>No Power, No Problem: A Practical Guide to Repairing Dead Switch-Mode Power Supplies with the Help of Service Manuals

Ask anyone who repairs electronics for a living what the most common fault is, and you will hear the same answer again and again: "It's dead. Nothing happens when I press the button." Televisions, monitors, audio amplifiers, satellite receivers, printers, game consoles, network equipment, LED drivers, washing machine control boards – almost every modern device gets its energy from a switch-mode power supply (SMPS), and that power supply is, statistically, the part most likely to fail.

The good news is that a dead power supply is also one of the most rewarding repairs you can learn. The parts are usually cheap, the failure patterns repeat across brands, and with a service manual on the bench you can turn a "beyond economical repair" verdict into a working device in an afternoon. This article walks through the whole process: how an SMPS works, how to work on one safely, how to read the power supply section of a service manual, and a step-by-step diagnostic method that covers the vast majority of "no power" faults.

Throughout the article we will refer to the resources our community uses every day: the free archive at eserviceinfo.com, the manual search engine at jabse.com, and the Service Manuals group on Facebook (facebook.com/groups/servicemanuals), where thousands of technicians and hobbyists share schematics, tips and requests.

Safety First: This Is the Dangerous Part of the Device

Before anything else, a clear warning. The primary side of a switch-mode power supply is directly connected to the mains. It is not isolated, it can kill, and it can stay dangerous after the plug has been pulled. Please read this section even if you have years of experience – most accidents happen to people who know what they are doing and get careless.

Why an unplugged power supply can still hurt you

Mains AC is rectified and stored in a large electrolytic capacitor, often called the bulk or main filter capacitor. On a 230 V supply this capacitor charges to roughly 325 V DC (230 V × 1.414). On equipment with active power factor correction (PFC), the bus is boosted further to around 380–400 V. On 120 V mains the unboosted figure is about 170 V. A healthy supply normally has bleeder paths that discharge this capacitor, but a dead supply is exactly the case where the circuit that would normally drain the capacitor may not be working. The capacitor can hold its charge for minutes or even hours.

Basic safety rules

Always discharge the bulk capacitor before touching the board. Use a resistor of roughly 1–10 kΩ rated at 5 W or more, held with insulated probes or clips, across the capacitor terminals. Never short it with a screwdriver – the arc can damage the capacitor, the board and your eyes.
Measure after discharging. Confirm with your multimeter on the DC voltage range that the capacitor is below a safe level before touching anything.
Use an isolation transformer when you need to make measurements on the live primary side, especially with an oscilloscope. Without one, connecting a mains-earthed scope ground clip to the primary side can create a short circuit through the earth.
Use a series lamp limiter (more on this below) when powering up a repaired supply for the first time.
Know the hot and cold sides. Most service manuals and many PCBs mark the primary side with a hatched area, a warning symbol, or a dashed isolation line. Learn to spot it.
Wear eye protection. Electrolytic capacitors and semiconductors can explode when power is applied to a fault.

How a Switch-Mode Power Supply Works (In Plain Language)

You do not need to be an electronics engineer to repair an SMPS, but you do need to understand the basic chain of events. Once you see the supply as a sequence of blocks, troubleshooting becomes a matter of finding the first block in the chain that is not doing its job.

Input and protection. Mains enters through a fuse, sometimes an NTC thermistor to limit inrush current, a varistor (MOV) to clamp surges, and an EMI filter made of capacitors and a common-mode choke.
Rectification. A bridge rectifier (four diodes, or one bridge package) converts AC into pulsating DC.
Bulk storage (and PFC). The bulk capacitor smooths the DC. In larger supplies, a PFC stage with its own controller, inductor and MOSFET sits in front of the bulk capacitor.
Start-up. The PWM controller IC needs a small supply voltage to begin oscillating. At first, it gets this from high-value start-up resistors (or an internal high-voltage current source) fed from the bulk voltage.
Switching. The controller drives a power MOSFET (or an integrated switch inside the IC) on and off tens of thousands of times per second, chopping current through the primary winding of a small high-frequency transformer.
Self-supply. Once running, an auxiliary winding on the transformer, rectified by a small diode and capacitor, takes over powering the controller. This is called the bootstrap or Vcc winding.
Secondary rectification. On the isolated side, fast or Schottky diodes and low-ESR capacitors convert the transformer output into the DC rails the device needs (for example 5 V, 12 V, 24 V).
Feedback and regulation. A reference (very often a TL431) compares the output voltage to a target and drives an optocoupler. The optocoupler sends the information across the isolation barrier to the controller, which adjusts the duty cycle to keep the output steady.
Standby and control. Many devices have a separate standby supply that is always on, and a main supply that is switched on by a signal from the main board (often labelled PS-ON, STB, POWER_ON or similar in the service manual).

