Cleanroom Head Replacement and Donor Matching
Head-stack replacement is precision, donor-dependent work done in a particle-controlled work zone, because at the flying height of a modern head ordinary dust is enormous by comparison. A controlled environment, a genuinely compatible donor and controlled acquisition afterwards are three separate conditions — and even together they make partial read access possible, not guaranteed.
- What matters is verified particle control in the actual work zone; an ISO class describes air cleanliness, not lab competence or recovery success.
- Same model number is only the start of a donor search; family, actual head configuration, adaptation, tool fit and donor condition are separate checks.
- A successful swap can restore part of the read access; it cannot rebuild a recording layer that is already damaged.
- A disk that spins after a swap proves neither the right donor nor healthy platters; repeated donor attempts are not an unlimited solution.
- Plan around whether the areas holding the needed files are readable, not around whether the drive starts.
A room that looks clean is not a controlled work zone
The reason internal HDD work needs a controlled environment isn't cosmetic. At the flying heights heads normally operate at above the platter surface, an ordinary dust particle is enormous by comparison — large enough to lodge under a slider, score the platter, or create debris that spreads across multiple surfaces.
A whole cleanroom, a clean zone inside a larger room, and a laminar-flow clean bench are different architectures. What matters is that the particle control is verified for the zone where the drive is actually open, and that the way people work in that zone keeps it that way. ISO 14644-1 classifies air cleanliness by the concentration of airborne particles, and labs often describe their work zones by such a class; ISO Class 5 is commonly cited for this kind of work. A particle class, however, is a property of the air in a zone, not a certificate of the technicians' competence, of the lab's donor matching, or of the recovery outcome. Nor should any particular class be read as a universal regulatory requirement for data recovery.
Opening the drive is a diagnostic step — with limits
Once an HDD is opened under controlled conditions, a technician can inspect for visible head damage, platter scratches, contamination or debris, and what is found can change the plan. Visible scoring or debris changes the prognosis and the acquisition strategy; visible contamination is a separate problem from a failed head.
Surfaces that look clean, though, do not make a head swap straightforward. Visual inspection cannot confirm the integrity of the magnetic layer, the servo information, the condition of every head, or which tracks will actually be readable. Those questions are answered, if at all, by what the drive can read after the work.
Why the label is not enough for a donor
Two drives sold under an identical retail model can be poor donors for each other, because the things that decide compatibility are not printed on the label. Matching is a set of separate questions, and their weight differs between drive families; there is no universal checklist or ranking. The specific parameters professional donor databases track are covered on HDD donor parts and compatibility; the table below explains the logic of each check.
Scroll horizontally to see the full table →
| Evidence | What is checked | Why the label is not enough | Limit |
|---|---|---|---|
| Family and revision | That patient and donor belong to the same internal family and a compatible revision | One model name can cover more than one internal design | Necessary in most cases, never sufficient on its own |
| Actual head configuration | How many heads and surfaces are physically present and in use on both drives | Drives with the same capacity label can be built with different head counts or configurations | A matching configuration does not mean matching head behavior |
| Relevant adaptation | Whether patient-specific adaptive or calibration data must be considered for this family, and whether it is available | Adaptive data belongs to the individual drive, not to the model | Need and available operations are family-specific |
| Physical and tool compatibility | Whether the donor assembly fits mechanically and whether the handling tool fits this family | Fit depends on internal mechanics that the label does not describe | A tool that fits does not answer whether the heads are compatible |
| Donor condition | That the donor's own heads and mechanics are in working order and have not been damaged or contaminated | A correct-looking donor can itself be faulty | A faulty donor can look like an incompatible one, or like worse patient damage |
Details such as manufacturing date, site or firmware version can help narrow the search in some families, but they are not a universal rank, and passing every check is not a guarantee. Tool developers make the same point from their side: one tool vendor notes having to deal with multiple drive families sharing the same model's name but requiring different tools.
Fit, tooling, compatibility and calibration are separate questions
Mechanical fit, tool fit, donor compatibility and calibration are often merged into a single “matching donor,” but each can fail independently. A head-change tool is designed for particular families and solves a handling problem; the developer manual for one such tool states plainly that it doesn't solve the head compatibility problem.
