Temperature Sensor
Temperature sensors sit on every hashboard. When one of them goes silent, the firmware stops the whole miner: it does not work blind.
How temperature control is built
Each board carries several sensors, usually at the ends of the chip chain: where the air comes in and where it leaves already hot. The firmware compares their readings, drives the fans, and decides when to drop frequency and when to shut a board down.
Errors like temp lost or failed to read temp mean a sensor stopped responding. The firmware's reaction is harsh and correct: without a thermometer, overheating gets noticed by smell, so the whole miner stops even if the other sensors are alive.
The causes are almost always the same as with a failed chip: corrosion after moisture, a solder crack from thermal cycling, a cable damaged during careless disassembly. The sensor itself costs pennies, only the diagnosis is expensive.
Custom firmware can ignore a dead sensor and run on the remaining ones. It is a crutch to limp along until the repair: the patch of board around the blind sensor is left unwatched, and the cause still has to be fixed.
The short version
| Where they sit | at the ends of each board's chip chain |
| Typical error | temp lost, failed to read temp |
| Firmware reaction | the whole miner stops |
| Common causes | corrosion, solder joints, a cable |
| Custom firmware workaround | ignore the sensor, only until repair |
What to do about temp lost
Start with the board's cables and connectors: after a move or a cleaning, half of these errors are cured by reseating.
Check whether one sensor is silent or all of them on one board: all at once points at the cable and controller, a single one at the sensor itself.
Do not leave the miner running long on firmware with a disabled sensor: a blind zone on the board means failed chips down the road.
Bring the error text from the log to the repair shop: which board and which sensor, that is half the diagnosis.
An example from the workshop
A miner dies a minute after startup, the log shows failed to read temp on the second board. The cable is fine, the sensor at the air exit is not answering: under a microscope there is corrosion at one pin. Sensor replaced, that patch of board washed, an hour of work, and the miner holds its rated numbers around the clock again.
Related terms
Where to go next on the site
Hardware
- ASIC Miner knowledge base
- ASIC Chip and Hashboard knowledge base
Software
- Software firmware
Questions and answers
Why does the whole miner stop over one sensor
Because blind overheating costs more than downtime: a board without a thermometer burns quietly. The stop is protection, not a firmware whim.
Can I just disable the sensor in the settings
Custom firmware allows it, but that is driving without a speedometer: fine to reach the repair shop, not a way to live.
The sensor shows odd numbers but is not silent
Jumping or frozen readings are a fault too: the firmware steers the fans by a lie. The diagnosis is the same as with temp lost.
How the terms connect
Every link in the chain is clickable. Orange marks where you are now.
Looking for an ASIC miner
The catalog holds 212 models. You can compare them by hashrate and by joules per terahash, then plug your own rate into the calculator and see what stays in your pocket.
Page written and checked by Denys Klimchuk. Updated .