When you understand this chain, the question "why is it dead?" becomes "which link is broken?" – and the service manual is the map that shows you every link.

Why the Service Manual Matters for Power Supply Repair

It is possible to repair simple supplies by experience alone, and many technicians do. But a service manual changes the work from guesswork to a methodical process. Here is what a good manual gives you that the board itself does not:

The schematic. It shows how every component connects, including tracks hidden under parts or running on the other side of the board.
Component values and ratings. Burnt resistors are often unreadable. The schematic or parts list tells you the original value, wattage and sometimes a "safety critical" marking that means you must use an identical replacement.
Test point voltages. Many manuals print expected voltages on the schematic or in a separate table, sometimes for both standby and power-on states.
Waveforms. Higher-quality manuals include oscilloscope screenshots for key points such as the MOSFET drain or gate.
Troubleshooting flowcharts. Manufacturers often include a "No Power" or "Power Supply Does Not Start" chart that walks you through checks in the order the factory engineers recommend.
Board identification. The same TV model may have been built with several different power boards. The manual, together with the board's own part number, helps you confirm you are working on the right version.
Protection circuits explained. Over-voltage, over-current and thermal protection circuits are a frequent cause of "dead" units that are actually in shutdown. The manual tells you what triggers them.

Finding the right document

When searching for a manual, look beyond the product model number. Power boards usually carry their own identification printed on the PCB – something like a part number, a board code or a code on a sticker. Searching for that board code on jabse.com or eserviceinfo.com often finds a schematic even when the full device manual is hard to locate, because the same power board is shared across many models and even across brands.

It is also worth searching by the PWM controller IC part number. The IC datasheet is not a service manual, but it contains the pin functions, start-up thresholds and a typical application circuit, which is often very close to what is on your board.

If you cannot find a document anywhere, post a request in the Service Manuals Facebook group with clear photos of the board, the board code and the device model. Someone may have the exact schematic in their personal collection – and once it is shared, it can be added to the archive so the next person does not have to ask.

The Tools You Need

You do not need an expensive laboratory. Most power supply repairs can be done with a modest set of tools.

Recommended tools for SMPS repair
Tool	Why you need it	Priority
Digital multimeter	Voltage, resistance, continuity and diode tests – the basis of all diagnostics	Essential
Discharge resistor with insulated leads	Safely discharging the bulk capacitor and other high-voltage capacitors	Essential
Series lamp limiter	Limits current on first power-up so a remaining fault does not destroy new parts	Essential
Temperature-controlled soldering iron	Replacing components without lifting pads or overheating the board	Essential
Desoldering pump or braid	Removing through-hole parts cleanly	Essential
ESR meter or component tester	Finding electrolytic capacitors that look fine but have high internal resistance	Highly recommended
Isolation transformer	Safer live measurements on the primary side, required for scope work there	Highly recommended
Variable bench power supply	Powering the controller IC directly to test start-up behaviour without mains	Recommended
Oscilloscope	Seeing switching waveforms, ripple and intermittent faults	Useful for advanced work

Building a series lamp limiter

The series lamp limiter (often called a "dim bulb tester") is the single cheapest tool that will save you the most money. It is simply an incandescent lamp wired in series with the mains feed to the device under test, mounted in a proper enclosure with a socket for the device. If there is still a short in the supply, the lamp lights up brightly and limits the current, instead of the fuse blowing or your new MOSFET exploding.

