INTRODUCTION: This is one of the most serious damages to mechanical hard drives. SSDs and other Flash drives are not affected by this type of damage, as they do not contain heads or other moving mechanical parts.

CAUSE: It is usually due to mechanical impact on the drive (falling to the ground, generally hitting an object, or striking in some way, e.g., punching). It can also be due to: manufacturing defect, poor quality materials, extreme operating temperatures, wear over time, surface defects causing head damage, defective PCB.

SYMPTOMS: Clicks of Death. In some cases, if head damage only involves one or a few heads that are not the system heads (usually 0 and 1), the drive may not produce Clicks of Death but may “freeze” when trying to read.

LABORATORY TREATMENT: The drive must be opened and its surfaces (platters) inspected. Then, microscopic inspection of the heads is absolutely necessary. If they are clean, the heads are replaced with compatible ones, and the drive is set for cloning, following procedures that help stabilize it during surface reading. If they are not clean, this means there is damage on the surface corresponding to that head, as the dirt visible on the heads microscopically is actually magnetic material scraped from the drive’s surface. In this case, the platters must be removed, the damage identified, and appropriately assessed by a very experienced data recovery engineer who will devise a plan for partial data recovery if possible.

DIY COMPLICATIONS:
1. Head damage often leads to surface damage, especially if the drive’s owner repeatedly powered the drive, hoping it would work even for a little while. This repeated power supply can cause the damaged heads to destroy/scrape the drive’s surface, causing irreparable surface damage and data loss.
2. Not all heads fit all drives, but the opposite: for the heads to fit and read the surfaces with the data, they must be fully compatible. Often, this is not easy, and the wrong choice of parts can cause problems in the process, as we stress the drive unnecessarily.
3. Head replacement is a difficult process requiring great precision and extensive experience; it is particularly demanding for specific drive models.

DIY RISKS:
1. While the process may seem easy, it is not at all. At each step, there are risks that can cancel the process and significantly reduce recovery chances, even for professionals. These risks diminish only after many years of practice.
2. Hard drives must be opened in a Clean Room, not an uncontrolled environment. See why here and see what happened to a drive opened in an uncontrolled environment and then powered.
3. In the common scenario where head replacement fails, and the “new” heads are already damaged as they were mishandled during the DIY process, the drive is powered, causing the damaged heads to rise above the surfaces and destroy them.
4. Surface contamination from poor workmanship and carelessness, e.g., fingerprints, scratches, screwdriver slips, etc.
5. Head replacement is often the easiest part of the recovery process, although, as stated, it is already complex. Even if the head replacement is successful, the drive must then be connected to special equipment to help its parameterization for cloning the desired parts. There is a rule followed in these failure cases: Each sector must be read ONCE, during the drive’s cloning. There may not be a second chance.
6. Often, a noisy hard drive is opened for visual inspection by its owner to determine “where the noise is coming from.” This leads to the fatal decision to power the drive while the lid is removed. This results in the drive becoming unrecoverable due to severe surface damage. The reason for this is the fact that the drive lid has a dual role: to keep the drive’s interior sealed/clean and to maintain stable internal pressure. This pressure creates a microscopic air layer on which the heads fly above the platters. If the lid is open, the pressure becomes equal to the atmospheric, the air layer cannot form, and the heads cannot fly, causing them to fall onto the surfaces and scrape them.

INTRODUCTION: The definition comes from the term “STUCK,” which describes the state where the heads, instead of “parking” where they should (either on the parking ramp designed for this purpose or in the special area towards the motor axis, known as the “parking zone”), have stuck to the disk surfaces. As a result, the drive cannot rotate the platters because the heads are stuck on them, and the drive makes the characteristic sound of a stuck motor.

CAUSE: There are two possible scenarios for the cause of the failure: 1. Drop/impact of the drive and 2. Sudden pulling of the USB cable while the drive is in operation.

SYMPTOMS: Buzz! The drive audibly attempts to rotate the platters but fails. The drive will not be detected by the system.

LABORATORY TREATMENT: The drive is opened in the Clean Room, and a special technique is used to return the heads to their position. Subsequently, it is absolutely necessary to remove and examine the heads microscopically for alterations and damages. This is done because if the stiction is due to the drive falling or being impacted, the likelihood of damage to the heads is enormous, and they should not be used again on the drive to avoid potential surface damage. If the heads appear fine under the microscope, they can be used with caution: Since they have already touched the surfaces, there is a potential for damage both to the heads and locally to the surfaces.

