Common ASIC Miner Errors and How to Troubleshoot Them

05 September 2026

ASIC miners are rugged machines, but they fail in predictable ways. Most errors fall into three categories: hardware, connection, or firmware. The fix might be a cable reseat, a new PSU, or a trip to the electronics recycler. Here is how to tell the difference.

Hardware Errors

Hash board failure. This is the most common serious error. One board stops hashing while others continue. Check the physical connection first - reseat the ribbon cable. If the board still shows zero chips, inspect for blown capacitors. A board with visible damage is not worth repairing. Replacement boards cost less than the labor to diagnose and repair.

Chip temperature errors. Miners report these when one or more chips exceed the thermal threshold. The cause is usually dust buildup or a failed fan. Clean the miner thoroughly. Replace any fan that does not spin freely. If temperatures remain high after cleaning, the thermal paste under the heatsink may have degraded. This is repairable if you have the skill.

PSU undervoltage. The power supply cannot deliver stable voltage. This often appears as random reboots or a "power good" flag. Test with a known good PSU first. If the error clears, replace the power supply. Cheap PSUs fail frequently. Do not attempt PSU repair unless you are qualified to work with high-voltage DC.

EEPROM read error. The miner cannot read its configuration data from the board's memory chip. This typically means the hardware is failing. Try reflashing the firmware first. If the error persists, the board is dead. Replace it.

Connection Errors

Socket closed. The miner disconnected from the pool server. This can be a network issue, a pool outage, or a miner crash. Check the miner's network cable and your router. If other devices work, restart the miner. A persistent socket closed error points to a failing network interface on the miner's controller board.

Stratum timeout. The miner sent work but got no response from the pool within the timeout window. This is almost always on the pool side or the network between you and the pool. Try a different pool port. If the error follows you to multiple pools, your internet connection is the problem - check for packet loss.

Worker not authorized. The miner sent the correct pool address but a wrong worker name or password. Double-check the worker string. Common mistakes: extra spaces, missing period after the username, or using a wallet address instead of a worker name. This is a configuration error, not a hardware problem.

Firmware Errors

Signature verification failed. The firmware file did not pass the miner's integrity check. The file is corrupt or was not signed by the manufacturer. Download the firmware again from the official source. Verify the checksum. If it still fails, the miner's bootloader may be damaged - this requires a JTAG reflash or a new controller board.

Invalid extranonce. This is a stratum protocol error. The pool sent an extranonce value the miner's firmware cannot process. It usually resolves when you update to current firmware. If the pool is old and uses non-standard extranonce handling, switch pools. This is not a hardware fault.

When to Repair vs. Replace

Set a time value for your labor. A single hash board repair takes two hours minimum. If the miner is older than three years, replacement boards may not exist. A miner with two failed boards is not worth fixing - sell it for parts.

PSUs are replaceable. Controller boards are sometimes replaceable. Fan failures are trivial. Everything else on a modern ASIC is a surface-mount nightmare that costs more in tools and time than a new machine.

General Troubleshooting Sequence

Always start with the basics. Power cycle the miner. Check all cables. Verify the network connection. If the error persists, look it up in the manufacturer's error code table. If no table exists, the community forums have seen it. Do not flash new firmware until you have ruled out hardware causes. A failed firmware flash can brick a miner that only needed a new fan.

One final note. Miners generate error events constantly, and not all of them matter. A single chip temperature spike that self-corrects is normal. A pattern of repeated errors is not. Watch error frequency, not one-off alerts.

Back to mining