Review the task before going to the rack

Make the SD card correctly

1. 1Use a SMALL, name-brand microSD: 2-16 GB (SanDisk/Samsung/Kingston). Cards >32 GB frequently fail - the bootloader can't read exFAT/SDXC.
2. 2Format the card FAT32.
3. 3Download the correct recovery image for the EXACT model from the manufacturer.
4. 4UNZIP the image and copy its CONTENTS onto the card (don't just drop the .zip on it).
5. 5Tools: balenaEtcher (easiest, cross-platform), Win32DiskImager, or Rufus (Windows).

Antminer SD recovery procedure

Before using this checklist: SD-slot position (left vs right of the Ethernet port) changes the procedure on some models - check that model's page.

1. 1Power OFF the miner.
2. 2Insert the prepared microSD into the control board's card slot.
3. 3Power ON - the board auto-loads from the card (~30 s to a few minutes).
4. 4When the green LED goes STEADY, flashing is complete. Do NOT interrupt power.
5. 5Power off, REMOVE the card, power on - it boots the recovered firmware.
6. 6Reconfigure network + pools (config was wiped) and verify the version.

Braiins Toolbox (install & manage Braiins OS+)

Braiins Toolbox is the GUI+CLI app for installing and managing Braiins OS+ firmware across Antminers - no SD card needed for most units.

1. 1Download Braiins Toolbox from braiins.com and the firmware package for your exact model.
2. 2Verify the download checksum (SHA-256) against the published value.
3. 3Confirm the miner is reachable on the LAN; note its IP.
4. 4GUI: point Toolbox at the miner IP(s) and run Install. CLI equivalent: bos-toolbox install <MINER_IP>.
5. 5Enter miner credentials when prompted (factory default often root/root).
6. 6Wait ~5-10 min - Toolbox auto-BACKS-UP the stock firmware, then the miner reboots into Braiins OS+.
7. 7Browse to the miner IP, change the password, set a Power/Hashrate target and pools.
8. 8Uninstall/revert: bos-toolbox uninstall <MINER_IP> restores the backup.

Bitmain APMinerTool (Antminer batch)

1. 1Scan one or more IP ranges to enumerate every Antminer (IP/MAC/model/status).
2. 2Select many units -> push a pool config, firmware image, or reboot in one batch.
3. 3Batch-convert DHCP -> static for production racks.
4. 4Export the discovered list for your site map/inventory.

MicroBT WhatsMinerTool (Whatsminer batch)

Before using this checklist: Whatsminers reject write-API and many monitoring platforms until you change the default password AND flip the API switch in WhatsMinerTool. It's the #1 onboarding gotcha.

1. 1Enter an IP range and Start - it lists every Whatsminer with status, temps, and error codes.
2. 2CHANGE the admin password and ENABLE the Miner API Switch (required before any write-API/monitoring works).
3. 3Batch pool config, firmware upgrade, reboot, power-mode toggle (Low/Normal/High), Fastboot, Restore DHCP.
4. 4Export status to CSV for records.

The RMA workflow

1. 1Confirm it's a HARDWARE failure (rule out cable/power/config/airflow); capture the exact error string.
2. 2Check warranty by serial (the white label near the LAN port). Watch typo traps: O/0, I/1, B/8.
3. 3Document: serial, model, purchase proof, fault + error string, photos.
4. 4Tag & quarantine the unit (fault + serial + date); pull it from the production row.
5. 5Open the vendor ticket and WAIT for a written RMA number (unauthorized returns are refused).
6. 6Ship (you pay inbound); keep tracking, update the ticket.
7. 7Track credits in a fleet RMA ledger: RMA#, ship date, resolution; reconcile on return.

Walk a no-pool / offline fault by layer

Do not jump to the pool or the hardware. Localize the break one layer at a time, lowest layer first.

Before using this checklist: Each layer you clear rules out everything below it. Work bottom-up and you find the break instead of guessing at it.

1. 1NIC + switch port: link light on both ends, cable seated, correct port. No link, and nothing above it matters.
2. 2IP / gateway / DNS: miner has a valid IP, reaches its gateway, and resolves the pool hostname. A bad gateway or dead DNS looks identical to a dead pool.
3. 3Firewall / egress: the stratum port is open outbound - often 3333, or 443 for TLS/encrypted stratum. A blocked egress port stops shares cold.
4. 4Pool status: only after your side checks out, load the pool's own status page. The outage may not be yours.