In any case of doubt, the heads are immediately replaced.

The drive is then set for cloning as usual.

DIY COMPLICATIONS:
1. If the stiction was caused by a fall, there is a serious chance of head damage, as explained above.
2. The coating film of the surfaces (platters) in some hard drive manufacturers consists of a particularly sticky material, causing the head sliders to stick intensely to the platters, dramatically increasing the likelihood of their destruction. It is common to observe sliders stuck to the surfaces that must be removed using a special technique.
3. Repeated attempts to detach the heads by re-powering the drive can have unpleasant consequences for both the heads and the drive’s surface, making data recovery nearly impossible. A frequent phenomenon is the physical detachment of the slider after persistent attempts by the drive to rotate the platters, causing internal surface damage. In a more extreme phenomenon known as wobbling, persistent attempts to rotate the platters by the drive owner re-powering the drive cause the drive shaft to warp, making the platters rotate with “waves.” This results in axis destabilization, strong system vibrations that can destroy the heads and surfaces, and a risk of loss of both alignment and the servo marker.

DIY RISKS:
1. Moving the heads to their position must be done with a special technique that ensures zero further damage to the surface they have touched and maintains the heads’ integrity if they exist.
2. The drive must be opened in a controlled environment, and the technique must be followed by an experienced data recovery engineer. Otherwise, surface contamination with dust and particulates will occur, and if the drive is powered in this condition, it will be unrecoverable.
3. At every stage of the process, there is a risk of causing damage to the exposed drive surface, e.g., fingerprints, scratches, screwdriver slips, etc.

INTRODUCTION: This is a mechanical failure related to the motor that spins the platters. The diagnostic process is not easy, as the symptoms are exactly the same as stiction, since both failures share the common characteristic that the drive tries to spin the motor but fails. This is simultaneously one of the most difficult failures in data recovery.

CAUSE: There are two possible scenarios for the origin of the failure: 1. SEVERE drop/impact of the drive and 2. Defective lubrication of the motor during its assembly at the factory or a poor-quality motor.

SYMPTOMS: Buzz! The drive audibly attempts to spin the platters but fails. The drive will not be detected by the system. The symptoms are exactly the same as stiction, and there is often confusion.

LABORATORY TREATMENT: This is one of the most difficult failures to handle, as it almost always involves transferring the platters to a new chassis.
Empirically, there are 3 possible solutions:
1. Transfer the platters to another, similar chassis with a healthy motor,
2. Unblock the motor with a special key, and
3. Replace the motor with a special setup.

DIY COMPLICATIONS:
1. Transferring the platters to another chassis is an extremely risky process that requires delicate and skillful handling, with a high likelihood of failure, so it is chosen even by professionals only if there is no other solution. The platters must obviously remain completely intact, and in most cases, the alignment between them must be maintained, which is factory-set with nm (nano-meter) accuracy. Any misalignment will lead to a phenomenon called platter eccentricity at best, or loss of the servo at worst, making recovery impossible if there are no servo markers on each surface.
2. The heads must be replaced de facto.
3. In the case of drives with spacers between the platters, the process’s complexity skyrockets.

DIY RISKS:
1. Dropping or damaging the platters in any way means automatic loss of all the data contained on the drive.
2. Often it has been observed that a platter is mistakenly installed upside down. In this case, the drive is unrecoverable.
3. At every stage of the process, there is a risk of causing damage to the exposed drive surface, e.g., fingerprints, scratches, screwdriver slips, etc.
4. The drive must be opened in a controlled environment, and the technique must be followed by an experienced data recovery engineer. Otherwise, surface contamination with dust and particulates will occur, and if the drive is powered in this condition, it will be unrecoverable.
5. The heads must be replaced with compatible ones, which, as described above, is a particularly complex process.

INTRODUCTION: A burnt PCB (Printed Circuit Board) is the most common electronic failure of hard drives.

CAUSE: It is usually due to incorrect power supply. A classic scenario is when an external drive (normally powered by 12V) is mistakenly powered by a laptop (powered by 18-20V), causing the drive to burn, often accompanied by smoke. Other scenarios of burnt PCBs include the presence of water/other liquid/moisture on the drive (e.g., accidental spilling of a glass of water or other liquid on the external drive or laptop), defective drive manufacturing at the factory, defective board in the external case, wear over time resulting in contact oxidation, etc.

