Antminer Z11 Hashboard Repair Service

Antminer Z11 hashboard repair covers the board the miner has stopped listing and the board still hashing short. The Z11e and Z11j carry the same board and get the same repair. A healthy board reports three chips, one per power domain, and any other number is a fault. Component-level repair in the USA takes 4-5 business days.

Antminer Z11 hashboard repair usually begins with a board that was blamed too early. On this machine a power supply that sags under the full three-board load produces exactly the picture a dying board produces — a chain that comes and goes, a chip marked x, output that will not stay up — and after enough months of that the board's input power stage burns for real, and the guess becomes true. So every board here starts on a supply that can actually feed it, and only then goes through the route any hashboard takes in our USA-based lab: board-level diagnostics, component-level repair down to the chip, cleaning and thermal work, and validation under load inside a running miner. The Z11e and Z11j are handled on this page as the same job. Working out which of the three boards failed is not something you need to do before writing to us.

Technical Diagnostics and Common Issues with Antminer Z11

Quick Symptom Checklist

These are the readings Z11 owners describe most often. Chain numbers, chip counts and rates differ from machine to machine — match the shape of what you see, not the digits.

• One of the three chains is missing from the status page — sometimes from the moment the miner boots, sometimes after hours of hashing — and a restart brings it back for a while
• Total output settles at two thirds of what the machine is rated for — around 90 kSol/s where 135 belongs on a Z11 — and stays there
• A chain lists 0 chips, or 2, where a healthy board shows 3 — the count is the message, whichever line it appears on
• ASIC status shows an x for one chip on a chain that used to show three good ones
• The miner is stable with one or two boards fitted and falls apart only when all three are in and hashing

Five lines from the kernel log of one of our own Z11s, in the order the log wrote them, from the minute one of its three boards went short. The timestamps, the process number and the chain letters will be different on your machine; what to compare is the number after has, and whether it stays the same from one pass to the next.

• Sep 15 03:07:39 (none) local0.notice cgminer[971]: check_asic_reg: chain J0 has 3 ASIC, a board answering with all three of its chips
• Sep 15 03:07:40 (none) local0.notice cgminer[971]: check_asic_reg: chain J1 has 3 ASIC, the second board, the same
• Sep 15 03:07:42 (none) local0.notice cgminer[971]: check_asic_reg: chain J2 has 2 ASIC, the third board answering with two — this is the fault
• Sep 15 03:07:57 (none) local0.err cgminer[971]: reg_handle CRC error crc = 02, a reply from the chain arrived damaged: the board is talking, and what it says cannot be trusted
• Sep 15 03:08:11 (none) local0.notice cgminer[971]: software_set_address_chain chainID2 asics 1 addrInterval 64, the next pass over the same board finds one chip, not two

Two things are worth ruling out before anything ships. A machine that does not appear on the network at all, or gives no web interface, is most often a control-board or network matter, and that case is handled on our control-board service rather than here — with one exception: if the power supply itself will not start, a shorted hashboard is among the causes, and the boards are worth checking before the control board takes the blame. And a supply on a 110 V circuit that is not rated for the whole machine will drop boards that have nothing wrong with them — its own label states what it delivers on that line.

Model-Specific Patterns We See on Antminer Z11

A chip that looks whole and is not. A hairline crack in the package of a BM1744 does not show under ordinary light: held in the hand, or photographed without that glare, the chip reads as intact, and only light skimming across its face at a low angle brings the line up. That is a trait of these chips rather than a rare case, so a visual once-over that finds nothing proves nothing about the board — its chips are measured one by one whether or not any of them looks damaged. That is the crack's quiet face. Its usual one is louder: a cracked chip, as a rule, shorts the board, and then there is no count of 2 of 3 to read — the chain enumerates no chips at all, and sometimes the power supply will not start either, so the whole miner sits dark and looks dead. A Z11 that will not come up at all can be carrying exactly this board.

The board that took the blame for the supply. A chain that comes and goes, then reports no chips, then comes back after a restart: on this machine that picture is produced by a supply sagging under the full three-board load at least as often as by the board itself. The tell is in the checklist — the miner runs for a day on two boards and fails on three. Left on a sagging supply long enough, the board's own input power stage burns, and at that point the board is genuinely faulty. A supply problem turns into a board repair with time, which is why the first thing we do is separate the two.

