# Controller A volatile reset-witness result

Status: passed
Type: evidence
Updated: 2026-07-23
Next Action: use the proven J-Link-off/discharge sequence for the next armed data-durable power-cut attempt
Campaign Reason: Historical record from campaign 013-close-hpt-media-2-initial-release-qualification; retained independently after campaign closure and requires no separate owner-queue execution.
Campaign: standalone

## Result

Controller A `1751136` on J-Link `69664082` passed a RAM-only reset witness
using Debug S-record SHA-256
`4BEDE424D276796271DDB275787ED1D02ED2205B3A8616FF631C01141602EB60`.
The witness did not hold or modify an SD media write.

Before the power cycle:

- diagnostic command was set to `0x52535457`;
- diagnostic status word zero was set to `0x575354A5`; and
- normal raw-HPT logging remained active.

The operator then:

1. powered J-Link off while Controller A remained powered;
2. waited 10 seconds;
3. powered Controller A off;
4. left both unpowered for 30 seconds;
5. powered Controller A on with J-Link still off and waited for boot; and
6. powered J-Link on and reconnected it for read-only inspection.

Post-cycle verification passed:

- command witness cleared to zero;
- status witness cleared to zero;
- `gSdCardState = SD_CARD_STATE_RAW_HPT` (`6`);
- HPT media state was `HPT_MEDIA_STATE_HPT_READY` (`5`); and
- HPT status was zero.

This proves the operator sequence executes the normal MCU startup-clearing path
and that powering J-Link after boot does not recreate diagnostic RAM. Use this
exact sequence, including both discharge waits, for the next armed cut.