SYMPTOMS: The drive is completely dead and does not spin up.

LABORATORY TREATMENT: It must be ensured that the cause of the PCB failure (overvoltage?) did not destroy the head preamplifier [see “Complications” below]. This is done by measuring the preamplifier’s control points with a multimeter or Logic Analyzer of the signals. Once we confirm that the preamplifier is fine, the board is replaced with a compatible one, and the unique adaptive data [see “Complications” below] is transferred to the donor board. The drive is then set for cloning.

DIY COMPLICATIONS:
1. The vast majority of modern drives use the “adaptive” data philosophy, resulting in the PCBs being “paired” with the hard drive. There are extreme implementations of this logic, with manufacturers whose boards are completely unique (e.g., Seagate, Toshiba, some WD models) and less strict implementations (Samsung, some WD models) where the boards can be recreated even if they are completely burnt or lost.
2. Direct PCB replacement from the same exact drive is not applicable to modern drives. It was applicable to very old drives, but not anymore. The transfer of adaptive data to the donor board is necessary; otherwise, the drive will not function, resulting in clicks of death.
3. The preamplifier is a microscopic device on the drive’s heads, amplifying the signal from the heads to the PCB, allowing the drive to communicate with its various parts. For example, without the preamplifier, the PCB (controller, to be precise) would send a signal that—for instance—the spinup sequence completed successfully, and the heads should “unload” onto the platters to read the Service Area, but the heads would “not hear” the information. If surplus currents pass into the drive, the preamplifier burns out, causing the drive to not function even if the electronic issue is resolved.

DIY RISKS:
1. The risk exists when the preamplifier is burnt, but corresponding measures (head replacement) are not taken, and the electronic failure (PCB replacement) is resolved. In this case, there is a serious chance the drive will zap the Service Area, making the drive unrecoverable or creating a recovery labyrinth.
2. In many cases, the chips containing the unique data that must be transferred to the donor board are unique. The transfer process usually involves the use of a soldering iron or hot air gun, increasing the chances of chip damage. If the chip is damaged due to poor workmanship or overheating, recovery becomes impossible or extremely complex (and expensive). Modern recovery labs do not use this technique to avoid the possibility of such an error.
3. Using an incompatible board will lead to new destructive paths, as if the board is not compatible or the adaptive transfer is not done correctly, the drive will not spin up, which in the DIY world almost always means opening the drive, something that never had a pleasant outcome.

INTRODUCTION: This is one of the most common failures in SSDs. Since SSDs consist solely of circuits, the currents passing through these circuits must be stable; otherwise, various components on the board may burn out.

CAUSE: It is usually due to incorrect power supply or defective board construction.

SYMPTOMS: The drive is completely dead, with no activity, and is not recognized by the computer.

LABORATORY TREATMENT: The board is analyzed to identify which part is shorted or burnt out and is then repaired.

DIY COMPLICATIONS: 1. In many cases, the shorted part is the drive’s controller, which cannot be repaired or replaced, making the drive unrecoverable.

DIY RISKS: 1. There are board components unique to each drive, e.g., the NAND chips, and often the controller encrypts data for each drive separately. Damage to these components due to carelessness or poor workmanship means the drive will be 100% unrecoverable.

INTRODUCTION: This is one of the most common failures in data recovery, affecting all manufacturers. The firmware, or microcode, is the “heart” of the drive. It is a hidden section where the drive’s code is written, based on which all drive functions are performed. It includes the translator, defect lists, SMART, factory logs, recording zones, and more. In the data recovery language, this is called the Service Area (SA) and is a vital part of the drive, without which it cannot function. In the vast majority of manufacturers, the SA is written on the platters in a special area inaccessible to the end user, except for Toshiba, where most of the SA is written on the PCB [*1].

Writing the SA on the platters “carries” all vulnerabilities that platters may have. For example, surface damage to the platters means the SA cannot be read, as well as damaged or failing heads. On the other hand, writing the SA on the PCB offers the advantage of faster SA reading and almost instant boot of the drive. However, the boards are particularly vulnerable to power instabilities, and any fluctuation can render the drive unrecoverable.
[*1] Toshiba abandoned this logic in its more modern SMR drives (e.g., MQ04xxxxx), and now almost the entire SA is written on the platters.