A few practical notes:

Use a classic incandescent lamp, not an LED or compact fluorescent – only an incandescent filament has the right resistance behaviour. Halogen lamps also work.
For small supplies, start with a 40–60 W lamp. For larger TV supplies, 60–100 W.
A healthy supply typically makes the lamp flash briefly at switch-on (while the bulk capacitor charges) and then glow dimly or go dark. A lamp that stays at full brightness indicates a short or heavy overload.
The enclosure must be built to mains wiring standards. If you are not qualified to build one, buy a ready-made unit.

Step One: Observe Before You Measure

Many repairs are solved in the first five minutes, before a meter even touches the board. Remove the board, put it under good light, and look carefully. Smell it too – burnt components have a very distinctive odour.

What to look for

Bulging or leaking electrolytic capacitors. Domed tops, brown crust at the base, or a capacitor sitting crooked because its rubber seal has pushed out.
A blackened or shattered fuse. A glass fuse that is blackened inside or has a vaporised wire usually means a hard short on the primary side. A fuse with a neatly broken wire may have failed from age or a surge.
Cracked or burnt components. Look at the MOSFET, the bridge rectifier, the NTC thermistor, the varistor and resistors around the controller.
Dry or cracked solder joints. Especially on transformer pins, chokes, large capacitors and heatsink mounting points – parts that heat up and move with thermal cycling. A ring-shaped crack around a pin is classic.
Discoloured PCB. Brown or dark areas show where something ran hot for a long time.
Corrosion and contamination. Liquid spills, insects, dust mixed with humidity or kitchen grease can create leakage paths, especially around high-impedance start-up and feedback circuits.

Photograph the board from both sides before you start removing parts. Those pictures are invaluable when you reassemble, and they are exactly what you will need if you post a question in the Facebook group.

Step Two: Cold Checks with the Power Off

After discharging the bulk capacitor and confirming it is safe, start with resistance and diode tests on the most likely failure points. Work from the mains input towards the output, following the chain described earlier.

The primary side checklist

Fuse: continuity test. If it is blown, do not just replace it and plug in – find out why it blew.
NTC thermistor: should read a few ohms to a few tens of ohms at room temperature. Open means failed.
Varistor (MOV): should read open circuit. A short or visible crack means it took a surge and failed.
Bridge rectifier: use diode mode to test each of the four diodes. A shorted diode shows near zero in both directions.
Bulk capacitor: check for a short across its terminals (after the initial charging reading settles). Check the capacitance if you have a meter that can do so.
Power MOSFET: measure between drain, source and gate. A failed MOSFET often reads as a short or very low resistance between drain and source. Many fail with all three pins shorted together.
Current sense resistor: a low-value resistor (often below 1 Ω) between the MOSFET source and ground. When the MOSFET shorts, this resistor often goes open. Check it with care – a meter on its lowest range should read close to the marked value.
Controller IC: when the MOSFET fails violently, the gate drive pin of the controller is frequently damaged too. Check for shorts between the gate pin and ground. If in doubt, replace the IC along with the MOSFET.
Gate resistor and gate-source zener: small parts that are easily destroyed in the same event.
Start-up resistors: high-value resistors (often hundreds of kilohms) feeding the controller from the bulk voltage. These are a very common failure – they drift high or go open over the years because they carry high voltage continuously.
Vcc capacitor and auxiliary diode: the small electrolytic capacitor on the controller's supply pin and the diode from the auxiliary winding.

The secondary side checklist

Output rectifier diodes: Schottky diodes on the low-voltage rails often fail short. A short here can make the supply "hiccup" or refuse to start.
Output capacitors: the most common failure of all. Check ESR, not just capacitance.
Short circuits on the rails: measure resistance from each output to ground. A very low reading may point to a shorted diode, capacitor, or a fault on the device's main board that is dragging the supply down.
TL431 and optocoupler: harder to test in circuit, but visual damage or a reading inconsistent with the schematic is a clue.

Electrolytic Capacitors: The Number One Suspect

If you only remember one thing from this article, remember this: most dead power supplies have bad capacitors. Heat is the enemy of electrolytic capacitors, and power supplies are hot places. Over time the electrolyte dries out, capacitance drops, and – most importantly – the equivalent series resistance (ESR) rises.

A capacitor with high ESR can look perfect, show nearly full capacitance on a basic meter, and still be useless in a switching circuit operating at tens of kilohertz. That is why an ESR meter is such a valuable tool.

