The short answer: If a drive, computer or NAS won’t start after a storm, smells burnt, or has disappeared from the system, unplug it from power and stop trying to revive it. A drive hit by a power surge usually has burnt electronics, but the magnetic recording on the platters normally survives the hit — what actually costs you the data are the attempts that follow: a donor PCB, repeated power cycling, or connecting the drive to a damaged power supply. Swapping in an „identical" board doesn’t work, because a modern drive keeps adaptive data in ROM that is unique to that specific unit. Have the drive diagnosed before your neighbour or the nearest walk-in shop starts experimenting on it.
A summer thunderstorm is the most common scenario behind the drives with burnt electronics that arrive at our lab. The story is always the same: a storm rolls through overnight, in the morning the computer won’t power on, the NAS is dark, the external drive looks dead. Surges reach your equipment through the mains, but with a NAS or router they also travel along the Ethernet cable or the antenna downlead. This article explains exactly what burns out inside a drive, why the popular advice to „buy the same drive and swap the PCB" is more likely to finish the data off, and what a safe route to recovery looks like.
What a surge actually burns out in a drive
The voltage spike travels from the wall socket through the computer’s power supply or the external adapter into the drive’s power rails. There it meets the first line of defence: the protective TVS diodes on the power input. They are designed to act as a fuse — under a surge they break down into a short circuit and sacrifice themselves so the spike goes no further. When that works, the drive looks dead (the shorted rail prevents it from starting), but the rest of the electronics is usually fine.
A stronger jolt, or repeated hits, will get past them. Then the damage lands on the motor driver, the controller, or the area around the ROM chip on the PCB. In the worst case the surge reaches the sealed chamber and burns the preamplifier on the head stack — that sits right next to the heads inside the drive, and damage there means a mechanical intervention in a clean environment, not a board repair.
The important good news: data is stored magnetically on the platters and an electrical spike does not harm it by itself. So after a surge you are usually not dealing with destroyed data, but with a destroyed electronic path to it. Which is exactly why it would be a shame to lose it to a careless repair attempt.
The first few minutes after the storm
- Disconnect the affected device from power. With a NAS or router, pull the Ethernet cables too — the surge may have arrived over the network.
- Don’t keep switching it on. Partly damaged electronics can degrade further with every start and take down whatever has survived so far.
- Don’t connect the drive to a computer or power supply that took the same hit. A damaged PSU can feed the drive the wrong voltage and finish off a unit that survived the storm.
- Document the event for your insurer. Date and time of the storm, photos of the equipment, receipts if you have them. The diagnostic report is useful for an insurance claim.
- Write down what matters on the drive. Accounting, projects, family photos — the technician then verifies the important things first.
How to decide based on specific symptoms (the drive won’t spin up, it clicks, it reports the wrong capacity) is covered in detail in HDD data recovery: what to do based on the type of failure. With a surge-damaged drive, though, the decision is simpler than with other failures: if the drives hold the only copy of important data, send them straight through data recovery diagnostics — it’s free for HDDs, SSDs and RAID alike, and it tells you exactly what burnt out and what the recovery will involve before you pay anything.
Why buying the same drive and swapping the PCB isn’t enough
The most widespread piece of internet advice for a burnt drive goes like this: get the same model, unscrew the board, swap it over, done. Fifteen years ago that occasionally worked. On modern drives it’s a trap, for two reasons.
Adaptive data in ROM. Every drive goes through individual calibration during manufacturing: read head parameters, compensation for production tolerances, defect maps, servo parameters. These „adaptives" are stored in the ROM on the PCB and are unique to that specific unit — two drives of the same model from the same box have different ROM contents. A donor board with foreign adaptives will either fail to start the drive at all, or the drive comes up and reads nonsense; repeated starts with the wrong calibration also put needless strain on the heads flying above the platters.
The same model number doesn’t mean the same hardware. Manufacturers change board and firmware revisions during production without changing the model name. Even an „identical" drive bought six months later can have incompatible electronics.
The correct procedure therefore looks different: the ROM is transferred from the original board (by desoldering the chip or reading out its contents) to a compatible donor board, shorted TVS diodes and damaged circuits are checked, and only then — with working electronics and the original adaptives — is the drive powered up safely for the first time. That is a job for a microsoldering station and service equipment, not a screwdriver on the kitchen table.
When the surge goes deeper than the board
If the drive won’t come up even with repaired electronics and a transferred ROM, the hit went further in — typically to the head preamplifier or the motor. At that point it is no longer a board repair but a mechanical recovery: opening the drive in a clean environment and replacing the head stack with one from a donor drive.
To give you a sense of the numbers: data recovery with us starts at CZK 2,900 (typically electronics and logical failures), while mechanical work involving opening the drive comes to roughly CZK 18,900. Diagnostics are free and we work on a no recovery, no fee basis — if we can’t recover the data you asked for, you don’t pay for the recovery. You get an exact quote after diagnostics, not over the phone; nobody can honestly tell from „it died after the storm" whether only a protection diode sacrificed itself or the preamplifier is gone.