A donor head stack that fits and seats correctly can still differ from the original in gain, flying characteristics or calibration behavior. In some families recovery tools apply adaptive changes after a swap before the drive reads reliably; in others different operations, or none, are needed or available. It would be wrong to assume that every swap needs the same adaptation, or that any adaptation exists for a given family. Donor selection and whatever adaptation the family requires are planned together, but mechanical fit by itself does not guarantee readable calibration.
Possible states after targeted work
A replacement that completes technically can leave the drive in quite different states. These are possibilities to consider, not ready-made diagnoses:
Scroll horizontally to see the full table →
| State | What it may mean | Decision to consider | What it does not prove |
|---|---|---|---|
| Priority range readable | The donor heads read at least the area that holds the needed data | Acquire that range first, before broader testing | That the rest of the drive will read, or stay readable |
| Partial access, unstable head | One or more heads read intermittently or degrade during use | Limit work to priority ranges and surfaces that read; record which head and range each error belongs to | Whether the cause is the donor, the adaptation or the surface |
| Severe surface damage | Parts of the recording layer are already gone or are producing debris | Stop and reassess; continued reading can spread damage to heads and other surfaces | That another donor would read those areas |
| Dependent service area inaccessible | The drive cannot read service information it needs, so the user area stays unavailable | Treat as a separate firmware and service-area question (see HDD service area and translator recovery) | That the new heads are faulty, or that they are fine |
The trade-off between testing longer and acquiring earlier depends on the symptoms and on which data is needed. Extended testing tells more about the drive, but it spends operating time that a fragile head and damaged media may not have.
Training scenario: same-model donor, the disk spins, errors remain
Heads from a donor with the same model number were installed. The disk spins up and the drive responds, but read errors continue across large areas.
Spin-up and a response show that the mechanism runs. They do not show that the donor is the right one, that any family-specific adaptation is correct, or that the platters are healthy. Any of these could explain the errors.
A qualified reassessment of donor evidence, adaptation and surface condition, followed by controlled acquisition of priority ranges if the drive reads them.
“Just try another swap” is not a conclusion this observation supports; each attempt adds handling and operating time on damaged media.
Post-swap imaging and records
A drive that has just had its head stack replaced is not a drive returned to normal service — it's an acquisition instrument, and often a fragile one. Donor heads reading a patient's already-damaged media can themselves degrade quickly, especially over regions where the original heads left scratches or debris. Controlled, prioritized imaging (see imaging bad sectors and unstable HDDs and how to image a failing drive safely) still applies in full after a mechanical repair, and can matter even more because the repaired mechanism may have little operating life left.
The record of the work is part of the result. It should identify the patient and donor drives, state what was changed, which heads and ranges were read, where errors occurred, and what limits applied to the acquisition. The original parts and that evidence are kept. Repeated donor attempts are not an unlimited solution: each one adds handling, operating time and risk to the same surfaces.
What new heads cannot do
If a failed head already scratched the magnetic surface before replacement, the new heads meet that same damage from the first pass. A compatible, successful head swap can restore part of the read access; it cannot recreate a recording layer that is already destroyed. A completed swap does not guarantee that the full capacity is readable or that the files on it are intact. For planning the outcome, what matters is whether the areas holding the needed files can be read — not whether the drive starts.
“Same model number” is the start of a donor search, not the end of one. A clean work zone, a usable donor and controlled acquisition are separate conditions, none of them guarantees the result, and this is lab work rather than a parts swap anyone can do at a workbench.
Technical references: ISO 14644-1:2015 — air cleanliness classed by airborne particle concentration · HddSurgery head change tool guide (WD BB/BD/JB/JD, 1 platter) — tool fit is not head compatibility · HDDSurgery: Development of ToolFinder — one model name, different families and tools. Only the public ISO abstract was consulted, so no numeric class limits are given here. Both HDDSurgery items are a tool developer's product documentation for particular older drive families and products; they are not a general donor-matching specification.
Related: HDD Heads, Platters and Mechanical Failure · HDD Donor Parts and Compatibility · Imaging Bad Sectors and Unstable HDDs · How to Image a Failing Drive Safely · HDD Firmware, Service Area and Translator Recovery · Mechanical Hard Drive Failure · Water or Fire Damage · Directory of physical recovery providers