Typical symptoms caused by bad capacitors

The device takes several attempts to turn on, or only starts after warming up.
The standby LED blinks or the supply makes a ticking or chirping sound.
The device turns on and immediately off, or cycles repeatedly.
Whining or squealing from the transformer.
Output voltages that are low, unstable or noisy.
Completely dead, because the controller's Vcc capacitor can no longer hold the supply voltage during start-up.

The small capacitor on the controller IC's Vcc pin deserves special attention. It is often tiny – 10 µF to 100 µF – and frequently sits right next to a hot heatsink. When it dries out, the controller tries to start, the Vcc voltage collapses before the auxiliary winding can take over, and the cycle repeats. The result is a supply that looks completely dead or that "ticks". Replacing one capacitor worth a few cents fixes an astonishing number of devices.

Choosing replacement capacitors

What to match when replacing electrolytic capacitors
Parameter	Rule of thumb
Capacitance	Use the same value. A modestly higher value is usually acceptable on output rails, but avoid large changes in timing or feedback circuits.
Voltage rating	Same or higher. Never lower.
Temperature rating	105 °C types are strongly preferred in power supplies, even if the original was 85 °C.
ESR / ripple current	On secondary outputs, use low-ESR, high-ripple-current types designed for switching supplies. A general-purpose capacitor may fail quickly.
Polarity	Double-check before soldering. A reversed electrolytic capacitor can fail violently on power-up.

Step Three: Understanding the Classic Failure Patterns

Over time you will notice that dead power supplies fall into a handful of recognisable patterns. Identifying which pattern you are dealing with points you directly to the right group of components.

Common SMPS fault patterns and where to look
Symptom	Likely causes	First checks
Fuse blown, blackened inside	Shorted MOSFET, bridge rectifier, bulk capacitor or PFC transistor	MOSFET, sense resistor, controller gate pin, bridge, PFC section
Fuse intact, no voltage on bulk capacitor	Open NTC, open bridge diode, broken track, damaged input filter choke	Trace the mains path from the input connector to the bulk capacitor
Bulk voltage present, no output at all	Controller not starting: open start-up resistor, dry Vcc capacitor, failed controller	Voltage on controller Vcc pin, start-up resistors, Vcc capacitor
Ticking, chirping, or "hiccup" cycling	Overload on secondary, shorted output diode, bad output capacitors, failed auxiliary supply	Secondary diodes and capacitors, rail resistance to ground, Vcc capacitor
Outputs present but too low or unstable	High-ESR capacitors, faulty feedback loop, drifting resistors	Output capacitors, TL431, optocoupler, feedback divider resistors
Outputs much too high	Feedback loop broken – open optocoupler, failed TL431 or open divider resistor	Do not run under load; check the feedback components immediately
Standby works, main supply does not turn on	Missing power-on signal from main board, faulty switching transistor, protection latch	Power-on signal at the connector, protection circuit status
Works cold, fails when warm (or the reverse)	Cracked solder joints, temperature-sensitive capacitor, failing semiconductor	Reflow joints, use freezer spray or gentle heat to locate the part

The dangerous case: output voltage too high

A broken feedback loop deserves a special mention. If the optocoupler, TL431 or a resistor in the voltage divider fails open, the controller no longer receives any information about the output voltage. Depending on the design, it may push the output far above its rated value, which can destroy everything connected to it. Good designs include over-voltage protection to shut the supply down, but do not rely on it. If you measure an output that is clearly too high, disconnect the load and fix the feedback circuit before going further.

Step Four: Testing the Controller Without Mains

One of the most useful techniques – and one that keeps you away from mains voltage – is powering the PWM controller IC directly from a low-voltage bench power supply.

Most controller ICs have an under-voltage lockout (UVLO): they will not start until their Vcc pin reaches a turn-on threshold, and they stop when it falls below a lower turn-off threshold. These values are in the IC datasheet and sometimes in the service manual. For example, in the widely used UC384x family, the UC3842 and UC3844 have a turn-on threshold of about 16 V and a turn-off threshold of about 10 V, while the UC3843 and UC3845 start at about 8.4 V and stop at about 7.6 V.

