Power Management for ANTMINERs: Power Off, Reboot, and Restart

Antminers have no power button. They draw straight from the PSU the moment it energizes, so “powering off,” “rebooting,” and “restarting” all come down to controlling the AC feed or issuing a software reboot from the control board. This guide covers both the correct hard-power procedure and the software reboot paths across the S9, 17, 19, and 21 generations, plus the mistakes that cost people boards.

The single most important thing to know first: a software reboot and a factory reset are not the same operation. A reboot restarts the firmware and re-scans the hashboards. A factory reset wipes your pool credentials, tuning profile, and network config. Confusing the two is the classic way people lock themselves out of a running miner.

Before you touch anything: tools and prep

• The miner’s IP address (find it with your router’s DHCP table or Bitmain’s IP Reporter, then open http://<miner-ip> in a browser on the same network).
• Web-UI login (stock Antminer default is user root / password root unless changed).
• Optional: an SSH client for a clean CLI reboot on firmware that exposes it.
• For a hard cut: access to the branch-circuit breaker or the PSU’s AC inlet(s).

How to power off / shut down a miner

Because there is no on/off switch, you shut the machine down by killing its AC supply. In order of preference:

1. De-energize the branch circuit at the breaker feeding that outlet, if the miner is on a dedicated circuit. This is the cleanest cut and keeps you away from live connectors.
2. Unplug the PSU’s AC input at the wall/PDU. Pull the plug, not the low-voltage DC harness to the hashboards.

Whenever practical, stop the miner from hashing first (Web UI → Miner Configuration, or issue a reboot and cut power during the boot pause) so you are not yanking power mid-workload. In practice Antminers tolerate a hard cut, but a settle interval matters on restart — see below.

Dual-input PSUs (APW12 and hydro variants): these high-wattage supplies have two AC input sockets that must be fed together. Energize or de-energize both. Running a dual-input PSU on a single live input will fault or badly under-power the machine, and repeated single-leg starts stress the supply.

On voltage and circuits (read this before you plug a big miner into 120V): Bitmain’s older APW-series supplies (APW3, APW8, APW9/APW9+) are rated for 200–240V but will run on North-American 110/120V — they simply deliver reduced output. The higher-wattage APW12 is an exception: it expects a 200–240V input and should not be run on 120V. Either way, a modern S19- or S21-class miner needs its full rated wattage to hit spec, and a single 15A/120V circuit physically cannot supply that draw. So the practical rule is not “it won’t turn on at 120V,” it’s: expect throttled or unstable operation on the 120V-tolerant supplies, and give any full-size Antminer a proper 200–240V circuit sized to its nameplate current.

How to restart / reboot a miner

You have three routes. Pick the least invasive that solves your problem.

1. Software reboot from the Web UI (preferred)

A firmware reboot flushes a hung cgminer/bmminer process and re-initializes the hashboards without dropping your configuration. On stock Antminer firmware:

• 17 & 19 series (stock): log into the Web UI, open the System menu, and choose Reboot. The unit drops offline for roughly a minute, then re-scans boards and resumes hashing.
• 21 series and current firmware builds: the same System → Reboot action exists; the menu label may read “Reboot” or “Restart.” Look for an explicit reboot/restart control — never for a reset.

Do NOT click “Restore Factory Settings” / “Restore Initial Settings” to reboot. That button erases your pool URLs, worker names, and any tuning profile, forcing a full reconfigure. It is a recovery tool, not a restart. Use Reboot.

2. Reboot over SSH (if your firmware exposes it)

Where the firmware allows shell access, SSH in and issue a reboot:

ssh root@<miner-ip> then reboot

Stock Bitmain firmware often restricts the shell; open-source and third-party firmware exposes it fully. This is the fastest way to cycle a miner you can reach on the network but not through a wedged web interface.

3. Power cycle (hard restart)

When the UI is unreachable and SSH is closed, cut AC power as described above, then repower. Wait at least 30 seconds before re-energizing. That pause lets the PSU’s internal rails bleed down and its protection logic reset, and lets the control board release its NAND cleanly. Snapping the power straight back can leave the supply latched in a fault state or the controller mid-write.

The IP-Report and reset buttons are not reboot buttons — handle with care

The control board has an IP-Report button and, separately, a reset button — don’t confuse them. A quick tap of the IP-Report button broadcasts the miner’s IP to Bitmain’s IP Reporter tool. The reset button is a distinct control that wipes your config, exactly like the UI factory reset. Neither is a reboot button: if you only want to restart, leave both alone and use the Web UI or a power cycle instead.

How to confirm the restart worked

1. Give it 60–120 seconds after power returns; the controller boots before the hashboards report in.
2. Reload the Web UI status/miner page. Confirm all hashboards are detected (e.g. 3/3) and each is returning its expected chip count — a missing board or a short chip count after a restart points to a hardware fault, not a power-management issue.
3. Check that chip temperatures and fan RPM climb to normal ranges and that the unit re-registers on your pool with accepted shares.
4. If a board that was present before the reboot is now missing, or the hashrate comes back low, that is a diagnostic signal — move to fault-finding rather than repeated power cycling.

Safety notes

• Never open the PSU shell. The genuine shock hazard lives inside the sealed supply — the mains-side and PFC bulk capacitors hold a lethal charge long after unplugging. There are no user-serviceable parts in there.
• The low-voltage DC bus feeding the hashboards (roughly 12–21V depending on model) is not a shock risk, but it can push enormous current. A dropped tool or a slipped probe across those terminals can arc and vaporize metal. Kill AC power before working near the DC harness, and keep metal out of the bay.
• Do not disconnect or reseat hashboard power/data connectors with the PSU energized. Power down fully first.
• Let the machine reach a full stop before moving it; hot hashboards and spinning fans don’t appreciate being jostled.

Related: If a board fails to come back after a restart, work the symptom with the ASIC Fault Finder or open a ticket through Start a Repair. For model-specific power and reset details, pull the original documentation from our miner manuals library.