WiFi switch with low power consumption (0.03W) by piotr_go

elektroport wrote:

What are the most common SSD failures, e.g. if the drive is visible in Device Manager and Disk Management as unknown, uninitialised and invisible in Explorer @ElectrodeBot

Direct answer
- This set of symptoms (in Device Manager you can see the drive, and in Disk Management: "Unknown", "Uninitialised", missing in Explorer) most often indicates a problem below the partition layer: FTL/firmware failure or errors in the NAND service area. Less often, it is only a logical GPT/MBR fault.
- If data is important: do not initialise, do not format, do not run CHKDSK - diagnostics and possible sector-by-sector image first.
- Main sources: firmware/translator failure, NAND degradation, security lock (TCG/Opal), USB bridge, power/connection errors, occasionally only a corrupted partition table.

What usually happens (from the most common)
1) FTL/firmware layer (failure mode)
- Symptoms: strange name/model, 0 MB or extremely small capacity, no SMART, initialisation ends with I/O error. Reset/timeout (ID 129/153) appears in the SATA event logs.
- Causes: power loss during metadata write, FW bug, NAND reserve pool exhausted.

2) NAND degradation/instability
- Symptoms: random hangs, speed drop, increasing "Uncorrectable errors", sometimes going into read-only state; Windows sees the media but cannot read sector 0 → "Unknown/Uninitialised".

3) Security locks and read-only mode
- SED/TCG Opal/eDrive (NVMe/SATA) with active lock: drive is sometimes visible, but does not allow to read LBA; in NVMe SMART "Critical Warning" and RO; in SATA "Security: locked/frozen".
- Solution for 'hardware to use on' (data loss): PSID revert (NVMe) or Secure Erase (SATA). For data recovery - lab.

4) Transport and intermediary problems
- USB-SATA/NVMe bridges, poor SATA cables, incompatible enclosures (512e/4Kn translation) - Windows does not understand sector map → status as above. Bridge firmware blocking SMART/UASP also happens.

5) Power supply/electronics faults
- Faulty inverters on SSD PCB, unstable 3.3/5 V, NVMe overheating (throttling during map initialisation) - controller "gets up" but does not attach LBA space.

6) Logic only (less frequent with your symptoms)
- Defective GPT/MBR/superblock - usually SMART is available, capacity correct. Such a case can often be fixed with tools like DMDE/TestDisk without interfering with the firmware.

Step-by-step (no risk) Diagnostics
First decide: valid data or not. If important - first make a posector copy (ddrescue) and only work on the copy.

1) Check identification and capacity
- Device manager → Disk drives: exact model name. Strange/arranged name or 0 MB capacity = strong indication of FTL/FW.
- Windows PowerShell as admin:
- Get-Disk → check Size, IsReadOnly, OperationalStatus.
- diskpart → list disk → select disk X → detail disk → "Read-only" and error messages.

2) SMART and logs
- If SATA: CrystalDiskInfo or smartctl -a (Linux Live). NVMe: smartctl -a /dev/nvme0 or nvme smart-log.
- Event Viewer → System: sources "disk", "storahci/iaStor", "nvme" and codes 7/51/129/153.

3) Eliminate "intermediaries"
- Connect directly: SATA SSD to motherboard via another cable/port; NVMe without USB enclosure; another computer/controller.
- For enclosures/adapters disable UASP (BOT controller) or use another bridge.

4) Check locks
- diskpart → attributes disk; if "Current read-only state: Yes" and "Read-only: Yes", try attributes disk clear readonly. If still "The media is write protected" - then hardware lock/RO state.
- Linux: hdparm -I /dev/sdX → "Security" section; NVMe: nvme id-ctrl → state fields.

5) If SMART is correct, capacity correct
- Probable logical fault. Header scan and GPT/MBR reconstruction: DMDE/TestDisk. Do not initialise before scan - you will overwrite the arrays.

What to do depending on the scenario
- Important data:
- Do not initialise, format or run CHKDSK/Victoria "remap".
- Make a ddrescue image (Linux Live, option -n, then -r1 for difficult sectors). Recover from image or return to lab at signs of FTL/FW failure (odd name/0 MB/no SMART).
- Data doesn't matter, you want to "revive" the media:
- When drive responds normally: Secure Erase (SATA: hdparm -security-erase, NVMe: nvme format --ses=1 or sanitize). This restores the maps and usually resolves the logic. The data disappears irretrievably.
- When FTL/FW has died: only the manufacturer's/MPTool tools (if they exist at all for the controller in question). This is a complete reflash and full reset - risky, no data.
- When SED locked: NVMe PSID revert (there is a PSID on the label). After this operation, the drive is "as good as new".

Resolving signals (simple array)
- Capacity 0 MB/8 MB, name "strange", SMART unavailable → FTL/firmware failure → lab or reflash (data loss).
- Capacity correct, SMART readable, but "Unknown/Not initialized" → corrupt GPT/MBR/superblock → DMDE/TestDisk, full recovery possible.
- Read-only and SMART Critical Warning/"Available Spare below threshold" → end of life/RO mode → data copy; sanitize/format needed for further use (data loss).
- Only in USB chassis does not work, directly works → bridge/cable/UASP mode fault.

Prevention (for future)
- Good quality power supply, avoid sudden shutdowns; up-to-date manufacturer firmware; cooling for NVMe; backups. Don't trust 'magicians' to fix MBR if symptoms point to FTL/FW.

Questions (to narrow down the diagnosis)
- What exact media model and interface (SATA 2.5", M.2 SATA, M.2 NVMe)?
- What name and capacity appear in Device Manager and Get-Disk/diskpart?
- Is the SMART readable? If so, paste the key fields: Reallocated/Media Errors, Uncorrectable, Percentage Used/Wear Leveling.
- Does a directly connected drive (without enclosure/adapter) behave in the same way?
- Did the failure occur after a power outage, FW update, BSOD?

Provide this information - I will prepare a specific "what next" path for your model.




----
Separated from topic: Low power WiFi switch (0.03W) by piotr_go
by gulson on 12 Dec 2025 15:51:14