Mining Academy
- Home
- Academy
- Maintenance and Repair
Antminer Errors: Overheating, Fans, Hash Boards, Power
Antminer Errors: Overheating, Fans, Hash Boards, Power
Most Antminer errors come down to four causes: overheating and fan problems, unstable power, a lost connection to a hash board, and wrong network or pool settings. The first thing to do is open the Kernel Log in the web interface and find the line that names the error. It almost always points to a specific part: a fan, a temperature sensor, a board or the power supply. Below are the most common messages, a diagnostic order, and the point at which repair is better left to a service shop.
Where to start diagnosing
Before you open the unit, gather facts. It takes about ten minutes and often saves a lot of unnecessary work.
- Open the miner's web interface by its IP address and check the Miner Status tab: how many boards are visible, the hashrate of each, temperatures and fan speeds.
- Read the Kernel Log right after boot. The first two or three minutes are the most informative, because most faults show up while the boards initialize.
- Recall what changed before the failure: a firmware update, a move to another location, hot weather, a power outage, a new extension cord or socket.
- Save the log to a file. If the unit goes to repair, the technician will need it.
Message wording differs between models (S9, S17, S19, S21, L7, L9) and firmware versions, so the fragments below are typical rather than exact. Search the log for keywords such as temp, fan, chain, asic, power and voltage.
Overheating: ERROR_TEMP_TOO_HIGH and related messages
Messages like ERROR_TEMP_TOO_HIGH or temp too high mean that a board or chip sensor passed the allowed limit and the miner stopped hashing to protect the chips. Causes, from most to least common:
- Dirty heatsinks and fans. Dust clogs the fins, airflow drops and temperature climbs. Blowing the unit out with compressed air (powered off and unplugged) solves most cases.
- Hot intake air. An ASIC is designed to run in a reasonably cool room. If hot exhaust air gets pulled back into the intake, overheating is inevitable. Check how fresh air comes in and how hot air leaves.
- An overly aggressive performance mode. Overclocking or a high-performance mode raises heat output. Return to the stock frequency.
- Dried-out thermal paste or poor heatsink contact. This is workshop territory.
If one board runs noticeably hotter than the others in the same airflow, suspect that board rather than the environment. More on temperatures and ventilation is in the article on mining at home.
Fans: ERROR_FAN_LOST
ERROR_FAN_LOST (or "fan lost", "fan speed error") appears when the fan tachometer gives no signal or the speed is below normal. The fan is not always broken. Check in this order:
- Whether debris or a cable is jamming the impeller, and whether it spins freely by hand.
- Whether the fan connector sits firmly on the control board. Vibration loosens contacts over time.
- Whether replacement fans were swapped between intake and exhaust positions on models with two fans.
- Whether the replacement fan has the wrong rating. A weak fan technically spins but reports a speed error.
If the fan spins and the error stays, the connector on the control board may be at fault. Do not run the unit in this state: without cooling the chips can be damaged within minutes.
Hash boards not detected: "only find N asic", chain not found
The most common search is "Antminer not detecting hash board". Miner Status shows fewer boards than expected (two instead of three, for example), and the log contains lines like chain[1] only find 0 asic, find N asic, will power off hash board or no asic. The logic is simple: at startup the controller polls the chips along the chain, and if not all of them answer, the board is switched off.
What to check:
- Data cables and board power. Power the unit off, reseat the ribbon cable between board and controller, and the power connectors from the power supply. Oxidized or burnt contacts are a frequent cause.
- Swap the board with a good one. If the missing board stays missing in a new port, the board is faulty. If the problem stays at the original port, suspect the cable or the controller.
- The number of chips found. If the miner finds, say, 76 of 114, it is usually a burnt-out section of the chain: one dead chip breaks the voltage chain. Such boards need oscilloscope diagnosis and chip replacement.
- Visible damage. Burn marks, darkened components, swollen capacitors, a smell of burnt electronics.
Repeated restarts in the hope that the board will wake up do not help and only add wear. If reseating the cables did not bring the board back, move on to repair.
