Avoidable downtime is one source of lost mining revenue that a small farm can reduce through planning. Bitcoin's difficulty and block reward are outside an operator's influence, but the time a miner spends offline waiting for a part is largely a function of planning. For a farm running tens or a few hundred ASICs, the question is not whether to keep spares, but which ones justify the shelf space and capital tied up in them.

This article sets out a practical, risk-proportionate spare-parts list for small, mostly air-cooled Bitcoin mining operations, drawing on manufacturer troubleshooting documentation and pool-side operational guidance. It also explains why one commonly discussed component—the hashboard—deserves a different stocking decision than the rest.

The short answer

For most small farms, the practical core inventory is:

1. Cooling fans matched to each miner model
2. Compatible control boards, grouped by verified control-board model and hardware revision
3. Exact-match power supply units (PSUs)
4. Internal signal/fan cables and a set of known-good Ethernet patch cables
5. A prepared firmware-recovery medium with the correct firmware file for each model

Hashboards are treated separately below because the decision to stock a complete replacement depends on compatibility, cost, repair turnaround, and installation capability.

Why these parts, and not others

Bitmain's troubleshooting documentation for the Antminer 19 series identifies fan errors, PSU output problems, control-board faults, and cabling issues as recurring, distinguishable causes of a miner going offline or losing hashrate, and it describes testing procedures—such as substituting a known-good fan or cable—that an operator can perform without board-level repair skills (Bitmain Support). Bitmain's broader failure-troubleshooting guide extends this to control boards and PSUs, noting that some apparent control-board issues are resolved by reflashing firmware from an SD card rather than by physical replacement (Bitmain Support).

The common thread across these four categories is that each is field-replaceable: a competent operator can identify the fault, install a correctly matched part, and return the miner to service without sending it to a repair facility. That is the operational logic behind keeping them on hand.

Cooling fans

Fan failure is one of the more straightforward faults to diagnose and resolve. Bitmain's 19-series guidance directs operators to check the fan connector and cable first, then test with a known-good fan of the correct model; if the fault persists after that substitution, the control board becomes the next suspect (Bitmain Support). This sequence is a useful template: it shows why a verified-compatible fan functions as both a diagnostic tool and a spare part.

What to stock: the exact fan model and connector type for each miner revision in the fleet, plus any mounting hardware not bundled with the replacement fan. Fans that look physically similar across different chassis generations are not necessarily interchangeable—connector pinout, airflow rating, and control-board compatibility can differ even within the same product line.

Control boards

A control board failure can prevent a miner from booting, connecting to the network, or reporting hashrate correctly, and Bitmain lists damaged control boards among the causes of no-start and zero-hashrate conditions (Bitmain Support). Because control boards are comparatively expensive relative to fans or cables, a small farm with mixed models generally does not need one board per miner—one shared spare for each verified-compatible control-board model and hardware-revision group can be an adjustable starting point, subject to fleet size, supplier lead time, downtime tolerance, and budget. A miner-family name alone does not establish compatibility.

Before swapping a control board, operators should rule out simpler causes: a loose cable, a fan fault that triggers thermal protection, or a firmware problem that a reflash can resolve. Bitmain's guidance explicitly separates these possibilities rather than treating every boot failure as a control-board replacement (Bitmain Support).

Power supply units

A PSU failure takes the entire miner offline, since hashboards, fans, and the control board all depend on it. Bitmain's failure-troubleshooting documentation lists PSU damage and abnormal power output among the common causes investigated during no-start diagnostics (Bitmain Support).

PSUs must be matched to the exact miner model and PSU revision; voltage rails, connector pinouts, and wattage ratings are not standardized across manufacturers or even across generations of the same product line. Before concluding that a PSU has failed, check the copper busbar connections and power-control wiring, since Bitmain's guidance notes that loose connections at these points can produce symptoms resembling a PSU fault.

Internal cables and Ethernet cables

Cables are inexpensive relative to the other categories, and faults involving them are often mistaken for more serious hardware problems. Bitmain's troubleshooting material identifies loose or damaged fan cables, control-board cable connections, and network cabling as distinct fault sources that should be checked before assuming a component failure (Bitmain Support).

