Whatsminer M60S – Power Supply Failure

Hard-cycle the miner at the breaker for 60 full seconds, not a soft reboot — some P221A protection latches only clear with a full de-energization of primary-side bulk caps. Power back up. Check WhatsMinerTool for fault codes. If the miner comes back and holds for 30 minutes, you had a transient grid event and you are done.

Export the MinerTool `power` log via `Remote Ctrl` → `Export Log`. Open with any text editor. Record hex sub-codes and decimal fault codes from the last 24 hours — this is your single most valuable piece of evidence for every subsequent step. Screenshot it before you change anything.

Verify your power cable and outlet. M60S needs a proper C19 cord rated for 15 A continuous, plugged into a 200-240 V circuit. If you are on 120 V household, stop — the M60S does not run on 120 V. You need a dedicated 240 V split-phase or 208 V commercial circuit. No firmware work will change that fact.

Check ambient and intake. P221A thermal protection is aggressive. Ambient >35 °C, clogged filters, or a miner in a closed rack with no exhaust path triggers `243/245` and `253/254`. Shop-vac the intake filter, verify ≥15 cm clearance front and back, confirm ambient ≤35 °C at the intake.

Check WhatsMinerTool firmware version. If you are on a pre-`20240914` build and the P221A MCU has been updated, the handshake-window mismatch manifests as spurious `200` / `267` codes. Update to the latest MicroBT-signed firmware via MinerTool — never via unauthorized channels. M60S signature verification is strict and a failed flash bricks the control board.

Measure AC input voltage at the C19 inlet under load. Multimeter on AC, hot to neutral, while the miner is hashing at full power. Target 200-240 V sustained, <3% sag. Below 195 V under load triggers `206`/`250/251`/`270/271`. Sag means the problem is upstream of the PSU — undersized circuit, marginal breaker, tired PDU. Electrician work, not miner work.

Power off at the breaker. Disconnect the PSU-to-control-board data ribbon. Inspect both connectors for bent pins, black heat damage, or greenish oxidation. Clean with 99% isopropyl on a lint-free swab. Re-seat firmly until you hear the click. Swap the ribbon with a known-good one if you have one. This clears roughly a third of `dead PSU` cases per repair-bench data.

Torque the copper-bar output screws. Zeus Mining's repair guide calls out loose copper bars as a common overcurrent-protection trigger (`236`/`268` family) — loose copper = resistance = heat = protection trip. Use an insulated 10 mm socket. Firm, not gorilla-grip — MicroBT does not publish a torque spec but mechanical feel should be `solid, no further rotation under steady pressure`. Verify no bar is bent or scorched.

Measure PSU output voltage at the copper bars under load, miner hashing. Multimeter on DC, target 14.5 V ± 0.3 V sustained. Drift beyond ±0.5 V indicates regulation failure inside the P221A. Below 14.0 V sustained = PSU sagging — either caps are tired or the secondary-side regulation loop has drifted. Time for Tier 3 or a swap.

Hashboard isolation — the single most valuable Tier 2 move for `236`/`268`/`240` codes. Power off. Unplug all three hashboards from the PSU output. Boot the miner with just the control board powered. Does the PSU come up cleanly and hold? If yes, PSU is fine; a hashboard is pulling it into protection. Reconnect boards one at a time, powering between each — the board that trips protection is the problem. This step alone prevents most unnecessary PSU swaps.

Swap the P221A with a known-good unit of the same revision. M60S-specific: a P221A from an M50S may or may not work depending on firmware match — the handshake windows differ and you can trigger fault code `201` (PSU/config mismatch) or `8700` (PSU/miner model mismatch). Always match PSU revision to miner model. If the swap fixes it, the original is a bench/repair candidate. If not, the problem is not the PSU.

Open the P221A enclosure (warranty void — out-of-warranty only). Discharge primary-side bulk electrolytics with a bleed resistor before probing anything — they store enough energy to hurt you badly. Inspect all electrolytics for bulging, dome swelling, or electrolyte weep. Secondary-side caps near the synchronous rectifiers are the most common aging failure. Replace with equivalent or better 105 °C Nichicon/Rubycon — same or higher V and temp rating.

Inspect PFC MOSFETs and rectifier bridge for scorching, solder cracks, or visible shorts. Check the inrush current limiter (NTC thermistor or relay) for open circuit — failed inrush limiters are a common `PSU will not power on at all` cause on high-wattage miners. Replace failed components with identical or higher-rated equivalents. Reflow cracked joints with proper flux and a preheater.

Internal fan replacement. P221A internal fan is an 80 or 90 mm 12 V brushless — verify exact part number on your unit before ordering. Cheap fans do not survive 24/7 duty; choose ball-bearing or mag-lev rated 50k+ hour MTBF (Sanyo Denki, Delta, Sanyo 9G series). A $40 fan swap can save a $600+ PSU replacement.

Firmware recovery. If fault codes point to control-board / PSU MCU mismatch, flash the latest MicroBT-signed firmware via WhatsMinerTool SD-card recovery path. Never attempt unsigned or downgraded firmware on M60S — signature verification is enforced in the bootloader and a failed flash hard-bricks the control board, requiring manufacturer JTAG recovery. Document the `before` firmware version, MinerTool version, and exact sequence before touching anything.

Stop DIY when: (a) a second known-good P221A still throws `PSU_ERR` on the same miner, (b) visible electrical damage on the P221A (burnt PCB, vented caps, scorched MOSFETs), (c) hashboard isolation clears the PSU but you cannot identify the loaded-fault board without bench equipment, (d) firmware recovery is needed and signature verification has failed. Book a D-Central ASIC Repair slot — M60S bench work is becoming a core queue as the generation enters its mid-life cap-drift phase.

D-Central bench process for M60S PSU: programmable AC source (100-277 V variable) to isolate grid sensitivity; programmable DC load bank to test PSU output under stepped load (1 kW to rated); oscilloscope on the data ribbon to capture the PSU↔control handshake live; IR thermal imaging under load for 30-minute soak; component-level cap and MOSFET replacement using brand-matched parts (no generic substitutes); 24-hour post-repair burn-in on a hashing M60S before return.

Ship safely. Pack the M60S (or just the PSU + control board if sending subset) in original foam if you still have it — better engineered than most generic inserts. Double-box with ≥5 cm foam on every side. Include a written note with observed fault codes, export log timestamp, firmware version, and what you have already tried — saves us two hours of diagnostic work and drops your repair cost. Canada-wide shipping; US and international welcome from our Quebec workshop.