Common Antminer faults and how to clear them
What to do when an ASIC behaves oddly
We collected the cases that come to us most often, and the order of work for each one. Half of them are solved without opening the ASIC.
Temperature sensor initialisation error
A classic. The order of work, from simple to hard:
- Check the power. Voltage spikes on the mains give this error regularly.
- Check that the hash board ribbon cables sit tight. Reseat them.
- Inspect the connectors and contacts: dust and oxidation get in the way of polling the sensors.
- Try a power supply you know is good.
- Cut power to the ASIC for 10 to 15 minutes and switch it on again.
If the error stays, it is the sensors on the hash board itself, and the chips need replacing. As a stopgap you can turn on ignoring of faulty front sensors in the advanced settings, but that is a delay, not a cure.
Other typical cases
| Symptom | Cause | What to do |
|---|---|---|
| The CV board went back to stock after a power cut | Normal CVITEK behaviour | Phoenix-scanner on the network, or a repeat Toolkit scan with CV recovery |
| "Dev fee offline", "DEV FEE ERROR" | No connection to the fee pool | Check the network and DNS, update the firmware |
| Hashrate dropped in a step | A hash board went offline or the profile stepped down | Look at the board table and the logs |
| The ASIC reboots in a loop | Build and board revision mismatch, overheating, power | Logs, temperature, rollback to stock and reinstall |
| The root password does not get you in | On AsicBoost firmware the password is admin | Use admin, root is stock |
| Pool addresses do not resolve | DNS is not working | Set 8.8.8.8 and 8.8.4.4, run the network test |
| Fan in Error status | Connector, cable, a seized fan | The order of work is in the cooling section |
| The ASIC will not start, hydro | Temperature spread over 10 degrees or the loop colder than 20 °C | Warm the loop up, look for an air lock |
Decoding of specific log lines and error codes is collected on a separate page, Antminer logs and errors. What to watch a whole fleet with is listed in the diagnostic tools section.
Third-party management systems
Working through HiveOS and panels like it there is one typical error: the system tries
to log in to the ASIC with the password root and can neither reboot it nor apply
settings. The reason is that on custom firmware the password is admin. You change
that in the settings of the panel, not in the firmware.