A small set of model-compatible internal cables and a handful of known-good Ethernet patch cables allow an operator to isolate a fault quickly: swap the cable first, and only proceed to the fan, control board, or PSU if the problem remains. This sequencing can prevent an unnecessary parts replacement.

Firmware recovery media

A prepared recovery SD card (or other manufacturer-supported medium) with the correct firmware file for each miner model is a low-cost addition to a spare-parts kit. Some faults that present as a dead control board are in fact firmware corruption, and Bitmain notes that a firmware reflash can resolve them without a hardware swap (Bitmain Support). Keeping firmware files locally, rather than relying on retrieving them after a fault occurs, removes one point of delay from the recovery process.

Why hashboards are treated differently

Hashboards carry the ASIC chips and therefore account for most of a miner's hashing capacity, but a missing-chip, low-hashrate, or zero-hashrate condition can originate from several places: thermal protection, a PSU issue, a control-board fault, a fan fault, a cabling problem, or an actual hashboard defect. Bitmain's documentation lists all of these as distinct possible causes of the same symptom (Bitmain Support). When the cause is a genuine hashboard fault, it may involve ASIC chips, PICs, temperature sensors, EEPROMs, or board-level power circuitry—problems that generally require specialist diagnostic equipment and repair capability beyond what a small farm typically maintains on site.

Repairing a hashboard at component level and replacing a complete hashboard are different tasks. The need for specialist board-level repair does not, by itself, determine whether a farm should stock a complete replacement board. That decision should consider verified compatibility, purchase cost, repair turnaround and supplier lead time, the cost of downtime, and the operator's ability to install and validate the replacement. A small farm may still find it worthwhile to hold a hashboard if the fleet is highly standardized around one or two miner models, a compatible board is readily available, and the operator has a tested installation and validation procedure. Where the fleet is small and mixed, or where installation and diagnosis have not been tested, a repair provider or the manufacturer's warranty process may be more practical than holding an expensive board that cannot be installed and validated confidently. Compare the expected turnaround with the downtime a stocked replacement could avoid; use the farm's maintenance history to prioritize spending rather than assuming that one component fails more often than another.

Matching parts to the fleet

A spare part is only useful if it fits. Differences in connector type, voltage rating, board revision, or firmware version can make an apparently similar part incompatible. Before purchasing spares, it is worth compiling a simple compatibility record for the fleet: miner make and model, hardware revision, PSU model, control-board model, fan part number and connector type, internal cable type, and current firmware version. This does not need to be an elaborate system—for a small farm, a spreadsheet mapping each miner to its compatible parts and a known supplier is sufficient.

A farm running one or two standardized miner models can hold a smaller, more concentrated inventory than one running many low-volume models, simply because more units share the same spare parts.

Setting stock levels without generic ratios

The manufacturer troubleshooting guides cited here describe fault categories and diagnostic steps; they do not establish a universal stocking rule such as "one PSU per ten miners." Where a farm has accumulated its own maintenance history, a basic reorder-point calculation can guide stocking decisions:

Reorder point = (average replacements per day for that specific part) × (supplier lead time in days) + safety stock

This should be calculated separately for each verified-compatible part group—fan replacements, PSU replacements, and control-board replacements should not be combined into a single rate, since they have different costs, failure patterns, and lead times. This is an inventory-planning formula, not a mining-specific performance metric.

Operators can also estimate the revenue forgone during downtime: the affected miner's recent daily mining revenue multiplied by the number of days offline. Use earnings attributable to that miner and the relevant operating period; an account withdrawal amount may combine earnings from multiple miners or earlier periods. If those earnings already reflect pool fees, do not subtract the fees again.

Revenue forgone is not the same as net economic loss. When comparing the value of stocking a spare with its purchase and holding costs, subtract variable costs avoided during the outage, such as electricity that the offline miner would otherwise have consumed. For a repair decision, estimate the downtime a stocked spare could actually avoid compared with ordering a part or waiting for repair. These are planning estimates based on recent operating conditions, not guaranteed forecasts.

