Hashboard Not Detected — Troubleshooting Guide

Overview

When a hashboard is not detected, the miner's control board cannot establish communication with it. The web UI shows fewer boards than installed (e.g., 2 out of 3), the missing board contributes zero hashrate, and the miner operates at reduced capacity. This is a distinct problem from "missing chips" — here, the entire board is invisible to the control board.

Hashboard detection failures range from a trivially fixable loose connector to serious hardware faults like a fried PIC microcontroller or EEPROM corruption. This guide provides a systematic approach to isolating the root cause.

Symptoms

• Web UI shows 2/3 boards (or 1/3, or 0/3) when all three are physically installed
• "Chain X not found" or "Chain X: find 0 pcb" in kernel logs (Antminer)
• Whatsminer: "Hash board X error", fault code 110, or board slot shows as empty
• Avalon: "No AUC found" or board count shows fewer than installed
• Status page shows a board but with 0 chips detected — this is borderline between "not detected" and "missing chips" and may have overlapping causes
• LED indicator: Board power LED is off or blinking error pattern while other boards are solid

Quick Checks

Diagnostic Flowchart

Hashboard not detected
│
├─ Reseat connector — detected now?
│  ├─ YES → Connector issue. Clean and monitor.
│  └─ NO ↓
│
├─ Swap board to different slot — follows the board?
│  ├─ NO (stays on slot) → Control board port or cable fault.
│  │  └─ Try a new cable. If still fails, control board
│  │     port is damaged. See "Control Board Port Failure."
│  └─ YES (follows board) ↓
│
├─ Check for short on hashboard power input.
│  ├─ SHORT detected → Find and repair shorted component.
│  │  See: Abnormal Chip Voltage guide for domain testing.
│  └─ No short ↓
│
├─ Check EEPROM — can it be read?
│  ├─ NO → EEPROM chip failure or I2C bus fault.
│  │       Replace EEPROM or repair I2C pull-ups.
│  └─ YES ↓
│
├─ Check PIC microcontroller (Antminer) / management chip.
│  ├─ PIC not responding → PIC failure. Replace PIC chip.
│  └─ PIC responds ↓
│
├─ Check signal chain — can the first chip be addressed?
│  ├─ NO → First chip in chain is dead, or CLK/CI trace
│  │       from connector to first chip is broken.
│  └─ YES → Partial detection issue. See Missing Chips guide.
│
└─ After all checks: if board still not detected with
   known-good control board, cable, and no shorts,
   the board likely has a fundamental power or
   communication circuit failure requiring board-level repair.

Common Causes (Ordered by Probability)

1. Loose or Poorly Seated Connector

Probability: ~30%

The most common cause is simply a connector that is not fully seated. Thermal cycling (heating and cooling over days/weeks) causes connectors to slowly work loose. Vibration from fans compounds this.

Indicators:

• Problem is intermittent — board appears and disappears across reboots
• Problem appeared after the miner was moved or shipped
• Reseating the connector fixes the issue

Fix: Reseat the connector firmly. If the connector has a locking tab, ensure it clicks into place. If the problem recurs, the connector receptacle may be worn — replace the connector or use a small amount of hot glue to secure it (easily removable for future maintenance).

2. Damaged Connector Pins

Probability: ~15%

Bent, corroded, or broken pins in the connector prevent electrical contact. This often occurs from rough handling, forced insertion at an angle, or liquid exposure.

Indicators:

• Reseating does not help
• Visual inspection shows one or more pins that are bent, missing, or discolored
• Melted plastic on the connector housing (indicates a pin was arcing under load)

Fix: For bent pins, carefully straighten with fine tweezers or a pin straightening tool. For broken or missing pins, replace the entire connector. If the connector housing is melted, replace both the connector and the mating receptacle — heat damage may have warped the alignment.

3. EEPROM Corruption or Failure

Probability: ~15%

Each hashboard contains an EEPROM chip (typically AT24C02 or similar) that stores the board's identity, calibration data, and configuration. The control board reads this EEPROM during startup to identify and configure the board. If the EEPROM cannot be read, the board is treated as absent.

# SSH into the miner
ssh root@<miner-ip>

# Check kernel log for EEPROM errors
dmesg | grep -i eeprom

# Attempt to read EEPROM directly
i2cdump -y 0 0x50

ssh root@<miner-ip>
cat /tmp/btminerlog | grep -i eeprom

Fix: If the EEPROM chip is dead, replace it and reprogram using data from a known-good board of the same model (adjust the serial number). If the data is corrupt but the chip is functional, reprogram it. EEPROM programming requires an I2C programmer or can be done through the miner's diagnostic interface on some firmware versions.

4. PIC Microcontroller Failure (Antminer)

Probability: ~12%

Antminer hashboards use a PIC microcontroller (PIC16F1704 on newer models) as an intermediary between the control board and the hashboard. The PIC handles voltage enable, temperature monitoring, and board identification. If the PIC fails, the control board cannot communicate with the hashboard at all.

Indicators:

• Board is not detected even after connector/cable swap
• EEPROM reads OK when tested with an external programmer, but the miner still cannot see the board
• No voltage appears on the hashboard power domains (PIC has not issued the voltage enable command)

How to diagnose (requires I2C scanner):

• The PIC responds at I2C address 0x20–0x27
• If no device responds in that range, the PIC is not functioning

Fix: Replace the PIC microcontroller. This requires soldering equipment and a pre-programmed replacement PIC (the firmware is model-specific). See PIC Microcontrollers for details.

