Whatsminer Error 240 – Low Voltage Protection Shutdown

Confirm fault is Code 240 and not a related output-side code via curl http://<miner-ip>:4028 -d '{"cmd":"status"}' from a laptop on the same subnet. Parse the error_codes array. [240] alone confirms output-undervoltage protection. [240, 233] or [240, 234] confirms multiple output-side envelopes tripping (bench territory). [240, 206] or [240, 218] confirms input problem driving output sag (fix input first via input-voltage diagnostic). Sub-code combination drives the entire fix path; do not skip this 5-second step.

Power-cycle at AC mains, not chassis switch. Kill AC at the PDU or wall breaker for 60 seconds to fully discharge primary-side bulk caps and let the PSU MCU clear its fault latch. Some firmware revisions hold the latch through a chassis-switch reboot and Code 240 re-appears immediately. If the error clears on AC cycle and stays clear under load for hours, you have a transient or marginal condition; if it re-fires within minutes under load, you have a real hardware fault and Tier 1 reset alone won't fix it.

Drop the tune by one power-mode step in WhatsMinerTool or the Web UI. If Code 240 started after a recent tune push (low-power -> high-power, or +10% overclock), revert to the previous mode and observe for 60 minutes. ~35% of Code 240 reports clear at this step alone per D-Central's repair queue. Plan the hardware fix on your timeline; run reduced-tune meanwhile. The ~10% revenue cost beats the 100% cost of a melted PSU after a sustained protection-bypass attempt.

Read the PSU model label and confirm it matches the chassis power class. Pull the PSU enough to read the metal nameplate: model number (P21 / P221 / P222 / P223), rated wattage, rated current. P21 (3.4 kW) for M30S / M30S+ / M31S; P221 (3.8 kW) for M30S++ / M50 / M50S; P222 or P223 (5.4 kW) for M50S+ / M50S++ / M60S+ / M66. An under-spec PSU in a high-power chassis throws Code 240 predictably. Most common 'why does my refurb miner keep tripping' cause.

Verify no recent service event triggered this. Check your service log for the last 30 days: PSU swap, miner relocation, lid opened, busbars touched, harness re-routed, breaker swapped, surge event. A Code 240 within 30 days of a service visit usually traces to a connection loosened during that visit (Tier 2 torque check is the fix). If no service events but the fault appeared with the season change (cold to hot, summer onset), the cap-aging hypothesis leads (Tier 3).

Verify intake ambient and airflow. Pop the intake grille, look for dust, check fan RPM in the dashboard. Healthy: ambient <=30C, fans at >=3,500 RPM on M30S class or >=5,500 RPM on M50S+ class. Ambient >35C sustained or fans bottoming out raises internal PSU temp, accelerates cap aging, and sags the rail faster on every duty cycle. Rearrange airflow, duct exhaust away from intake, or move the miner. Bitcoin Space Heater configuration is the cleanest permanent fix in a Canadian basement.

Measure the output rail at the control-box busbar under full load. Multimeter on DC, probe black on chassis ground, red on the +12V (or +14.5V on high-power) lug at the control-box busbar - not at the PSU, the BOARD side. Boot the miner, let it ramp to full tune, watch the meter for 60 seconds. Healthy: within +/-3% of nominal sustained (11.6 - 12.4V on 12V; 14.0 - 14.9V on 14.5V). Marginal: 11.0 - 11.4V (PSU at the floor). Failing: any sag below ~11.0V on 12V rail (Code 240 correctly firing on real undervoltage). The number drives the rest of diagnosis.

Torque every M6 copper-bar bolt to spec. Kill AC, wait 60s, remove the PSU-side and control-box-side covers. With a calibrated torque screwdriver at 3.0 N-m, check every M6 copper-bar bolt - typically four at the PSU output and four at the control-box input. Loosen each 1/8 turn, re-torque to 3.0 N-m (MicroBT spec is 2.5 - 3.5 N-m). Look for blackened or blued copper, missing lock washers, or bolts that never seated flat. Re-assemble, hash for 60 minutes, remeasure rail voltage.

Replace the 12V output harness if visibly damaged. Walk the harness from PSU to control box under bright light. Look for jacket discolouration (heat marks indicate prior overload), visible crimp damage at the ring terminals, kinks, pinch points, rodent damage, greenish corrosion on the lugs. Squeeze along the length - soft spots or crunching = internal conductor damage. Genuine MicroBT replacement harness is CAD $30-60. A damaged harness adds 0.2 - 0.5V of unwanted drop under load and can cause Code 240 even when both PSU and busbars are clean.

Compare AC input voltage at C19 inlet under boot load. Clamp meter on L-L at the inlet, watch during the 3-5 second inrush and the first 60 seconds of ramp. Healthy 240V nominal stays >225V; 208V Wye stays >205V. A sagged input plus aged PSU can fire 240 without simultaneously firing 206 (different sampling). If input is clean, exclude this category; if input sags, fix the input first - dedicated 240V 30A circuit with #10 AWG on runs >8m solves half the 'mystery' Whatsminer faults forever.

Swap to a known-good PSU of the same model family. If you own multiple Whatsminers, pull a known-good PSU from a healthy miner of the same model. Wire it into the faulty chassis, hash for 60 minutes, recheck Code 240 and busbar voltage. Swap clears the fault -> original PSU is end-of-life or has a real internal fault -> Tier 3 cap rebuild or Tier 4 replacement decision. Swap doesn't clear -> fault is downstream of PSU (chassis, harness, hashboard); re-check Steps 8 / 9 / 12. Cleanest way to isolate 'bad PSU' from 'PSU being blamed for downstream problem'.