One chip, one domain. A Z11 board carries three chips and three power domains, one chip in each. When a domain's power stage fails the count on that chain drops by exactly one — from 3 to 2, not to zero — and the chip behind it may be perfectly intact, waiting for its power to come back. A board reporting 2 of 3 is the partial failure this hardware produces, and it is a domain repair before it is a chip replacement. A board at 0 of 3 is a different fault: the chain signal itself has stopped, at a chip or in the path between them.

Wall power that looks like a dead board. The machine draws about 1418 W and expects a 12 V supply of 1.2 kW or more behind it. On a 110 V residential line the usual supplies do not deliver that, and the miner reports the shortfall the only way it can — a chain dropping out, a chip marked x, a rate that will not climb — so the owner buys a replacement board and nothing changes. Nothing on the board was wrong.

Hardware Notes

What the Log Is Telling You

You do not need to act on any of this. It is here so you know what state your board is in before you decide to ship it.

Diagnostics Focus

Supply before board. On the Z11 a sagging supply and a burnt board are indistinguishable from the dashboard, so nothing is judged until the board has run on a bench supply that holds 12 V at full load. A board that only ever failed on a weak supply goes home unrepaired, and the report says so — that is a cheaper answer than a repair, and it is the true one.

Domain before chip, and every chip measured. With three chips and three domains there is no group to hide a weak chip in and no majority to average it away. We prove which domain is down first, then measure each of the three chips on its own, so a chip that merely lost its power is not replaced and a chip that is genuinely dead does not go back because the board looked complete on a cold bench.

Our Professional Repair Process

Gotchas

We do not condemn a board on a weak supply. Every Z11 board that comes in is powered on a supply that holds its 12 V under the full load before a word is said about it. Boards that failed only because the miner's supply was tired are not rare on this model, and they go back untouched with a note saying what actually happened.

A dark chip is not yet a dead chip. One chip per domain means a domain failure switches off exactly one chip, and that chip is not billed until its domain has been repaired and it has been measured on its own. A chip count read off a dashboard or a photograph is a guess; the count that goes on your invoice is taken after the power is back.

Typical Service Scenario

Bought in a lot, switched on, one board short. Z11s change hands used, often several at a time, with a supply of unknown age in the box. The first full load in months finds the weakest of the three boards, and the buyer's first sight of the machine is a chain that will not stay up.

A supply that faded slowly. A third-party supply that held its output when new and, after a year or more, no longer does under three boards. The board that draws hardest is the one whose input stage gives way, and by the time it reaches us the supply has already been replaced — too late for that board.

A residential circuit. The machine set up at home on a 110 V line with a supply that cannot feed all three boards there. Boards drop, a replacement board is bought, the replacement drops too. The board that then arrives here for repair is usually the healthiest part of the setup.

What Happens After Intake

The choice above between Hashboard only and Complete miner sets what arrives here, not what it costs — the price is the same either way, and the complete miner is the right pick when you would rather not pull the board yourself, because we find it on the bench. Every board follows the same route through our USA-based lab: intake and incoming inspection, then diagnostics, then component-level repair, then cleaning and replacement of the thermal interface, then testing — on STASIC and ASIC REPAIR fixtures first and for a minimum of one hour inside a live miner under load afterwards. Extended burn-in beyond that hour is available as a separate add-on. Chips, where they turn out to be genuinely dead, are billed separately from the repair, because on this board the number cannot be known until the domain feeding them is back.

Diagnostics and Validation Equipment

Bench measurement on a supply rated for the full machine, STASIC and ASIC REPAIR test fixtures for chip count and work verification, each of the three chips measured on its own, and final validation inside a working miner under real load. The fixture proves the chain; the per-chip measurement finds what a count of three can hide; only the miner proves the board.

All repairs are performed personally by Alex, who has been servicing ASIC miners since 2019. See credentials →

Contact our repair team today and get your miner back to full power. Contact us