Innosilicon A4 Dominator — Maintenance & Repair Guide

Technical Specifications

Before you touch a screwdriver, know your machine. The Innosilicon A4 Dominator has a distinct hardware profile that sets it apart from the Bitmain Antminer and MicroBT Whatsminer families. Understanding these specifications is essential for accurate diagnostics and effective maintenance.

Before You Begin

Critical Safety Warnings

Routine Maintenance

Preventive maintenance is what separates a miner that runs for years from one that dies in months. The A4 is a well-built machine, but it was designed for data center environments with controlled airflow and regular maintenance cycles. Home mining setups — garages, basements, spare rooms — expose the miner to dust, humidity, temperature swings, and pet hair that Innosilicon never designed for. Regular maintenance compensates for that gap.

Recommended Maintenance Schedule

Visual Inspection

Start every maintenance session with a thorough visual inspection. Power down the miner, disconnect the PSU, and wait 60 seconds. Then examine the following:

• Exterior chassis: Look for physical damage, dents, corrosion, or rust on the metal casing. Check that all screws are present and tight. Loose screws can cause vibration noise and poor grounding.
• Fan grilles: Inspect intake and exhaust grilles for dust buildup, blockages, or debris. A clogged grille restricts airflow and forces the remaining fans to work harder, reducing their lifespan.
• Power connectors: Examine all 6-pin PCIe power connections. Look for discoloration, melting, or burn marks on connectors — these indicate loose connections or overloaded cables. A discolored connector must be replaced immediately.
• Ethernet port: Verify the Ethernet cable is firmly seated. Check for bent pins or corrosion in the RJ45 jacks.
• Hashboard connectors: If you open the chassis, inspect the ribbon cables and power connectors between hashboards and the control board. Loose hashboard connectors are a common cause of chain detection failures on the A4.
• PCB surfaces: Look for any signs of liquid damage, corrosion (green/white residue), burnt components, or bulging capacitors on both hashboards and the control board.

Cleaning Procedures

Dust is the number one killer of home-operated miners. The A4’s fans pull air across the hashboards continuously, and every particle in that air eventually deposits on heatsinks, PCBs, and fan bearings. Here is how to clean properly:

Quick External Clean (Every 1–3 Months)

1. Power down and disconnect from PSU. Wait 60 seconds.
2. Take the miner to a well-ventilated area (outside or a garage is ideal — you will be blowing dust everywhere).
3. Using compressed air held 6–8 inches from the unit, blow through the intake side in short bursts. The goal is to push dust out through the exhaust side, following the original airflow direction.
4. Clean fan blades individually — hold each fan still with a finger while blasting compressed air (letting fans free-spin during cleaning can generate back-EMF that damages the motor driver on the control board).
5. Use a soft-bristle brush to dislodge stubborn dust from grilles and heatsink fins.

Deep Internal Clean (Every 6 Months)

1. Remove the top and side panels by unscrewing the Phillips-head screws along the chassis edges.
2. Carefully disconnect and remove the front and rear fan assemblies.
3. Remove the hashboards by disconnecting their ribbon cable and power connectors from the control board. Note the orientation and slot positions — the A4 has two hashboard slots and they should return to the same positions.
4. Using compressed air and a soft brush, clean each hashboard individually. Pay special attention to the gaps between heatsink fins where dust accumulates.
5. Clean the control board, focusing on the Ethernet ports, header pins, and any connector sockets.
6. Inspect and clean the interior of the chassis — dust collects on the bottom surface and in corners.
7. Clean each fan with compressed air and a brush. Check for bearing noise by spinning the fan gently with your finger — a grinding or clicking sound indicates a worn bearing.
8. Reassemble in reverse order, ensuring all connectors are firmly seated.

Thermal Paste & Thermal Pad Replacement

The A4 uses thermal adhesive (thermal glue) to bond heatsinks to the ASIC chips. Over time, this compound dries out, cracks, and loses its thermal conductivity. When chip temperatures rise despite clean heatsinks and functional fans, degraded thermal interface material is usually the cause.