CAUSE: Often, firmware failure may practically result from mechanical failure, with damaged or failing heads unable to read the drive’s firmware section, causing the drive to exhibit firmware failure symptoms, though the actual cause lies elsewhere.
Usually, firmware failure results from a drive malfunction that has led to corruption of the firmware section necessary for the drive’s proper operation.
Firmware failure can also result from “normal wear” of the drive. A classic example is a WD drive with surface defects known as bad sectors due to wear. The drive’s SA identifies the problem and starts re-allocating the bad sectors to another area to preserve data integrity, recording the action in the corresponding defect list. If this failure extends, eventually, the defect list fills up, and the drive cannot continue functioning, leading to extremely slow operation despite being functional. The defect list is full (relo failure). [4]

SYMPTOMS: The drive may sound normal but not be recognized by the system, make eccentric noises unlike the click of death, or be recognized by the system after a while and respond extremely slowly.

LABORATORY TREATMENT: A detailed diagnostic check is conducted, access to the SA is obtained, the SA is analyzed for damages, and a knowledgeable data recovery engineer manually repairs specific sections or replaces mechanical parts if such underlying damage is found. The drive is then set for cloning.

DIY COMPLICATIONS:
1. In example [4], the classic mistake (unfortunately even by some who call themselves “professionals”) is resolving the relo and attempting to read the drive. This analogy is like sweeping dirt under the rug: The relo appeared for a reason and somehow formed. The cause must be identified and repaired. Otherwise, we will face the phenomenon where the relo appeared from defective heads unable to read surfaces correctly, filled the relo with errors that do not exist, the relo was cleared by uninformed individuals, the same heads were called to read, were completely destroyed, taking the surfaces with them. Then, knowledgeable people are called to solve it. This is an almost daily phenomenon for quality data recovery labs.
2. Managing firmware/SA failures REQUIRES DEEP KNOWLEDGE of data recovery and drive microcode structure and SHOULD NOT be done by uninformed individuals, as it will ALMOST CERTAINLY lead to unpleasant paths.
3. Often, attempts to “flash” drives with files from various websites or even manufacturers’ official sites are observed. It must be clear that hard drives DO NOT work like mobile phones for updates or like graphics cards (for example) requiring updates to the latest drivers. This will lead the drive to a “brick” state, needing an engineer with serious knowledge to resolve it, if possible. Firmware update publications on manufacturers’ websites are for SPECIFIC purposes and not for all users. They should only be done following an explicit request from the manufacturer, such as in the case of Seagate with the Barracuda 7200.11 with firmware SD15, which had the notorious BUSY BUG issue, and Seagate asked everyone to “flash” their drive to version SD1A) AND ONLY THEN.

INTRODUCTION: This is one of the most common failures in data recovery related to SSDs. SSDs have the SA written in the NAND memories, while the whole process is supervised by the CPU (the Controller).

CAUSE: It is usually due to the low quality of NAND chips but, in general, it is due to the Flash technology they use: when data is written to a NAND cell, to rewrite data in that cell, the previous data must be erased.

This writing and erasing of data is done by sending electrons to the cells (to put it very, very simply). These electrons pass through an insulator. The position and location of these electrons determine whether current will flow to the cell or not and whether this cell is occupied.

When we write and erase data from a cell, as the electrons go back and forth, the insulator we mentioned wears out to the point where it struggles to keep the electrons in the place they should be, resulting in leakage or/and inability to determine whether the electrons are where they should be.

This, in turn, explains the “finite number of write cycles” we often mention. It means that the Flash technology itself has a limitation on the number of read and write cycles of data.

The wear of NAND cells will lead to ECC errors and very often to the drive’s translator, as it is written on the NAND chips.

SYMPTOMS: The drive does not make ID and is not recognized by any system, or makes ID but then goes into a BUSY state after any command. It will never exit the BUSY state, or if it does, it will re-enter the same state after any command.

LABORATORY TREATMENT: If the controller is supported, a virtual loader is created, and the drive starts from there. Then, the translator is “rebuilt” from scratch. Essentially, the usual boot sequence of the drive is bypassed.
If the drive has degraded its NANDs and ECC Errors appear, the result can partially improve with persistent re-reads, but a good result concerning the quantity and quality of data will be rare.

DIY COMPLICATIONS:
1. As understood from the above, the handling of these drives should be done by a recovery engineer with very good knowledge of how these drives work and, of course, having the appropriate equipment. Any deviation from this will have disastrous consequences for the data.
2. Using various software freely available on the internet may sound tempting for what they offer, but they essentially DELETE the data and return the drive to a functional state. If the failure is, for example, in the translator, then the restoration will be successful, BUT ALL DATA WILL BE PERMANENTLY LOST. If the failure is, for example, in the NAND (ECC errors, for instance), then these actions will obviously have no effect.