Power: ERROR_POWER_LOST and voltage drops
Power messages (ERROR_POWER_LOST, power voltage rise or drop) and spontaneous reboots point to unstable voltage on the boards. Causes:
- A power supply at its limit. Older units lose capacitor capacity and cannot hold peak load.
- A weak socket or extension cord. An ASIC draws several kilowatts. A household extension heats up and the voltage sags. Use a dedicated line of adequate cross-section with a breaker rated for the load.
- Low mains voltage. Evening hours often bring sags; measure the voltage under load with a multimeter.
- Poor connector contact. Scorched plastic is a sign of high resistance at the contact. Replace the connector and wire rather than reseating them.
Work on a power supply involves dangerous voltage. Do not open the unit unless you are qualified: capacitors keep their charge after disconnection.
Network and pool: the miner runs but the pool sees no hashrate
Sometimes boards and fans are fine and the problem is the connection. Signs: Miner Status shows the pool as Dead instead of Alive, real-time hashrate is present, yet the worker is offline in the pool dashboard. Check:
- The stratum URL has no typos and the port is correct. On RuPool the main BTC port is 3333 and the LTC and DOGE port is 3334; backup ports and SSL are covered in the article on connecting an ASIC to a pool.
- The Worker field uses the
subaccount.workerformat; the password can be anything. - The device has an IP address and internet access: ping from the miner itself (some firmware has a Network Diagnosis tool), the gateway and DNS.
- Your router or ISP is not blocking a non-standard port. In that case the backup port 443 or an SSL connection helps.
If all three pool entries are empty or wrong, the miner idles: it heats up and makes noise without a single accepted share.
When to repair it yourself and when to use a service
It is reasonable to handle only what needs no soldering and no opened power supply: cleaning, a fan swap, reseating cables, replacing a power cable, updating or rolling back firmware, checking network settings. Chip soldering, capacitor replacement and reballing without the right equipment usually cause more damage.
Go to a service shop if:
- after cleaning and reseating cables a board is still not detected or finds only part of its chips;
- there are burn marks, swollen components or a smell of overheated insulation;
- a power error repeats on a known-good line;
- the unit is still under warranty, since opening it yourself may void it.
Before sending a unit in, save the log and describe the symptoms: how many boards are detected, at what temperature the failure happens, what changed. This shortens diagnosis. If you are looking for a workshop, see the ASIC repair section of our site. If a unit turns out to be uneconomical to repair, take a look at the used equipment market.
How to prevent repeat failures
- Every one to three months (depending on dust) blow out the unit and check fans and connectors.
- Keep the room temperature within the manufacturer's recommendations and give hot air a way out.
- Do not overclock without a reason.
- Use a dedicated power line of sufficient cross-section and avoid cheap extension cords.
- Watch hashrate in your dashboard: a sudden drop is the first signal of trouble. For causes of drops, read why hashrate drops.
FAQ
My Antminer does not see one hash board. Is the board always broken?
No. The cause is often a poor ribbon cable or power connector contact, or an overheated controller. Power the unit off, reseat the cables and power, then swap the board with a working one to see whether the board or the port is at fault.
Can I run the miner if a fan reports an error?
You should not. Without proper cooling the chips overheat very quickly, and protection does not always shut the unit down in time. Replace or repair the fan first.
Why does the miner reboot every few minutes?
Most often it is a power problem: a weak socket, a voltage sag, a worn power supply or an overheating board. Read the Kernel Log just before a reboot; it usually records the reason.
Will reflashing firmware help?
Only when the cause is software: corrupted settings, a failed update, incompatible firmware. Hardware faults such as dead chips, bad contacts or a failing power supply are not affected. The manufacturer's official firmware is the safer choice: third-party firmware is installed at your own risk and may void the warranty.
How can I tell whether the problem is the hardware or the pool?
If boards show up in Miner Status, temperatures are normal, but the pool entry is Dead or there are no accepted shares, check the stratum address and port, the worker name format and internet access. If in doubt, write to RuPool support chat (available 24/7) or Telegram @rupoolpro_support.