Engineering Against Waste. We give electronics a second chance. Two engineers, one hot‑air station, and a clear mission: turning so‑called “trash” back into working tech. Instead of throwing things away, send us your defective devices — the ones everyone else has already written off. We perform real board‑level repairs and document every case with full transparency. Less waste. More repair culture.
Every episode is one board, start to finish. Nothing edited out — including the times we lose.
The Logitech Z313 Speaker System sits on the bench, silent. For years it has amplified movies, carried game soundtracks, and filled rooms with more bass than its size ever promised. Now it gives nothing back. No audio, no hum — only a power LED that blinks, slow and uncertain, like a heartbeat that has forgotten its rhythm.
We're waiting for the patient.
The Amazon Fire HD 10 Kids Edition arrives not in the usual battle‑worn state of a child’s device, but in something far more unsettling. A worried mother comes to us immediately, hoping to prevent further damage. Her son has spilled juice over the tablet — not a splash, but a slow, creeping flood. The SD card slot is sealed shut, the card itself glued into place, and every surface carries a tacky residue that clings to fingers and tools alike.
We’re checking the patient.
The Nintendo Wii U Premium powers on with its familiar blue glow, the SoC beginning its early initialization sequence. For a brief moment, the console appears stable, modules waking in quiet succession. Then, without warning, a fatal system error cuts through the startup routine. No controller input, no menu, no chance for recovery — only an abrupt halt in the earliest operational phase, as if the system loses its footing the moment it tries to stand.
The console is fully disassembled, cleaned, and inspected at hardware level. The installed Samsung eMMC shows clear signs of data corruption, but no physical damage. Thermals are stable, pads and cooling surfaces intact. Using the Raspberry‑Pi‑Pico recovery path, the system is re‑initialized. OTP and SEEPROM are safely dumped, the ColdBoot title is restored, and the unit is shut down cleanly. On the next power‑on, the Wii U wakes up completely: initial setup, system menu, input handling — all stable. The patient has returned to life.
The brew group stays outside like someone who decided they are not coming back in today. You push it in, it stops cold. The machine acts like it completely lost its sense of direction. Something inside is out of sync. The drive does not move into the position the machine expects. The mechanics look solid. No jammed guides. No visible damage. The machine behaves as if it had an internal blackout and no longer recognizes the brew group as part of its own system.
The brew group refuses to enter. It stops immediately, as if the machine no longer knows where “home” is. Mechanics are intact, no jams, no broken guides — but the drive sits out of its zero position. The drive is realigned to its reference mark. Once synchronized, the brew group slides in smoothly and the machine regains orientation. The patient is alive again.
The NVIDIA GeForce GT 545 spent its life as a reliable CAD graphics card, steady and uncomplaining. For years it rotated models, calculated edges, and kept designs precise. Now it has reached its thermal limit. The temperature climbs, the heatsink glows, and the card whispers: “I’ve done my job.”
We took the patient to the bench. The card was fully disassembled, every surface freed from dust and residue. Old thermal paste came off in dry, cracked flakes. The heatsink was cleaned, the fan restored, and several cold solder joints on the main PCB were located and fixed. Fresh paste went on, the cooler was re‑mounted, and the card was rebuilt piece by piece. A final FurMark run pushed the card back into the heat it once feared. Temperatures rose, stabilized, and held. No artifacts, no throttling, no complaints. The quiet workhorse didn’t retire that day.
The ASUS ENGT520 SILENT is tripping. It boots up totally fine, but minutes later the screen goes wild with heavy glitches, pixel artifacts, and nasty lines. A quick reboot fixes the drama for a second, but then the visual chaos returns immediately. This is a textbook sign of overheating or dying VRAM. Since this silent card has zero fans, old thermal paste or bad case airflow is basically cooking the GPU alive.
The patient arrived in classic “I’ve been in a drawer for ten years” condition: dust-packed fins, fossilized thermal paste, and RAM chips running naked like it’s 2011. Step one was a full decontamination. The heatsink got a proper cleaning, the fan blades lost their dust armor, and the PCI connector received a careful scrub until the contacts shone again. The old thermal paste was scraped off — it had the consistency of stale chewing gum — and replaced with a fresh, even layer of modern compound. Since the original design didn’t bother giving the RAM any love, I added thermal pads to bring the card up to present-day sanity. Reassembled, tested, and alive again. Temperatures dropped, stability returned, and the ENGT520 is officially back from the dead.
Display remains dark despite fresh battery installed. No continuity buzzer, no voltage indication, no measurable activity on any range. Test leads show intermittent connection and the battery compartment exhibits visible corrosion.
Root cause traced to a corroded battery clip causing a complete supply interruption, compounded by worn-out test leads with intermittent open circuits. Replaced the battery clip, fitted new test leads, cleaned corrosion residues from the PCB and battery compartment, and restored all electrical contact surfaces. Finished with a full functional verification across voltage, resistance, and continuity ranges.
Central display missing multiple segments. Odometer and trip information remain partially visible, but large sections of the LCD have become unreadable over time.
A fresh LCD display took the throne. The broken backlight trace was hunted down and revived with precision. The missing LED was restored to its rightful place, bringing balance back to the illumination realm. Now the cluster wakes up sharp, bright, and fully alive, pixels crisp, backlight steady, every LED reporting for duty.
Same four stages, every time. If we skip one, we usually pay for it later in the video.
What it did before it died, and what the owner already tried.
Schematic, multimeter, scope, thermal camera. Measure before touching anything.
Hot air, microscope, and the actual technique — filmed close enough to copy.
Works, or it does not. Cost of parts against the cost of a replacement.
The groundwork. Watch these first and the teardowns stop looking like magic.
Short answers about repair scope, location and diagnosis.
We repair automotive electronics, graphics cards, consoles, laptops, appliance control boards, power supplies and other electronics where board-level diagnosis makes sense.
Not Dead Yet is based in Salzburg, Austria and focuses on board-level electronics repair and repair documentation.
Yes. Graphics cards and consoles are typical repair candidates, especially no-display faults, power rail issues, HDMI faults and short circuits.
Diagnosis is currently listed as free on the homepage. Parts are charged at cost if a repair is possible and wanted.
Yes. Automotive electronics such as instrument clusters, control modules and vehicle display boards can be repair candidates when board-level diagnosis is possible.
Got something written off by a service centre? Tell us what it is and what it did. Interesting failures get filmed, and you get the board back either way.
We reply to everything, usually within a week.