1. Remove heatsinks carefully. The A4 uses thermal adhesive, not clips, to hold heatsinks. Gently pry heatsinks off using a plastic spudger. Do NOT use a metal flathead screwdriver — you will scratch the PCB traces or chip package. If a heatsink is stubbornly bonded, apply gentle warmth with a heat gun at low setting (150°C) for 15–20 seconds to soften the adhesive before prying.
2. Clean old thermal compound completely. Use 99% isopropyl alcohol and lint-free wipes to remove all residue from both the chip surface and the heatsink base. The surfaces must be completely clean and dry before applying new compound.
3. Apply fresh thermal paste. Use a high-quality non-conductive thermal paste (Arctic MX-4 or Noctua NT-H1 are excellent choices). Apply a thin, even layer to the chip surface — about the size of a grain of rice for each chip, then spread it evenly. Too much paste is as bad as too little; you want a thin uniform layer, not a thick glob.
4. Reattach heatsinks. Press each heatsink firmly and evenly onto the chip. Ensure full contact across the entire chip surface. If using thermal adhesive for re-bonding, apply a thin line around the heatsink perimeter (not on the thermal interface area).
5. Cure time. Allow thermal paste to settle for 10–15 minutes before powering on. Run the miner and monitor chip temperatures for the first 30 minutes to confirm temperatures have dropped.

Fan Maintenance & Replacement

The A4 uses four 120mm fans — two on the intake side and two on the exhaust side. Fan failure is one of the most common issues on aging A4 units. Here is what to watch for and how to address it:

• Noise changes: A healthy fan produces consistent airflow noise. Clicking, grinding, rattling, or whining indicates bearing wear. Replace the fan before it fails completely — a dead fan means the adjacent hashboard runs hot and throttles.
• RPM monitoring: Check fan RPM readings in the web interface under Miner Status. A fan reading 0 RPM is obviously dead. A fan reading significantly lower RPM than its partners is dying.
• Replacement procedure: The A4’s fans are connected via standard 2-pin or 4-pin headers. Disconnect the old fan, unscrew it from the chassis, install the replacement with the correct airflow direction (arrow on fan housing indicates airflow direction), and reconnect the header. Ensure the replacement fan matches the original specs: 120mm, 12V DC, equivalent or higher CFM rating.
• Bearing lubrication (temporary fix): If a fan is clicking but still spinning, you can extend its life by peeling back the label on the fan hub, adding a single drop of light machine oil to the bearing shaft, and replacing the label. This is a stopgap — plan to replace the fan within 1–2 months.

Diagnostics & Troubleshooting

When an A4 misbehaves, systematic diagnosis is the difference between a quick fix and days of frustration. The A4’s diagnostic interface is different from Antminer machines — there is no kernel log in the same format, no BMMiner or CGMiner status API. Instead, you are working with Innosilicon’s own firmware stack. Here is how to approach it.

LED Indicators

The A4’s front panel has LED status indicators that provide quick visual diagnostic information:

Web Interface Diagnostics

The A4’s LuCI-based web interface is your primary diagnostic dashboard. Access it by navigating to the miner’s IP address in any web browser and logging in with the default credentials (root/root).

Key diagnostic pages:

• Miner Status > Miner Overview: Displays real-time hashrate, chip temperatures, fan speeds, voltage readings, and pool connection status for each hashboard. This is your first stop for any performance issue.
• Miner Status > Pool Status: Shows accepted/rejected share counts, pool connection health, and current difficulty. High rejection rates indicate connectivity issues or stale work.
• System > System Log: Contains timestamped entries for hardware events, errors, and state changes. Look here for chain detection failures, temperature warnings, and communication timeouts.
• System > Kernel Log: Lower-level kernel messages including driver initialization, device detection, and hardware interrupts. Useful for diagnosing boot failures and hashboard detection issues.

SSH Diagnostic Commands

For deeper diagnostics, SSH into the miner directly. The A4 runs OpenWrt, so you have access to standard Linux utilities plus Innosilicon-specific tools.