INTRODUCTION: This is an almost daily occurrence: accidental deletion of a file or folder, or formatting the wrong drive or partition.

CAUSE: There is an old saying that goes “most computer problems are caused by a loose screw between the chair and the keyboard.”

SYMPTOMS: Your files, folders, partition, or drive are no longer accessible.

LABORATORY TREATMENT: Serious data recovery labs have higher quality software compared to those available on the market. However, many software options out there can recover deleted data, provided no new data has been written over it.

DIY COMPLICATIONS: 1. There are some very low-quality free software options, which may recover non-functional or raw state data. Many users, not knowing the basics, transfer these recovered files to the same drive where the “problem” occurred. This will result in overwriting old “good” files with new, problematic ones, making proper structured recovery impossible.

INTRODUCTION: The most common scenario for this case is installing the operating system on the wrong drive. “I wanted to format C: and put fresh Windows there, but I accidentally formatted D: where I had all my files, and by the time I realized, the installation was complete / I interrupted it midway,” we hear this daily.

CAUSE: The same saying as in 4a.

SYMPTOMS: Your files have been lost, and new ones have been written over them.

LABORATORY TREATMENT: The disk is scanned with higher-quality software, with parameters set by a data recovery engineer with the relevant knowledge. The files overwritten by the Windows installation (about 50GB) are permanently lost. The remaining 750GB might be recoverable with structured recovery.

Example-1: On a 1TB HDD containing 800GB of data, format and install Windows. The installation is completed normally. The files overwritten by the Windows installation (about 50GB) are permanently lost. The remaining 750GB might be recoverable with structured recovery.

Example-2: On a 1TB HDD containing 20GB of data, format and install Windows. The installation is completed normally. In this case, the recovery will fail, and only traces of the data will be recovered.

Example-3: On a 1TB SSD containing 900GB of data, format and install Windows. The data will not be recovered as SSDs have TRIM, and recovering deleted data from a drive with TRIM enabled is currently impossible.

DIY COMPLICATIONS: 1. Same as 4a.

INTRODUCTION: The most common scenario for this case is searching for older dates in a camera recording system (DVR/NVR). The need for this usually arises when it is discovered that something has happened in the monitored area (e.g., embezzlement of money from a cash register, etc.).

CAUSE: The problem lies in the fact that DVRs are set to record for a specific period (sometimes mandated by law) and once this period passes, they start over, overwriting older dates.

SYMPTOMS: A specific date is sought that no longer exists in the DVR.

LABORATORY TREATMENT: Serious data recovery labs have higher quality software compared to those available on the market. Based on the DVR manufacturer and the recording and encryption algorithm used, there may be a slim chance—and with some luck—to recover some footage. However, in the vast majority of these cases, the footage has been overwritten and, according to 4b, its recovery is impossible.

DIY COMPLICATIONS: 1. The only complication here, provided it is ensured that no new recordings will be made on the examined disk, is the waste of time.

INTRODUCTION: The most common scenario for this case is composing a text, accidentally deleting part of the text, and saving the file over the previous saved version.

CAUSE: The same principle as in 4a.

SYMPTOMS: You open the file where you were working on your assignment and see it incomplete.

LABORATORY TREATMENT: A search will be conducted for temporarily saved files by the text editor of the file in question. If none are found, a search will be conducted for all files of that type on the disk, using search criteria based on the metadata of the specific file.

DIY COMPLICATIONS: 1. Same as in 4a.

INTRODUCTION: The most common scenario for this case is the “lazy” method of transferring files via cut and paste. For many reasons, this process can fail, resulting in the loss of files from the source.

CAUSE: Due to a momentary malfunction of the computer, operating system, or disk. Due to user error. Due to mechanical failure of the disk.

SYMPTOMS: The source folder containing the files you wanted to transfer is now empty.

LABORATORY TREATMENT: The methodology for deleted files is followed, as described in 4a.

DIY COMPLICATIONS: 1. Same as in 4a.

INTRODUCTION: The most common scenario for this case is the compulsive nature of humans choosing to empty the computer’s recycle bin or the operating system itself automatically emptying the bin to save space and because it is configured that way.

CAUSE: Due to the compulsive nature of humans and the operating system settings.

