# F3 Command Reference — confirmed vs. uncertain

The Seagate **F3 architecture** terminal is common across modern Seagate drives,
but the *command set differs by family*. The well-documented recipes are from the
**7200.11 "Moose" / Barracuda** era (circa 2009, the famous BSY/0-LBA bricking).
A 16 TB **Exos X18 (ST16000NM004J)** is many generations newer and its F3 dialect
is **not reliably documented publicly**. Treat everything below accordingly.

## Ground truth (safe to rely on)

- **Electrical:** F3 UART is **3.3 V TTL, 8N1**. Baud is one of
  **9600 / 38400 / 115200** (sweep with the bridge's `~1/~2/~3` menu). 38400 is
  the most common default.
- **Entering the terminal:** after the boot banner, **Ctrl-Z** → `F3 T>` prompt.
  `F3 T>` = top level. `F3 2>` etc. = other diagnostic levels.
- **Level change:** `/` followed by a level id, e.g. `/2`, `/1`, `/T`, `/C`.
- **Help:** `?` at a prompt *may* dump that level's command list. **Always run
  `?` and record the output before issuing anything.** This is how you learn the
  *actual* command tokens for this firmware instead of guessing.
- **Logging:** every command may have side effects; keep `screen -L` running.

## CLAR200-specific expectation

This drive **boots and answers INQUIRY** (it is not BSY-hung), so you will
*probably* get an `F3 T>` prompt without the isolation trick. The lock is the EMC
firmware refusing **SCSI** FORMAT/READ-CAP — the F3 path sidesteps SCSI. The
realistic in-terminal goal is a **low-level format / translator regeneration to
512-byte logical sectors** (the drive is likely stuck at 520 B EMC-DIF and/or in
a "format corrupt / block size 0" state).

## Candidate commands — VERIFY with `?`, do NOT fire blind

These are **7200.11-era** and listed only so you recognize the *shapes* of
commands. **Confirm each against this drive's own `?` output before use.** Many
will not exist or will differ on Exos.

| Intent | 7200.11-era token | Status on Exos X18 |
|---|---|---|
| Spin down motor (level 2) | `Z` | unconfirmed |
| Spin up motor (level 2) | `U` | unconfirmed |
| Regenerate SMART (level 1) | `N1` | unconfirmed |
| Partition regenerate (level T) | `i4,1,22` | unconfirmed |
| Format / regen translator (level T) | `m0,2,2,0,0,0,0,22` | unconfirmed |
| Low-level format w/ sector size | `m0,6,2,,,,,22` (historic 512-format) | **likely different** |

⚠️ The `m0,...` translator/format commands are exactly the ones that, with wrong
parameters, can make a drive *worse*. On Exos, the safer modern route is often a
proper **Format Unit at the format level** with an explicit bytes-per-sector of
512 — but the token for that comes from `?`, not this table.

## BSY isolation trick (contingency only)

Only if the terminal is **completely unresponsive / the drive hangs BSY** at
power-on:

1. Power off. Slip a thin insulator (business card / Kapton) between the PCB and
   the **head-contact block** (or motor contacts, per the drive) so firmware
   boots without reading the system area.
2. Power on → you should now get `F3 T>` (firmware skipped the unreadable SA).
3. Carefully remove the insulator / re-seat contacts **without** powering down.
4. Spin up and proceed.

This is a 7200.11 technique. It is **unlikely to be needed** here (the drive
isn't BSY) and is drive-mechanically risky — last resort.

## Sources / further reading (verify, don't trust blindly)

- HDDGURU / MRT / PC-3000 community threads on "CLAR200" and "SS160520" Seagate
  EMC drives — the recurring finding is: SCSI tools fail, F3 terminal or array
  decommission is the path.
- Seagate 7200.11 BSY fix write-ups — for F3 terminal *mechanics* (prompt, `/`
  levels, `?`), not for the exact Exos command tokens.

When you discover the real Exos X18 tokens during the session, **record them
here** so ZR511KWK (and any future EMC drive) is a known quantity.