How to do it

Make sure the board is disconnected from mains and the bulk capacitor is discharged.
Find the controller's Vcc and ground pins using the schematic and the datasheet.
Connect a current-limited bench supply to Vcc and ground. Set the current limit low, around 50 mA.
Slowly increase the voltage past the turn-on threshold, and do not exceed the IC's absolute maximum Vcc rating.
With an oscilloscope, check for a switching waveform on the gate drive pin. On the UC384x family you can also check that the reference pin comes up (5 V) and that the oscillator pin shows a sawtooth waveform.

If the controller oscillates on the bench supply, the IC and its timing components are probably fine, and the problem is in the start-up path or the self-supply from the auxiliary winding. If it does not oscillate, check the IC's supporting components, then the IC itself.

Step Five: Replacing the Power MOSFET Correctly

When a MOSFET has failed short, it is tempting to drop in any similar transistor and switch on. This is how many technicians destroy a second and third MOSFET in a row. A few rules make the difference.

Find out why it failed. A MOSFET rarely dies on its own. Check the controller, the gate resistor, the gate zener, the sense resistor, the snubber network (the resistor-capacitor-diode clamp across the primary winding) and the output side for overloads.
Match or exceed the ratings. Drain-source voltage, continuous current, and on-resistance should be at least as good as the original. The parts list in the service manual is the best source for the exact original part number.
Watch gate charge and package. A MOSFET with much higher gate charge may switch slowly and run hot. The package and pinout must match the PCB footprint.
Restore the thermal path. Clean the old thermal compound, use a new insulating pad or washer if one was fitted, and check that the transistor is electrically isolated from the heatsink if the design requires it.
Replace the snubber parts if in doubt. A failed snubber diode or resistor lets voltage spikes on the drain exceed the MOSFET rating, leading to repeated failures.
Power up through the lamp limiter the first time, every time.

Step Six: Protection Circuits – When "Dead" Means "Protecting Itself"

Not every dead device has a failed power supply. Modern equipment is full of protection circuits designed to shut everything down when something goes wrong. A TV may have a perfectly good power supply that is being held off because the backlight driver, the main board or an audio amplifier reports a fault.

This is another area where the service manual is irreplaceable. It typically describes:

Which signals the power board monitors (over-voltage, over-current, temperature, signals from other boards).
How the protection behaves – whether it shuts down permanently until the mains is removed (latching), or retries periodically (hiccup mode).
Blink codes from the standby LED that identify which protection circuit was triggered.
The sequence of power-on events and signals between boards, so you can see which step is missing.

Isolating the power board

A common technique is to disconnect the power board from the rest of the device and test it alone with a dummy load. The service manual will tell you which pin carries the power-on signal and what voltage it needs. Resistive loads, such as power resistors or automotive bulbs of suitable rating, connected to the main outputs let you check whether the supply can deliver its rated voltages without the rest of the device. Many supplies will not regulate correctly without a minimum load, so do not judge a supply as faulty just because it misbehaves with no load at all.

If the supply works perfectly on the bench with a dummy load but shuts down when connected to the device, the fault is elsewhere – on the main board, in the backlight inverter, or in a part that is drawing too much current. The power supply may simply be doing its job.

Step Seven: Power-Up, Verify and Burn In

Once the faulty components are replaced, resist the urge to plug straight into the wall. Follow a controlled sequence.

Re-check your work. Look for solder bridges, reversed capacitors, wrong transistor orientation and leftover flux or solder balls.
Measure resistance again across the mains input and across each output to confirm there are no shorts.
Power up through the series lamp limiter. Watch the lamp: a short flash followed by a dim glow is a good sign.
Measure the bulk voltage and compare it with the expected value from the service manual.
Measure the outputs and compare them to the schematic or the voltage table in the manual. Check both standby and power-on states if the supply has them.
Apply a load. Connect a dummy load or the real device, now connected directly to the mains without the limiter.
Check ripple with an oscilloscope if you have one. Excessive ripple points to capacitors that still need replacing.
Burn in. Run the device for at least an hour, ideally several. Monitor temperatures of the MOSFET, output diodes and transformer. Intermittent faults often reveal themselves during this period.