SYMPTOMS: The recycle bin where the user relied on to find a file deleted 3 months ago and finally remembered they need it is no longer there.

LABORATORY TREATMENT: The methodology for deleted files is followed, as described in 4a.

DIY COMPLICATIONS: 1. Same as in 4a.

INTRODUCTION: In order to keep the drives in good condition and to “fight” with the diseases of the flash technology itself, manufacturers invented the TRIM function.

This is a technology embedded in SSDs to keep cells in good condition. Essentially, when a file is deleted from an SSD, the TRIM function ensures the drive knows that the cells corresponding to this file are no longer in use and shows them as zeroed.

As a result, when files are deleted, formatted, etc., the files are unfortunately lost forever and cannot be recovered, at least not by conventional means.

CAUSE: Due to the Flash technology on which SSDs are based. Flash memories constantly degrade with every action, even if it is a simple data write. This is why manufacturers incorporated TRIM into the drives, to keep the cells in the best possible condition.

SYMPTOMS: After formatting, the drive appears zeroed out, even at the Hex level.

LABORATORY TREATMENT: A specific methodology is followed to take advantage of the entries in the drive’s translator, essentially ignoring the TRIM. Under certain conditions, this way we can see the data that existed before deletion.

DIY COMPLICATIONS: 1. Even experienced users who have recovered deleted files with software in the past are surprised by the poor results in the case of SSDs. There is no way to recover data at home; complex procedures and methods requiring equipment and knowledge must be applied.

DESCRIPTION: In the case of flooding, data recovery depends on several factors. The most important of these is whether the drive was powered or not. This is because, in the case of flooding, it is assumed that water has entered the interior of the drive (through the filter) along with particles of mud, etc. The water contains various particles that should not be present inside the drive (metal salts, etc.), while seawater (e.g., in the case of a laptop falling into the sea) is highly corrosive and immediate measures must be taken by knowledgeable people.

LABORATORY TREATMENT: In any case, the drive must be opened in a controlled environment, and its interior must be cleaned meticulously. Usually, in these cases, it is preferred to transfer the platters outside the chassis, thoroughly clean them, and transfer them to a new chassis along with all the peripherals (heads, ramps, magnets, filters, etc.) which must also come from another donor drive.

DIY COMPLICATIONS: 1. Flooding is usually accompanied by a short circuit of the board, resulting in the drive not being powered. This is partly good, as the drive is protected from destructive actions such as powering it while its interior is contaminated. Unfortunately, many people think that “the drive is simply burnt,” replace the board, and power it, resulting in the physical destruction of the surfaces inside the drive, making it 100% unrecoverable.

DESCRIPTION: Empirically, in these cases, the burglar identifies a closed-circuit recording system (DVR/NVR) and either destroys it or takes it with them.
In the first case, the burglar might remove the disk from the recording system and vandalize it to destroy evidence. In these incidents, it is clear that the success of data recovery depends on how “good” a job the perpetrator did. We have encountered an incident where the cover of the disk was opened and the entire upper surface was scratched with a screwdriver, but this surface was actually not used by the disk, so we were able to recover 100% of the recordings.

LABORATORY TREATMENT: In any case, the disk must be thoroughly examined for its condition. If there is hope, all the steps described in the previous chapters are followed, depending on the damage.

DIY COMPLICATIONS: 1. The same applies as in the corresponding chapters describing the damages.

DESCRIPTION: In the case of a fire, data recovery chances are usually small. Everything depends on whether the fire reached the interior of the drive or not. There are many examples where the fire burned the computer but stopped at the metal casing of the hard drive, resulting in the data being recoverable after appropriate actions.

LABORATORY TREATMENT: It is very important to examine the magnetic properties of the drive. If the fire raised the internal temperatures of the drive beyond the Curie point, the magnetic properties of the drive are permanently lost, and therefore the data is also lost.
Older drives used magnetic coating on the (non-magnetic) platters from iron oxide, which loses its properties at ~900°C, but other materials in the drive like neodymium magnets lose their properties earlier, at about 500°C.

Modern drives usually use a magnetic coating film that consists mainly of a cobalt-based alloy, which has a higher Curie point (around 1050°C). However, it is believed that the magnetic properties of the hard drive are lost much earlier, specifically at 400°C.

DIY COMPLICATIONS: 1. A data recovery company is looking to hire anyone who has successfully recovered data from hard drives damaged by fire :)