IceRiver Error 300/301/302 Temperature Sensor Failure

Hard power-cycle the miner: switch off at the wall outlet for 60 seconds, then power back up. Some 300 events are transient I2C glitches; a clean cold boot re-initializes the bus and the firmware's sensor self-test. If the code clears and stays clear after 30 minutes of hashing, no further action needed. If it returns, escalate.

Log into the web UI, note the firmware version, compare against iceriver.io/firmware-download/ for your exact model. If you are on a beta or older build, plan a firmware refresh in step 8. If you are on the current stock build, do not re-flash unless other steps fail - flashing introduces its own risk.

Factory-reset via the reset button. Press and hold the reset button until the D3 and D4 LEDs start flashing simultaneously, then release and wait until D3/D4 go steady. This procedure is documented in IceRiver's official FAQ. The reset re-applies the stock firmware image on top of any drift and is a no-cost, no-tools software fix worth trying before opening the chassis.

Verify ambient and airflow even though 300/301/302 is a sensor fault - confirm the miner sits in <=35 C ambient (Normal Mode) or <=30 C (Performance Mode) with at least 15 cm clearance front and rear. A miner that has been thermally stressed is more likely to have aged sensor-circuit components, and any non-thermal diagnosis assumes baseline thermal hygiene.

Power off at the wall. Open the chassis (typically four corner screws on KS-class). Re-seat the data ribbon between each hashboard and the controller: disconnect, inspect the female header for bent pins, blow out any dust, reconnect until firmly clicked. This is the #1 documented IceRiver fix per the official KB and Apex To Mining's 2025 guide. Roughly the majority of 300 codes on otherwise-healthy boards clear at this step.

Re-seat the hashboard power connectors with the same care. Power connectors carry the high-current rail; loose fit shows up as voltage drop and can mechanically disturb the data ribbon next to it. Inspect for blackening, oxidation, bent pins. Reconnect and click firmly home.

Slot-swap the hashboards (KS3 / KS5 series only). Label slots 0/1/2. Move the board the dashboard fingered to a different slot. Power up, observe for 10 minutes. If the fault follows the board = bad board (escalate to sensor-level repair). If the fault stays in the slot = bad slot, cable, or controller path (escalate to controller-side diagnosis).

Reflash to the latest stock firmware. Download the model-specific image from iceriver.io/firmware-download/. Verify model match exactly - flashing a KS3M image to a KS3L bricks the controller. Apply via web UI; if the UI is unreachable, use SD-card recovery per IceRiver's recovery procedure. After the flash completes, factory-reset and observe for 30 minutes.

Multimeter the I2C pull-up rails on the suspect hashboard. With chassis open and miner safely off, identify the SDA / SCL pads near the sensor IC. With a multimeter in continuity mode, verify the pull-up resistors are intact and not open-circuit (typical values 4.7 kohm or 10 kohm - confirm against an intact pull-up on the same board). If a pull-up reads open, you have found a cracked joint - escalate to Tier 3 reflow.

Multimeter the ribbon-cable continuity end-to-end. With the ribbon disconnected at both ends, test each conductor. A cable that has been folded, kinked, or pinched can break a single conductor - exactly the failure mode that disables the I2C bus while leaving every other signal alive. Replace the ribbon if any conductor fails the test.

Reflow the sensor IC and immediate surroundings. Pull the hashboard. Apply flux to the sensor IC pads and the surrounding pull-up resistors and decoupling caps. Reflow with hot-air station: preheat the PCB to ~120 C from below if a preheater is available, then top-side hot air at 300-320 C for ~30 seconds. Let cool naturally, clean residue with 99% IPA, re-test. Cracked-joint failures are the #1 fix-by-reflow mode on aged hashboards.

Replace the sensor IC if reflow does not fix it. Source an LM75-family or compatible TMP-series digital temperature sensor in the same package (most KS-class sensors are SOIC-8 or SOT-23-5). Match the I2C address - check IceRiver's hardware-revision schematics or use a logic analyzer on the I2C bus to read the address the firmware tries to talk to. Solder in the replacement, reflow, re-test.

Replace cracked pull-up resistors found in step 9. Use the same value (typically 4.7 kohm or 10 kohm for I2C - confirm by reading an intact pull-up on the same board first). Surface-mount, 0603 or 0402 package depending on board revision. Reflow, clean, re-test.

Inspect for water and condensation damage under magnification. Look for white residue, greenish corrosion, or dark traces near the sensor area. If found, clean thoroughly with 99% IPA and a soft brush, dry under a heat gun at low setting, then reflow any suspect joints. Boards with corrosion past the sensor footprint should be flagged for full inspection - corrosion travels along PCB traces.

Roll firmware to a known-good prior version if a recent firmware update preceded the fault. Pull the prior version's image from your records or contact IceRiver support - prior versions are not always public. Apply via SD-card recovery. Some IceRiver firmware revisions have shipped with I2C driver regressions that throw spurious 300 codes on otherwise healthy hardware.

Stop DIY and ship to D-Central when: sensor reflow plus sensor replacement both fail; visible corrosion past the sensor footprint; controller-side I2C path suspect (controller MCU damage = controller swap territory); two or more boards fail simultaneously suggesting controller hardware; or you are not equipped for SOIC-8 or SOT-23 rework. Book a D-Central ASIC repair slot at d-central.tech/services/asic-repair/.

D-Central bench process: test fixture probes the I2C bus directly to confirm sensor health independent of the controller, sensor IC replacement with matched part numbers, controller-side I2C MCU diagnostics, full reflow of the sensor area plus surrounding components, post-repair 24-hour burn-in at nameplate with continuous sensor logging to verify stability. D-Central is positioning as the Western English-language IceRiver repair authority - the only direct competitor is Zeus Mining (China-based, trust-deficient).

Ship safely. Pack the suspect hashboard(s) - or the full miner if you cannot isolate - in anti-static bags, double-box with at least 5 cm foam every side. Include a printed note with: model and serial, observed error code(s) (300, 301, or 302), exact firmware version, board(s) affected, what you have already tried (cite the step numbers from this page), and your contact. The note saves diagnostic time, which saves you money.