NAS after a storm: the hit also arrives over the network cable
A NAS is more vulnerable in a storm than a computer, because two paths lead into it: power and Ethernet. A surge induced into the network wiring can destroy the NAS network section even when the power went through surge protection.
Most often a storm takes out the NAS power supply or adapter, and the drives inside are fine. The worse scenario: the spike passes through the PSU and hits every drive at once. That is a situation RAID cannot handle — an array is designed for drives failing one at a time, not for several units being damaged in the same second.
So after a storm, stricter rules apply to a NAS:
- Don’t try the drives in another NAS, and don’t buy „the same box" to test them — the new system may write to the drives and overwrite the array metadata.
- Don’t start a rebuild or a volume repair if the NAS partly boots and reports a degraded array.
- Don’t connect the drives to Windows — the „initialize disk" prompt is a trap that rewrites the partition tables.
- Label the drives by bay if you take them out.
The complete safe procedure, including an explanation of why Windows doesn’t understand NAS drives, is in the article on how to get data off a NAS when the NAS dies.
External drives and SSDs after a surge
With an external 3.5” drive, the power adapter stands in the spike’s way — and it is often the adapter that sacrifices itself, or the enclosure electronics. The drive inside is then usually untouched. Don’t buy a replacement adapter blindly, though: voltages and pinouts differ, and the wrong adapter can damage the drive more reliably than the storm did. And watch out for WD portable drives: on many of them the encryption is handled by the USB controller in the enclosure, so „I’ll take the drive out of the box and connect it directly" gives you nothing but unreadable garbage. What to do when an external drive doesn’t show up is covered in the article on an external drive that won’t load or isn’t visible.
SSDs come off worse after a surge than their reputation as rugged drives would suggest. The spike hits the power circuitry and the controller; the NAND chips holding the data often survive, but without a working controller the only route to them is the difficult one — reading the chips directly — and with proprietary controllers sometimes there is no route at all. Why recovery from an SSD is a discipline of its own is explained in the article on SSD data recovery.
How diagnostics of a surge-damaged drive work
Diagnosing a burnt drive doesn’t start by powering it on, it starts with measurements. First a visual inspection of the board (scorched components, dark traces), then measuring for shorts on the power rails and checking the TVS diodes and fuses. Only once the electronics is safe — repaired, or replaced by a donor board with the transferred ROM — is the drive powered up for the first time, and that happens on service equipment that can control the read process and stop it instantly if trouble appears.
The goal is not to repair the drive back into service, but to create a sector-by-sector image and hand over the data from it on a healthy medium. Don’t put a surge-damaged drive back into service, even if it looks healthy after the electronics repair — it has been through electrical stress that may show up later.
Prevention: surge protection, a UPS and backups
The only hundred-percent protection during a storm is physical disconnection — pulling both the power and the Ethernet cables out. Everything else is risk reduction:
- Multi-stage surge protection. Arresters in the distribution board absorb the raw energy, a surge-protected power strip cleans up the rest. A cheap strip on its own, without protection in the distribution board, won’t stop a big hit, and after a single strike it is often used up without showing it.
- A UPS for the NAS. A backup power supply mainly protects against power cuts, which a NAS handles badly — unfinished writes after an outage are a frequent cause of volume failure. Better UPS units also filter the power. A UPS doesn’t replace surge protection, though; it complements it.
- Watch the wires coming in from outside. Outdoor Wi‑Fi antennas, PoE cameras and long Ethernet runs are entry points for induced surges; they belong behind their own protective devices.
- Backups following the 3‑2‑1 rule. A storm is the classic event that takes out the computer and the external backup plugged into the same socket at once. That’s why at least one copy of your data has to exist disconnected or off-site — how to set that up in practice is described in the article on an external HDD as a backup and the 3‑2‑1 rule.
Frequently asked questions
Does data on a drive survive a lightning strike?
Very often, yes. A surge destroys the electronics, but the magnetic recording on the platters stays intact. Data is lost afterwards — to a donor PCB without a ROM transfer, to repeated power cycling, or to connecting the drive to a damaged power supply.
Can I swap the electronics for a board from an identical drive?
Not without transferring the ROM. The board carries adaptive data unique to that specific unit; with foreign adaptives the drive won’t start, or it reads incorrectly. The correct procedure is to transfer the ROM from the original board to a compatible donor and repair the damaged protection circuits.
The computer runs after the storm but the drive doesn’t — what now?
Disconnect the drive and stop testing it in that computer. The PSU may have taken a hit even though the computer boots, and a damaged PSU can finish the drive off with the wrong voltage. Have the drive go through diagnostics; ours are free.
Will a cheap surge-protected power strip help?
Only to a limited extent. It works as a final clean-up stage behind arresters in the distribution board; on its own it won’t catch a big hit, and after a single strike it is often quietly used up. A NAS also needs its Ethernet feed protected, and a power strip covers nothing there.
How much does data recovery after a surge cost?
It depends on how deep the damage goes. Recovery starts at CZK 2,900, while mechanical work involving opening the drive comes to roughly CZK 18,900. Diagnostics for HDDs, SSDs and RAID are free and we work on a no recovery, no fee basis — no recovered data, no charge for the recovery.