# Connect via SSH (default password: root)
ssh root@[MINER_IP_ADDRESS]

# Check miner process status
ps | grep cgminer

# View real-time system log
logread -f

# Check hashboard detection
cat /sys/class/hwmon/hwmon*/temp*_input

# View network configuration
ifconfig eth0

# Check memory and CPU usage
free -m
top -n 1

# View kernel messages (hashboard init)
dmesg | tail -50

# Check uptime and load average
uptime

# View pool configuration
cat /etc/config/cgminer

# Restart the mining process
/etc/init.d/cgminer restart

Common Error Codes & Their Meanings

The A4 reports errors through both the web interface and system logs. Innosilicon uses alphanumeric error codes that can also be viewed using the InnoMonitor batch management tool. Here are the most common ones you will encounter:

Hashboard Testing with Normal.bin and Repair.bin

Innosilicon provides specialized test software for diagnosing hashboard issues at the chip level. This is the A4’s equivalent of Bitmain’s test fixture process, but software-driven rather than hardware-fixture-driven.

Normal.bin — Diagnostic Testing

1. Download the Normal.bin test firmware for the A4 from Innosilicon’s support resources (or your archived copy — these files are increasingly hard to find).
2. Flash the Normal.bin firmware to the miner via the web interface: System > Upgrade > Flash Image.
3. Once flashed, the miner reboots and runs automated diagnostic routines across both hashboards.
4. Access the test results via the web interface or SSH. The software reports: chip count detected per chain, voltage per domain, temperature per sensor, and communication status for each chip in the daisy chain.
5. A healthy A4 hashboard should report all chips detected with uniform voltage readings across the chain. Missing chips, voltage anomalies, or communication timeouts indicate specific failure points.
6. Document all test results before proceeding to repair. Note which chain position(s) show failures.

Repair.bin — Chip Reset and Recovery

1. Flash the Repair.bin firmware using the same process as Normal.bin.
2. Repair.bin sends extended communication commands to each chip in the chain, attempting to reset stuck or unresponsive chips.
3. This software can recover chips that have entered a fault state but are not physically damaged — think of it as a hard reset at the silicon level.
4. After running Repair.bin, reflash the Normal.bin to re-test, or restore the production firmware to resume mining.
5. If Repair.bin cannot recover a chip, that chip has a physical failure (cracked die, broken bond wire, ESD damage) and requires physical replacement.

Common Repairs

The A4 is a repairable machine — more so than some modern miners where everything is soldered to oblivion with 0201-size components. But Innosilicon’s architecture has its own quirks. Here are the most common repairs and how to approach them.

Fan Replacement

Fan failure is the most common hardware issue on any aging ASIC miner, and the A4 is no exception. The dual-intake, dual-exhaust design means losing one fan degrades cooling performance by roughly 25%, causing the adjacent hashboard to run hotter and potentially throttle.

1. Identify the failed fan using the web interface RPM readings or by visual/audio inspection.
2. Power down and disconnect from PSU. Wait 60 seconds.
3. Remove the appropriate chassis panel (front for intake fans, rear for exhaust fans) — typically 4 Phillips-head screws per panel.
4. Disconnect the fan cable from the control board header. Note which header position corresponds to which fan slot.
5. Unscrew the fan from its mounting bracket (4 screws per fan, one at each corner).
6. Install the replacement fan with the correct airflow direction — the arrow printed on the fan housing shows airflow direction. Intake fans should pull air IN; exhaust fans should push air OUT.
7. Reconnect the fan cable to the same header position.
8. Reassemble the chassis panel.
9. Power on and verify the new fan is reporting correct RPM in the web interface.

Power Supply Troubleshooting

The A4 requires a PSU capable of delivering at least 1,200W with six 6-pin PCIe connectors. Underpowered or failing PSUs are a frequent source of instability, hashrate drops, and E02 errors.