Work causes cheapest-and-commonest first

Once classified, chase causes in cost order. The cheapest, most common fix goes first; opening an RMA goes last. Skipping the order is exactly how a good board lands in the RMA pile.

Before using this checklist: An SD reflash often revives a 'dead' or no-boot board before any hardware swap. Rule out software before you call it hardware.

1. 1Power - qualified verification of the exact PSU, supply voltage/phase and cabling against model/nameplate requirements; no live reseating.
2. 2Network / config - pool URL, worker name, IP, DNS.
3. 3Hashboard / ribbon SEATING - reseat the data ribbon AND the board power connectors.
4. 4Thermal - dust, blocked airflow, dried paste.
5. 5Then and ONLY then - swap the board or open an RMA.

Hydro install and service review

Use the exact model and facility procedure. This checklist is a review for authorized personnel, not a universal installation sequence.

Before using this checklist: A stopped miner or pump does not establish isolation. Hot, pressurized coolant and stored electrical energy must be controlled before service; do not open fittings to investigate a running loop.

1. 1Confirm authorization, the equipment procedure, approved coolant and lifting arrangements before work.
2. 2Before connecting, opening, bleeding or removing fittings, isolate the relevant electrical and fluid sources, control stored pressure, allow cooling, and verify safe conditions under that procedure.
3. 3Authorized personnel complete the specified connections and checks. Use the manufacturer's fluid-specific chemistry, pressure and flow limits; do not substitute a generic coolant recipe.
4. 4Before return to service, complete the required leak, flow and air-removal checks under the approved commissioning procedure; clear personnel and obtain restart clearance before enabling hashing.
5. 5For removal, repeat the isolation, depressurization and cooling procedure before disconnecting couplers; drain, contain and store fluid as specified by the manufacturer and SDS.

Leak response and PPE

Liquid cooling means liquid can leak, and the two systems fail differently. Hydro leaks are water under pressure near live electronics; immersion leaks are fluid loss that can eventually run a tank dry. Know the response before it happens.

Before using this checklist: Water and energized boards do not mix. On any suspected hydro leak near live electronics, de-energize the affected unit first, then deal with the water. Never reach into a wet, powered machine.

1. 1Spot it early: run leak sensors and a secondary containment tray under tanks and CDUs so a drip is caught and alarmed, not discovered later.
2. 2Keep clear of liquid near potentially energized equipment and report the leak. Authorized personnel must establish electrical, pressure and thermal safety before fitting work or cleanup; follow the fluid SDS.
3. 3Contain and absorb with pads kept near the equipment; keep spare fittings and couplers on hand for a fast reseat.
4. 4Wear the PPE: gloves, safety glasses, and closed-toe shoes when handling coolant or fluid; avoid skin and eye contact and follow the fluid's safety data sheet.
5. 5After any fluid work, dry and inspect every fitting and connector before you re-energize, and log the incident so a slow recurring leak gets caught.

Power-on sequence

Bring it up in order so you can tell a power fault from a network fault from a heat fault.

1. 1Confirm the PSU/model match and that DC cables (Antminer) or AC (Whatsminer/Avalon) are fully seated.
2. 2For hydro or immersion, confirm coolant is flowing / the tank is filled BEFORE applying power.
3. 3Energize the circuit and power the miner; watch that the fans spin up and the status LED comes alive.
4. 4Give it a few minutes to boot the control board and initialize the hashboards.
5. 5Confirm the Ethernet link light before you go hunting for the IP.

Verify it reaches nameplate

Don't call it commissioned until it's actually hashing at spec and the pool sees it. A fresh miner ramps for several minutes while the boards initialize and the firmware auto-tunes, so the first reading will be low. Give it time, then confirm the real numbers.

1. 1Wait 3-10+ minutes for all hashboards to initialize and the tuner to settle.
2. 2On the miner's dashboard, confirm the pool status shows Alive/Active and accepted shares are climbing.
3. 3Confirm the 5-minute/average hashrate is close to the model's nameplate, with all hashboards and all fans present.
4. 4Check board temps and fan speeds are in normal range, then log the unit as commissioned against its rack position.

