The six common faults
| Symptom | Likely cause | Check first | Usually needs |
|---|---|---|---|
| Hashrate down by about a third, one chain missing | Hashboard: failed chips, connector or temperature sensor | Kernel log for a chain with zero or missing chips; reseat the data cable | Lab repair of the hashboard |
| "Temperature too high" or hashrate throttles | Dust, failed fans, or intake air above about 40 °C | Intake temperature, fan speeds, dust in the heatsinks | Cleaning, fans or a cooler location |
| No power, or the miner restarts | Power supply or input power | Input voltage and the PSU’s fan and LEDs | PSU swap |
| Fan speed error, miner stops | A failed fan | Each fan spins and reports speed | Fan swap |
| Miner unreachable on the network | Control board, firmware or network cable | Link lights, IP scan, a firmware reflash | Control board swap or reflash |
| Hydro model: high coolant temperature or leak alarm | Flow, pump or a leak in the loop | Coolant flow and fittings | Repair of the loop, then a pressure test |
What you can check yourself
- Read the kernel log on the miner’s web page. Every chain should report the full chip count for the model; a chain with missing chips is a hashboard fault.
- Power-cycle once and reseat the data cables between the control board and each hashboard.
- Check intake air temperature and clean dust from fans and heatsinks; in the Gulf, sand-laden air clogs them faster than people expect.
- Reflash the manufacturer’s current firmware if the miner misbehaves after an update.
Don’t open a hashboard or the power supply yourself: hashboards need chip-level tools, and PSUs hold dangerous charge after they are unplugged.
When it needs a repair lab
A hashboard that has lost chips needs diagnosis to find the failed chip or voltage domain, rework, and a long test under full load to prove the fix. A quick test can pass a board that fails again within hours, so ask any repairer how long they test and what they guarantee.
How Segments handles repairs
- Open to any owner of Bitmain Antminer, MicroBT Whatsminer and Canaan Avalon machines, hosted with us or not.
- On site or in the lab. Fans, power supplies and control boards are swapped at our hosting centres; hashboards go to our UAE lab for chip-level work.
- Diagnosis, then a written quote. The diagnosis fee is credited against the repair if you go ahead.
- Tested under full load for 12 hours, with photos and a test report for every unit, and a 7-day warranty on replaced parts.
- Hosted machines: no hosting fee while a unit is in repair, and a spare unit on request.
To book a diagnosis, contact us with the model, serial number and what the kernel log shows.
Questions
- Why is my Antminer missing a hashboard?
- Usually a hashboard fault: failed chips, a bad connector or a temperature sensor. Check the kernel log for a chain reporting zero or missing chips and reseat its data cable; if the chain still fails, the board needs lab repair.
- Can a dead hashboard be repaired?
- Often, yes. A lab finds the failed chip or voltage domain, reworks it and tests the board under full load for many hours before it goes back into service.
- Why does my miner overheat in summer?
- Air-cooled ASICs are rated for intake air up to about 40 °C. In the Gulf summer, dust and hot intake air push them over; clean the fans and heatsinks and check the room’s airflow, or host the machine where cooling is engineered for it.
- Does Segments repair miners that aren’t hosted with it?
- Yes. Repairs are open to any owner of Bitmain, MicroBT and Canaan machines, with a written quote after diagnosis.