Symptoms of PSU problems:

• Miner fails to power on or reboots randomly
• Hashrate drops under load (PSU cannot sustain full power draw)
• E02 error codes in system log
• Discolored or melted 6-pin PCIe connectors
• Audible buzzing, clicking, or whining from the PSU
• Burning smell (immediately disconnect if detected)

Diagnostic steps:

1. With the miner unpowered, inspect all 6-pin PCIe connectors on both the PSU and miner side. Look for bent pins, discoloration, or melting. Replace any damaged cables immediately.
2. Using a multimeter, measure the PSU output voltages on the 6-pin connectors. You should see +12V DC ±5% (acceptable range: 11.4V to 12.6V). Voltage below 11.4V under load indicates a failing PSU.
3. If the PSU has a built-in fan, verify it is spinning. A PSU with a dead fan will overheat and shut down under load.
4. Test with a known-good PSU if available. Swapping PSUs is the fastest way to confirm whether the issue is power-related or miner-related.

Hashboard Issues

Hashboard problems on the A4 are more complex than on Antminers due to the daisy-chain chip topology. On an Antminer, a dead chip in one voltage domain reduces hashrate proportionally. On the A4, a dead chip can break the entire communication chain, causing the whole hashboard to report as missing.

Common hashboard failure modes:

• Chain detection failure (E04): The control board cannot detect one or both hashboards. Start by reseating the ribbon cable and power connectors. If the issue persists, the hashboard has a broken communication chain — likely a dead chip or a failed communication line component (resistor, capacitor, or trace).
• Reduced chip count: The hashboard is detected but reports fewer chips than expected. This indicates one or more chips in the chain have failed and been bypassed by the firmware. Hashrate will be proportionally reduced.
• Low hashrate on one board: If one board runs significantly slower than the other with a full chip count, suspect degraded thermal performance (dried thermal paste), a chip running at reduced frequency due to temperature throttling, or a marginal solder joint that causes intermittent communication errors.
• High hardware error rate: HW errors above 1–2% indicate a chip producing incorrect results. This is often a thermal issue (chip overheating and producing errors) or a chip nearing end-of-life. Check thermal paste and heatsink contact first.

Chip-level repair process (advanced):

1. Run Normal.bin diagnostics to identify the specific chip position that has failed.
2. Locate the chip on the hashboard using the board layout reference. Innosilicon chip positions are numbered sequentially along the chain.
3. Inspect the chip visually for cracks, discoloration, or damaged solder balls.
4. If replacing: use a hot air rework station at 350–380°C with medium airflow to remove the failed chip. Protect surrounding components with Kapton tape and aluminum foil shields.
5. Clean the pads thoroughly with flux, solder wick, and IPA.
6. Place the replacement chip with correct orientation (pin 1 marker alignment) and reflow using the hot air station.
7. After cooling, clean flux residue with IPA and inspect the solder joints under magnification.
8. Re-test with Normal.bin to verify the repair.

Network & Control Board Issues

The A4’s control board handles network communication, pool configuration, hashboard management, fan control, and the web interface. When it fails, the miner either does not boot, loses network connectivity, or cannot communicate with the hashboards.

Common control board issues:

• No network connectivity: Verify the Ethernet cable and port. Try each of the 4 LAN ports on the A4’s rear panel — if one port fails, another may still work. If no ports respond, the Ethernet controller on the control board may have failed.
• Firmware corruption: If the miner boots but the web interface is inaccessible or displays errors, firmware corruption is likely. Attempt a firmware reflash via TFTP recovery mode (hold the reset button during power-on for 5 seconds).
• Hashboard detection failure (both boards): If both hashboards fail detection simultaneously, the control board’s hashboard communication interface may have failed. Test each hashboard in a known-good A4 to confirm whether the issue is the boards or the controller.
• Random reboots: Can indicate NAND flash degradation (common in aging OpenWrt devices), RAM failure, or overheating of the SoC on the control board. Check for adequate airflow over the control board and inspect for bulging capacitors.

Firmware & Software Management

The A4’s firmware is based on OpenWrt with Innosilicon’s custom mining daemon. Managing firmware on an older miner like the A4 requires extra caution — Innosilicon has not released new firmware for this model in years, and the files floating around the internet are of questionable provenance.