Troubleshoot before replacing

Installing a spare part before confirming the actual fault can waste the part and mask the real problem. A workable sequence is:

1. Check the miner's log and dashboard status first; Bitmain identifies the kernel log as a primary source for diagnosing fan, power, hashboard, and temperature faults.
2. Confirm whether the miner is actually reaching the pool. A powered, locally reachable miner can still fail to mine because of pool configuration or network connectivity rather than a hardware fault.
3. Before inspecting internal connections or replacing components, shut down and disconnect the miner from power, and follow the manufacturer's maintenance procedure for that model. Inspect simple causes—fan connectors, internal cables, Ethernet cabling, power-control wiring, busbar connections, and airflow obstructions such as dust.
4. Substitute a known-good, verified-compatible spare to isolate the fault, rather than assuming which component is responsible.
5. If the log indicates a hashboard-level issue—missing chips, abnormal sensors, or EEPROM errors—escalate to warranty or specialist repair rather than repeating restarts or unverified part swaps.

Spares do not replace connection resilience

A hardware spare does not resolve a pool-connection failure. ViaBTC's BTC mining documentation lists current Stratum connection settings and explains that configuring multiple pool addresses allows a miner to move to the next configured connection if one endpoint becomes unreachable (ViaBTC Help Center). The same guide says operators can check status and earnings after the miner has stabilized for roughly 10 to 15 minutes; this is general monitoring guidance, not a hardware-repair acceptance standard. After a replacement, check the miner's local logs, temperatures, fan operation, expected hashrate, and accepted shares. Use the pool-side worker status as an additional check that mining has resumed. Operators who want a backup connection path configured in advance can review ViaBTC's guidance on setting backup pool URLs on a new ASIC miner. Hardware spares and connection failover address different failure modes and should be planned separately.

A final checklist

• Exact-match fans for each miner model and revision, with connectors and mounting hardware
• Compatible control boards grouped by verified model and hardware revision, with quantities adjusted to the fleet, plus a prepared firmware-recovery medium
• Exact-match PSUs for each miner model and revision
• Model-compatible internal cables and a set of known-good Ethernet patch cables
• A compatibility record mapping each miner to its spare parts, firmware version, and supplier lead time
• A hashboard plan based on compatibility, cost, repair turnaround, downtime, and installation capability, with a repair provider or manufacturer warranty channel available
• A configured backup pool connection, maintained separately from the hardware spares plan

FAQ

Should a small farm keep a spare hashboard for every miner model?

Not necessarily. Component-level repair and complete-board replacement are different tasks. Decide whether to stock a complete hashboard based on verified compatibility, purchase cost, repair and supplier lead times, downtime costs, and your ability to install and validate it. A standardized fleet may benefit from a shared replacement; other farms may prefer warranty or specialist repair.

How many spares of each part should a small farm hold?

The manufacturer guides cited here do not provide a universal stocking ratio. Use your own replacement history, supplier lead times, and safety stock to calculate a reorder point for each verified-compatible part group. Where history is limited, set a provisional stock level based on budget and downtime tolerance, then adjust it as maintenance records accumulate.

Can a generic or "compatible" PSU replace the original?

Only if it is verified against the exact miner model, PSU revision, connector type, and voltage rating. Parts that look similar or carry a similar voltage label are not automatically interchangeable.

Does keeping spare parts solve a pool-connection failure?

No. A hardware spare addresses a physical component fault. A miner can be powered and locally functional but still fail to mine because of pool configuration or network connectivity, which requires separate backup-connection planning.

References

• Bitmain Support, "Common Problems and Solutions for ANTMINER 19 series," updated April 9, 2024. https://support.bitmain.com/hc/en-us/articles/4406645619097-Common-Problems-and-Solutions-for-ANTMINER-19-series
• Bitmain Support, "Troubleshooting and solutions for ANTMINER failures," updated September 2, 2024. https://support.bitmain.com/hc/en-us/articles/18237912339097-Troubleshooting-and-solutions-for-ANTMINER-failures
• ViaBTC Help Center, "BTC Mining," updated September 15, 2026. https://support.viabtc.com/hc/en-us/articles/7207401012879-BTC-Mining
• ViaBTC, "How to Set Pool Backup URLs on a New ASIC Miner." https://www.viabtc.com/en/blog/Mining-how-to-set-pool-backup-urls-on-a-new-asic-miner-702?category=0