Current and continuous-load headroom

For standard-rated NEC-style continuous-load applications, 125% sizing corresponds to an 80% loading example. Properly applied 100%-rated assemblies have different conditions. Qualified electrical design determines the applicable conductor, breaker, assembly and site limits.

Before using this checklist: Electrical measurements and circuit changes require qualified, authorized personnel. A trip needs investigation and a safe-to-energize determination before any reset under the site procedure. Never increase protective-device ratings to mask a fault.

1. 1Review the manufacturer's maximum/input-current requirements and approved circuit design; a point-in-time measurement does not override them.
2. 2For a single-phase estimate, RMS current I = P / (V × PF). At unity power factor, 3500 W / 240 V is about 14.6 A; at illustrative PF 0.95 it is about 15.35 A. Neither PF value is an asserted miner specification.
3. 3For balanced three-phase total power, I = P / (√3 × line-to-line voltage × PF). Do not apply the single-phase formula to a three-phase total.
4. 4In the standard-rated 80% example, a 20 A circuit has a 16 A continuous-load ceiling. Confirm every applicable installation limit before adding or changing loads.

Balancing miners across breaker sections

A container or row is often split into breaker sections, and each section feeds a set of racks. The site's electrical design gives each section a planning target in kW. That target sits below the breaker's rating to leave headroom. When miners are deployed, they are placed so every section stays at or under its target. A model with higher wall power fills a section with fewer machines, so some slots stay empty on purpose. The Power balance tool in Technician Reference does this math for any section size and model mix, and prints a section worksheet.

Before using this checklist: Rated wall power is a planning figure. Real draw rises with a higher power profile or a hotter room, so keep the target's headroom. Phase balance, breaker trips and any circuit change belong to qualified electrical personnel, and moving or powering miners follows the site's authorization.

1. 1Find the section's planning target. If you only have the total approved load, divide it by the number of sections.
2. 2Divide the target by one miner's wall power in kW (watts ÷ 1,000) and round down. That is the most miners the section can take.
3. 3Spread those miners evenly across the section's racks, putting any extra on the first racks.
4. 4For mixed models, multiply each model's quantity by its kW and add the results. The total must stay at or under the target.
5. 5Swapping one model for another changes the load by the difference in their wall power. Check the new total before the swap.

Electrical safety: never work it energized

Everything above is high-current AC that can kill you or arc-flash you. The rule is simple: establish an electrically safe work condition before you touch the copper. NFPA 70E treats de-energizing as the default; energized work is only for qualified persons when de-energizing is truly infeasible. As a technician you rack, cable, and swap machines, tasks that should be done on a dead circuit with the miner unplugged.

Before using this checklist: Arc flash from a fault in switchgear or a PDU can exceed thousands of degrees and blast at lethal distance. Know the arc-flash boundary and required arc-rated PPE for the gear you are near, and never open energized distribution equipment you are not qualified and dressed to work on. Unqualified personnel do not perform lockout/tagout or energized work.

1. 1Confirm the task, authorization and exact equipment/site isolation procedure with the responsible team.
2. 2Authorized personnel identify and isolate all relevant sources and control stored energy using the required lockout/tagout procedure.
3. 3A qualified person verifies electrical isolation using the applicable procedure and suitable inspected test equipment. A switch position, dark LED or stopped fan is not proof.
4. 4Service begins only after safe conditions are established. Connection changes require isolation each time; never disconnect live board power connectors.
5. 5Complete the required inspection and personnel clearance before controlled return to service; re-isolate before further physical changes.

Reviewing a supply check

These are evidence-review steps for learners. They are not a live-probing procedure.

Before using this checklist: Only qualified, authorized personnel perform circuit testing. Do not infer conductor function from an unverified color or photo. In this US reference context, green identifies protective grounding, not a phase conductor.

1. 1Record the affected miner, PSU label, circuit identification and symptoms without opening electrical enclosures or touching conductors.
2. 2Ask the qualified, authorized electrical team to assess the supply using the site procedure and appropriate inspected test equipment.
3. 3Compare their documented results against the actual supply topology and the exact equipment input specifications.
4. 4Escalate abnormal or uncertain results; keep affected equipment out of service until the responsible team clears return to service.