Pull hashboards one at a time and recheck the rail. Kill AC, disconnect the 12V ring terminals from each hashboard one at a time. Boot the miner with surviving boards, measure the busbar voltage under load, note the change. If pulling a specific board restores the rail to nominal AND clears Code 240, that board has an internal short or a failing buck regulator pulling the rail asymmetrically. Branch to hashboard-level diagnosis. If pulling boards doesn't change the rail droop, the PSU itself is the fault.

Discharge bulk caps and inspect PSU primary-side internals. Kill AC, wait 60s, remove PSU outer cover. SAFETY-CRITICAL: primary-side bulk caps store lethal energy for hours after AC removal. Discharge through a 10kOhm / 5W resistor across each cap's terminals for 30 seconds, read terminal voltage with DMM (must be <5V) before proceeding. Inspect bulk caps (typically 2 x 470uF or 2 x 560uF at 400-450V): bulged tops, vented bottoms, brown leakage at base, cracked cases, or sticky residue near pads = end-of-life. Visual inspection alone catches most cases. Photo-document before any rework.

Measure ESR and capacitance on suspect bulk caps. With a Peak Atlas ESR meter or equivalent, measure ESR at 100kHz on each bulk cap (in-circuit acceptable for ESR; for capacitance lift one leg or pull the cap). Healthy: ESR <100mOhm, capacitance within +/-15% of nameplate. End-of-life: ESR >500mOhm, capacitance <70% of nameplate. Both metrics together give high-confidence diagnosis. Single high ESR with normal capacitance = early degradation, replace prophylactically. High ESR + low capacitance = definitively gone.

Cap rebuild - replace bulk caps with quality matched parts. If diagnosis is end-of-life caps, replacement is the fix. Source genuine Nichicon, Rubycon, United Chemi-Con, or Panasonic electrolytics - match voltage rating (>= original, never less), capacitance (+/-10% of original), temperature class (105C minimum), ESR class (low-ESR / high-ripple variant required for SMPS service). Avoid no-name marketplace caps (counterfeits fail in months). Desolder originals with hot-air station + flux + braid (large parts; iron-only damages PCB pads). Install new caps with correct polarity, leads short. Reflow with eutectic or SAC305 per board's original alloy.

Re-flash the BTMiner control-board firmware via SD-card recovery. A corrupted PSU calibration blob in flash occasionally mis-sets the output undervoltage threshold and triggers Code 240 on a perfectly healthy PSU. SD-card recovery via the WhatsMiner bootloader procedure factory-resets this. CRITICAL: the firmware image must match your exact model and sub-revision - cross-flashing an M50S image onto an M50 bricks the control board (MicroBT version-checks on the CB identity EEPROM). Verify image filename against chassis nameplate before flashing.

Inspect the secondary-side regulation circuit under magnification. If caps test healthy but the rail still sags under load, the failure is on the secondary side: synchronous-rectifier MOSFETs, sense resistor, output filter capacitors, or the regulation feedback loop. Under 10x-20x magnification, look for scorched packages, lifted pads, cracked silicon, swollen output electrolytics, or burn marks on the PCB silkscreen. Bench-rework territory - hot-air SMD work, replacement parts matched by silkscreen part number, programmable load to verify. If not equipped, Tier 4.

Verify against a programmable DC load on a bench (skill gate). Pull the PSU, connect a programmable DC load rated >400A @ 12V (TDI / Chroma / Itech-class). Run a duty-cycle profile mimicking the miner: 50% rated for 1 minute, 100% rated for 1 minute, 0% for 10 seconds, repeat for 30 minutes. Log V_out against time. Healthy: holds V_out within +/-2% of nominal across the profile. Marginal: sags >3% during 100% segments, sometimes recovering. Failing: sags >5% and trips. Reproduces what the chassis does to the PSU under load and isolates marginal behaviour conclusively.

Stop DIY when Tier 1-3 is clean but Code 240 still latches under load. A PSU with healthy bulk caps, correct spec, and clean busbars but still producing 240 under load is failing on the secondary-side regulation circuit (MOSFETs, sense resistor, OVP/UVP reference) or has a marginal PCB-level fault that only reveals itself under specific load + temperature conditions. Bench reproduction with a programmable DC load is the only reliable path. Book a D-Central ASIC Repair slot.

Stop DIY when you see scorched silicon, vented caps, or blackened PCB traces. Vented bulk caps, scorched secondary-side MOSFETs, blackened traces near the regulation circuit, or burnt smell are all signals of cascade damage. The first failure took down adjacent components, and replacing only the visibly-broken part leaves the unit vulnerable to immediate re-failure. D-Central's bench replaces the full input-protection chain, the bulk caps, the secondary-side MOSFETs, and inspects adjacent silicon under magnification - not just the visibly-burnt part.

Ship for combined-unit isolation when Code 240 correlates with a specific hashboard. If Step 12 isolated the fault to one hashboard pulling the rail asymmetrically, the fix is at the board level (failing PMIC, leaky buck cap, dying core regulator) - not the PSU. D-Central's hashboard fixture tests each board independently against rated current and identifies the failing component. Bench repair on the board runs CAD $100-250 per board depending on the failed part; cheaper than a new hashboard in most cases.

Ship with full context. Pack the PSU (optionally the chassis if shipping budget allows), a copy of your last btminer-api status JSON, the WhatsMinerTool PSU telemetry log if you captured one, your service history (recent moves, swaps, tune changes, ambient), and your busbar voltage measurements at idle and at load. Match chassis serial to PSU serial in your ship note - both are on metal plates. Canada-wide standard shipping; US / international welcomed. Turnaround 3-7 business days for isolated Code 240, longer if multiple faults surface on the bench.