5. Control Board Port Failure

Probability: ~10%

The control board has multiple ports (typically 3) for hashboards. If a specific port's driver circuitry fails, any board plugged into that port will not be detected.

Indicators:

• The problem stays on the same slot regardless of which board is installed
• The board works fine in a different slot on the same miner or in a different miner
• Other ports on the same control board work correctly

How to diagnose:

• Swap boards between slots as described in Quick Checks
• Try a new data/signal cable between the control board and the slot

Fix: Control board port repair is typically not cost-effective. Options:

• Use the remaining working ports (run with 2 boards instead of 3)
• Replace the control board
• On some models, the port driver IC can be replaced by a skilled technician

6. Hashboard Short Circuit

Probability: ~8%

A severe short circuit on the hashboard's power input will prevent the board from powering up. The PSU's overcurrent protection may trip, or the control board may refuse to enable power to the board.

Indicators:

• Resistance measurement across power input shows less than 0.5 ohms
• PSU clicks or restarts when the board is installed (overcurrent protection tripping)
• Burnt smell or visible burn marks on the board

Fix: Locate the short circuit. Common locations:

• Input bulk capacitor (shorted electrolytic or ceramic cap)
• Voltage regulator with a catastrophic failure
• ASIC chip with internal short (especially near the power input side)

Use a thermal camera or apply low voltage with a current-limited bench supply to find the component drawing excessive current (it will heat up). Replace the faulty component.

7. Broken Signal Trace (CLK/CI/RI)

Probability: ~5%

The communication path from the connector to the first ASIC chip may have a broken trace. Without this signal path, the control board cannot establish any communication with the chip chain.

Indicators:

• Board powers up (you can measure voltage on domains) but is not detected
• No short circuit present
• EEPROM and PIC are functional (on models where these can be tested independently)

Fix: Trace the CLK, CI (Clock In), and RI (Receive In) signals from the connector to the first chip using a multimeter in continuity mode. Repair any broken trace with a fine jumper wire.

8. Data Cable Failure

Probability: ~5%

The ribbon cable or wire harness between the control board and the hashboard can develop internal breaks, especially at stress points near the connectors.

Indicators:

• Problem follows the cable, not the board or the slot
• Flexing the cable changes the detection status
• Visible damage to the cable

Fix: Replace the cable. Use an OEM or compatible replacement — do not splice data cables.

Board-Swap Diagnostic Matrix

This table summarizes how to interpret board-swap results:

How to Access Error Logs

# SSH into the miner (default password: admin or root)
ssh root@<miner-ip>

# View kernel log for detection errors
dmesg | grep -i chain

# View miner application log
cat /tmp/log/bmminer.log | tail -100

# Check for EEPROM and PIC errors
dmesg | grep -iE "eeprom|pic|i2c"

# SSH into the miner
ssh root@<miner-ip>

# View main miner log
cat /tmp/btminerlog | tail -200

# Check for hardware detection errors
cat /tmp/btminerlog | grep -iE "board|chain|detect"

When to Seek Professional Help

Consider professional repair in these situations:

• The PIC microcontroller needs replacement and you do not have a pre-programmed replacement chip or the tools to solder the QFP/TQFP package
• EEPROM needs reprogramming but you do not have the original calibration data for the specific board
• The control board port has failed and you want to attempt port-level repair rather than replacing the entire control board
• Multiple boards are not detected on a single miner, suggesting a control board or PSU issue that requires deeper diagnosis
• The board has liquid damage affecting multiple components around the connector area

Frequently Asked Questions

Can I run a miner with only 2 out of 3 boards?

Yes. Most miners will operate with fewer boards. Hashrate and power consumption will be reduced proportionally. Some stock firmware versions may report warnings but will continue mining. Third-party firmware generally handles this gracefully.

Why does the board appear after multiple reboots but then disappear?

This is classic intermittent connector contact. The thermal expansion from heating pushes the connector just enough to lose contact. Clean the connector pins, ensure a firm seat, and consider applying a small dab of hot glue to prevent the connector from working loose.

My board was working yesterday but is not detected today — what changed?

In order of likelihood: the connector worked loose from thermal cycling, a marginal solder joint on the EEPROM or PIC finally failed, or a voltage regulator failed overnight causing a power-up short that prevents detection. Start with reseating the connector, then check for shorts.

How do I know if the EEPROM data is correct even if the chip reads OK?

Compare the EEPROM dump to a known-good board of the same model. Key fields include the board model identifier, chip count, and voltage calibration values. If any critical field is zeroed out or contains unexpected values, the data is corrupt. Some technicians maintain a library of known-good EEPROM dumps for each model they service.

Can a bad PSU cause a board to not be detected?

Yes. If the PSU cannot deliver enough power for all three boards to complete their power-up sequence simultaneously, one or more boards may fail to initialize. This is more common on older or undersized PSUs. Try powering up with only the missing board installed — if it is detected alone but not with the other two boards, the PSU cannot handle the combined inrush current.

Related Guides

• Missing Chips Troubleshooting
• Abnormal Chip Voltage
• EEPROM Chips
• PIC Microcontrollers
• How Hash Boards Work
• Antminer Error Codes
• Whatsminer Fault Codes