripple diagnosis guide provide the deeper test sequences this table starts.
Reading an SMPS Board: Component-to-Symptom
The last habit is the payoff: reading the board by component-to-symptom. A technician or engineer with the five-worker map can turn a symptom into a part in minutes, and the habit is the difference between diagnosis and parts-drawer guessing.
The read starts at the symptom. No output at all: check the switch first (is it switching?), then the controller (is it commanding?), then the rectifier and the fuse. Low output or sagging under load: check the rectifier’s loss and the capacitor’s health before the switch. Hot box or efficiency complaint: the loss terms of the switch and the rectifier, plus the thermal path, are the suspects. Ripple, noise, or instability: the capacitor, the filtering, and the controller’s loop are where the cause lives. The power supply repair article and the ripple diagnosis guide give the test order, and the field reliability checklist turns the read into a pre-power audit.
The five workers and their jobs are the skeleton of every SMPS, and the failure signature of each is the diagnostic shortcut. The controller decides, the switch chops, the transformer steps, the rectifier converts, and the capacitor smooths — and when the supply misbehaves, the symptom names the worker. Reading the board by role, checking the switch before the capacitor, and understanding the loss terms of the rectifier turns the SMPS from a black box into a diagnosable chain. The general rectifier category supplies the rectifier parts whose roles this article maps, and the SMPS components article is the block-diagram companion to the role map.