Firmware Updates & Recovery

Checking current firmware version:

1. Log into the web interface.
2. Navigate to System > System — the firmware version and build date are displayed at the top.
3. Alternatively, via SSH: cat /etc/openwrt_release

Firmware update procedure:

1. Back up your current firmware first (see the backup command in the SSH diagnostics section above).
2. Download the firmware file (a .bin or .tar file, depending on the release).
3. Navigate to System > Upgrade in the web interface.
4. Click “Choose File” and select the firmware file.
5. Uncheck “Keep settings” only if you need a clean slate (this will reset pool configuration, network settings, and passwords).
6. Click “Flash image” and confirm.
7. Wait for the process to complete — the miner will reboot automatically. Do NOT power off during flashing. This takes 2–5 minutes.
8. After reboot, log in and verify the new firmware version. Reconfigure pool settings if you chose not to keep settings.

TFTP recovery (for bricked miners):

1. If the web interface is inaccessible and SSH does not respond, the firmware may be corrupted.
2. Set up a TFTP server on your computer (Tftpd64 on Windows, built-in on macOS/Linux).
3. Set your computer’s IP to 192.168.1.2 (or the appropriate subnet for the A4’s recovery mode IP, typically 192.168.1.1).
4. Place the firmware .bin file in the TFTP root directory.
5. Connect the miner directly to your computer via Ethernet.
6. Power on the miner while holding the reset button for 5–10 seconds.
7. The miner will enter recovery mode and attempt to pull firmware from the TFTP server.
8. Monitor the TFTP server log for transfer activity. Once complete, the miner reboots with fresh firmware.

Configuration Best Practices

Optimizing your A4’s configuration ensures maximum efficiency and stability. These settings are configured through the web interface under Miner Configuration.

• Pool configuration: Always configure at least 3 mining pools in priority order. If your primary pool goes down, the miner automatically fails over to the next pool. Use the format stratum+tcp://pool.address:port with your worker name and password.
• Frequency settings: The A4’s default frequency is factory-optimized for the rated 280 MH/s. Increasing frequency beyond the default can yield marginal hashrate gains but increases power consumption, heat, and chip stress disproportionately. For a miner of this age, running at or slightly below stock frequency is the wisdom play — you are optimizing for longevity, not benchmarks.
• Fan speed: Auto mode works for most environments. If your ambient temperature is consistently above 30°C, consider setting fans to 100% manually to maximize cooling. For home miners in cold climates (Canadian basements in winter, for instance), you can let auto mode handle it — lower fan speeds mean less noise.
• Network settings: Use a static IP assignment (either via the miner’s config or a DHCP reservation on your router) to prevent the miner from changing IPs after a power outage or router restart.

Advanced Board-Level Diagnostics

This section is for experienced repair technicians. If you do not have a multimeter, soldering equipment, and familiarity with PCB-level troubleshooting, skip to the FAQ section. For board-level issues that are beyond your skill level, D-Central’s professional repair service is the safer path.

Impedance Testing

Impedance testing helps identify short circuits, open circuits, and degraded components on the hashboard without powering the board on.

1. Set your multimeter to resistance (ohms) mode.
2. With the hashboard completely disconnected from the miner and PSU, measure resistance across the power input connector pins.
3. Compare readings between the two hashboards — they should be very similar (within 10%). A dramatically lower reading on one board suggests a short circuit. A dramatically higher reading suggests an open circuit or broken trace.
4. Measure impedance across individual chip power rails if the board layout is known. Each chip should show similar impedance. An outlier indicates a failed chip or associated passive component.

Voltage Domain Testing

1. With the hashboard installed and the miner powered on, set your multimeter to DC voltage mode.
2. Measure voltage at the hashboard power input connector. Expected value: 12V ±5%.
3. If the board has test points for individual chip voltage rails, measure each one. Voltage should be consistent across all chips in the chain. A chip showing significantly lower or zero voltage is likely failed.
4. Record all readings for comparison with future tests — trending voltage data over time can reveal gradual degradation before a chip fails completely.

DEBUG Port Diagnostics

The A4 hashboard includes a DEBUG header that connects to the chip communication chain. Using a USB-to-serial adapter connected to this port, you can communicate directly with the chips on the board independently of the control board.

