Definition
Diagnostic is the structured process of testing a Bitcoin mining machine to find out why it has slowed down, lost a board, or stopped hashing entirely. It combines software readouts from the miner’s control board with hardware measurements on the hashboard to isolate a fault down to a specific domain, signal, or even a single chip.
Also known as: hashboard diagnostics, fault isolation, ASIC troubleshooting.
What a diagnostic actually checks
A mining diagnostic works from the outside in. The first signals come from software: the control board runs cgminer or bmminer and exposes an API (the devs query on port 4028) that reports detected chip count per chain, per-board hashrate, accepted shares, hardware-error counts, and readings from the on-board temperature sensors. A miner that reports, say, 29 of 108 chips on one chain has a chain break, and that number alone narrows the search dramatically.
When the software points at a hardware fault, the diagnostic moves to the bench. Unpowered, a multimeter measures domain impedance and power-input resistance to spot shorts. Powered through a test fixture, you sweep each voltage domain and probe the signal chain (clock, command, response, reset) at domain boundaries. An oscilloscope confirms the 25 MHz clock is clean, and a thermal camera reveals a cold dead chip among hot working neighbours or a hot spot on a failed regulator.
From symptom to root cause
Most ASIC failures fall into a handful of families that a good diagnostic separates cleanly. A dead chip breaks the daisy chain so every chip downstream goes invisible. A shorted chip can drag the whole board to zero detected chips. A failed LDO or buck converter starves a voltage domain. A cracked crystal kills the clock, and a cold solder joint on a domain-boundary resistor breaks signal forwarding even though the chip itself is alive.
To pin down exactly which chip breaks a chain, technicians use a binary-search (dichotomy) method: inject a valid response at the midpoint of the chain, re-run chip enumeration, and watch whether the detected count rises or stays put. Repeating that halving locates the faulty chip without desoldering anything. Note that voltage control on modern Antminers is applied per voltage domain, not per individual chip, so a single bad regulator can take out a whole cluster of chips at once.
Why a home miner should care
For someone running a single S19 or S21 in a garage or a Bitaxe on a desk, diagnostics are the difference between a quick fix and an expensive paperweight. A board that suddenly reports far fewer chips, throws a wall of hardware errors, or trips thermal protection is telling you something specific, and the readouts above let you decide whether the cause is dried thermal paste, a tired PSU, or a genuinely failed chip. Catching it early protects both your uptime and your ROI.
You do not need a full repair lab to start. SSH into the control board, pull the chip count and temperatures, and check the kernel log for chain errors before you ever pick up a screwdriver. Open and custom firmware stacks often expose richer telemetry and per-chip statistics than stock firmware, which makes self-diagnosis far more approachable. If you do reach the soldering stage, our ASIC troubleshooting resources walk through the test points and failure modes, and the same diagnostic discipline underpins the open-source tooling growing up around projects in the Bitaxe hub.
Diagnostics and decentralization
Repair is one more layer of decentralization. Every miner you diagnose and recover instead of discarding is hashrate that stays in your hands rather than being concentrated into a large operator’s replacement order. Transparent, documented diagnostics, the kind that open firmware and community test tools make possible, lower the barrier so ordinary sovereign Bitcoiners can keep older hardware alive, contribute to the network hashrate, and own the full lifecycle of their machines.
Related terms: Hashboard repair, Reflowing, Temperature sensor, Voltage regulator, ASIC chip, Custom firmware
In Simple Terms
Identifying faults in mining hardware using test equipment and systematic testing before repair.
Diagnostic is the structured process of testing a Bitcoin mining machine to find out why it has slowed down, lost a board, or stopped hashing entirely. It combines software readouts from the miner's control board with hardware measurements on the hashboard to isolate a fault down to a specific domain, signal, or even a single chip.
Also known as: hashboard diagnostics, fault isolation, ASIC troubleshooting.
What a diagnostic actually checks
A mining diagnostic works from the outside in. The first signals come from software: the control board runs cgminer or bmminer and exposes an API (the devs query on port 4028) that reports detected chip count per chain, per-board hashrate, accepted shares, hardware-error counts, and readings from the on-board temperature sensors. A miner that reports, say, 29 of 108 chips on one chain has a chain break, and that number alone narrows the search dramatically.
When the software points at a hardware fault, the diagnostic moves to the bench. Unpowered, a multimeter measures domain impedance and power-input resistance to spot shorts. Powered through a test fixture, you sweep each voltage domain and probe the signal chain (clock, command, response, reset) at domain boundaries. An oscilloscope confirms the 25 MHz clock is clean, and a thermal camera reveals a cold dead chip among hot working neighbours or a hot spot on a failed regulator.
From symptom to root cause
Most ASIC failures fall into a handful of families that a good diagnostic separates cleanly. A dead chip breaks the daisy chain so every chip downstream goes invisible. A shorted chip can drag the whole board to zero detected chips. A failed LDO or buck converter starves a voltage domain. A cracked crystal kills the clock, and a cold solder joint on a domain-boundary resistor breaks signal forwarding even though the chip itself is alive.
To pin down exactly which chip breaks a chain, technicians use a binary-search (dichotomy) method: inject a valid response at the midpoint of the chain, re-run chip enumeration, and watch whether the detected count rises or stays put. Repeating that halving locates the faulty chip without desoldering anything. Note that voltage control on modern Antminers is applied per voltage domain, not per individual chip, so a single bad regulator can take out a whole cluster of chips at once.
Why a home miner should care
For someone running a single S19 or S21 in a garage or a Bitaxe on a desk, diagnostics are the difference between a quick fix and an expensive paperweight. A board that suddenly reports far fewer chips, throws a wall of hardware errors, or trips thermal protection is telling you something specific, and the readouts above let you decide whether the cause is dried thermal paste, a tired PSU, or a genuinely failed chip. Catching it early protects both your uptime and your ROI.
You do not need a full repair lab to start. SSH into the control board, pull the chip count and temperatures, and check the kernel log for chain errors before you ever pick up a screwdriver. Open and custom firmware stacks often expose richer telemetry and per-chip statistics than stock firmware, which makes self-diagnosis far more approachable. If you do reach the soldering stage, our ASIC troubleshooting resources walk through the test points and failure modes, and the same diagnostic discipline underpins the open-source tooling growing up around projects in the Bitaxe hub.
Diagnostics and decentralization
Repair is one more layer of decentralization. Every miner you diagnose and recover instead of discarding is hashrate that stays in your hands rather than being concentrated into a large operator's replacement order. Transparent, documented diagnostics, the kind that open firmware and community test tools make possible, lower the barrier so ordinary sovereign Bitcoiners can keep older hardware alive, contribute to the network hashrate, and own the full lifecycle of their machines.
Related terms: Hashboard repair, Reflowing, Temperature sensor, Voltage regulator, ASIC chip, Custom firmware