Real-World Examples of Common Repairs

The LCD monitor that will not switch on

An older LCD monitor shows no power LED at all. The fuse is fine, and the bulk capacitor charges to the expected voltage. The controller Vcc pin, measured carefully through an isolation transformer, rises and falls repeatedly but never stays up. The small Vcc capacitor next to the heatsink tests at a fraction of its rated capacitance with a high ESR. Replacing it, along with two bulging capacitors on the 12 V output, brings the monitor back to life. Total parts cost: a few euros.

The TV that clicks but never turns on

An LED TV shows a red standby LED, clicks a relay when switched on, then returns to standby. The service manual's power-on sequence shows that the main board requests the main supply and then checks for a feedback signal from the backlight driver. Testing the power board alone with a dummy load shows correct outputs. The fault turns out to be a failed LED strip in the backlight, causing the driver to report an error. The power supply was perfectly healthy – without the manual's sequence diagram, many hours could have been wasted on the wrong board.

Common Mistakes to Avoid

Replacing the fuse with a higher rating or bridging it with wire. The fuse is protecting the house wiring and you – never defeat it.
Replacing parts without finding the root cause. A blown MOSFET is a symptom; the cause may be elsewhere.
Testing capacitors for capacitance only. ESR matters more in a switching supply.
Ignoring "safety critical" markings in the service manual. Parts marked with a warning symbol (often a triangle or an exclamation mark, sometimes a shaded area on the schematic) must be replaced with identical specified types – these include fuses, fusible resistors, Y-capacitors and isolation components.
Measuring the primary side with a scope that is not isolated. This can create a dangerous short circuit through the earth connection.
Assuming the power supply is at fault. In complex devices, protection circuits often shut down a healthy supply.

How Our Community Can Help You

Even experienced technicians regularly need a second opinion or a missing schematic. Here is where to find them:

eserviceinfo.com

A free archive of more than 200,000 service manuals, schematics and technical documents covering consumer electronics, appliances, computer equipment, audio, video and much more. Search by brand, model or board code, and download the documents you need for your repair.

jabse.com

The Jabse Manuals Search Engine helps you find service manuals and user manuals quickly. If you are not sure of the exact document name, try searching with the model number, the chassis code or the power board identification printed on the PCB.

The Service Manuals Facebook group

Our group at facebook.com/groups/servicemanuals brings together technicians, engineers and hobbyists from around the world. You can:

Request a service manual or schematic that you cannot find elsewhere.
Share clear photos of a board and ask for help identifying a burnt component.
Describe a fault and get suggestions from people who have seen it before.
Share your own repair experiences and documents, helping the next person with the same problem.

When you post, include the model number, the board code, a clear description of the symptom, your measurements and sharp photos of both sides of the board. And when your repair is successful, please come back and post what the fault was. That one short message may save someone else hours of work months or years from now.

Quick Reference: Dead SMPS Diagnostic Sequence

To finish, here is the whole method condensed into a short checklist you can keep on your bench.

Find and download the service manual or power board schematic.
Discharge the bulk capacitor and confirm with a meter.
Inspect visually: capacitors, fuse, burnt parts, solder joints, corrosion.
Test the input path: fuse, NTC, varistor, bridge rectifier, bulk capacitor.
Test the switching stage: MOSFET, sense resistor, gate components, controller gate pin, snubber.
Test the start-up path: start-up resistors, Vcc capacitor, auxiliary winding diode.
Test the secondary: output diodes, output capacitors (ESR), shorts on the rails.
Check the feedback loop: TL431, optocoupler, divider resistors.
If needed, test the controller with a bench supply on its Vcc pin.
Replace faulty parts with correct ratings, respecting safety-critical markings.
Power up through a series lamp limiter and measure voltages against the manual.
Test with a load, check ripple and burn in for at least an hour.
If the supply works alone but not in the device, investigate protection circuits and other boards.
Share your findings with the community.

Dead power supplies are not mysterious. They fail in predictable ways, and with a methodical approach, basic tools and the right documentation, most of them can be brought back to life. The next time someone tells you a device is "dead" and not worth repairing, open the case, find the service manual on eserviceinfo.com or jabse.com, and give it a chance. And if you get stuck, the Service Manuals group is always there to help.

Happy repairing – and stay safe around that bulk capacitor.