1. Connect a USB-to-serial adapter (3.3V logic level) to the hashboard DEBUG header. Pin orientation matters — consult the A4 hashboard schematic for TX, RX, and GND pin positions.
2. Open a serial terminal (PuTTY, Tera Term, or minicom) at the appropriate baud rate (typically 115200 for Innosilicon boards).
3. Send chip detection commands to enumerate all chips on the chain. Each chip should respond with its address and status.
4. Chips that do not respond are either physically dead or have a broken communication path upstream. Start from the first unresponsive chip and work backward along the chain to find the actual failure point.

Reassembly & Burn-In Testing

After any repair or deep maintenance, proper reassembly and a structured burn-in test ensure the miner is stable before returning it to production.

Reassembly Procedure

1. Hashboard installation: Slide hashboards back into their chassis slots. Ensure they seat fully and evenly. The A4 has a specific orientation for each board — if the connectors do not align, the board is upside down or in the wrong slot.
2. Connector check: Reconnect all ribbon cables and power connectors between hashboards and control board. Each connector should click or seat firmly. Give each cable a gentle tug to verify it is locked in — a loose ribbon cable is the most common post-repair failure.
3. Fan installation: Mount all four fans with correct airflow direction. Reconnect fan cables to the correct headers on the control board.
4. Chassis closure: Replace all panels and tighten all screws. Loose panels cause rattling under vibration and reduce airflow efficiency by allowing air to bypass the hashboards.
5. Final visual check: Before powering on, verify no tools, screws, or loose components are left inside the chassis. A dropped screw on a powered hashboard can cause a catastrophic short circuit.

Burn-In Test Protocol

1. Connect the miner to PSU and network. Power on.
2. Monitor the web interface for the first 30 minutes:
   • Both hashboards should be detected and reporting chips.
   • Hashrate should ramp up to near 280 MH/s within 5–10 minutes.
   • Chip temperatures should stabilize below 85°C.
   • Fan RPMs should be consistent and within normal range.
   • No error codes in the system log.
3. Continue monitoring for 24 hours. Log hashrate, temperature, and fan speeds at intervals. A stable miner will show:
   • Hashrate variance of less than ±5% from average.
   • No hashboard drops or chain failures.
   • Temperature within 5°C of initial stabilized reading.
   • Zero or near-zero hardware error rate (HW errors < 0.1% of accepted shares).
4. If the miner passes the 24-hour burn-in with no issues, it is ready for production deployment.

Frequently Asked Questions

When to Call a Professional

Not every repair is a DIY project. The A4’s Innosilicon-specific architecture, proprietary chip design, and aging component availability make some repairs best left to experienced ASIC repair technicians. Here are the situations where professional help is the right call:

• Chip-level hashboard repair: Replacing daisy-chained Innosilicon Scrypt chips requires a hot air rework station, the correct replacement chips, and experience with BGA/QFN rework. One wrong move can damage adjacent chips or lift PCB pads, turning a one-chip repair into a board replacement.
• Control board failure: If the control board SoC, Ethernet controller, or NAND flash has failed, repair requires specialized equipment and replacement ICs that are not commonly available.
• Firmware recovery failure: If TFTP recovery does not work and the miner is bricked, there may be hardware-level NAND flash corruption requiring direct flash programming with a clip or socket programmer.
• Multiple simultaneous failures: If your miner has both hashboard and control board issues, or multiple chips have failed across both boards, the repair becomes a rebuild project that is best handled by a shop with parts inventory and test equipment.
• Unknown or electrical damage: If the miner was exposed to a power surge, lightning strike, or liquid spill, the extent of damage may be far beyond what is visible. Professional diagnosis can identify all affected components before you invest in partial repairs.

Final Notes

The Innosilicon A4 Dominator is a machine from a different era of Scrypt mining — built before Dogecoin became a household name, before merged mining was standard, and before the current generation of highly efficient Scrypt ASICs. But it is a well-engineered piece of hardware that, with proper maintenance, can continue hashing for years beyond what most people expect from “old” mining equipment.

The key to longevity is consistent preventive maintenance: keep it clean, keep it cool, keep the thermal interface material fresh, and do not push it beyond its limits with aggressive overclocking. Monitor your error rates and temperatures regularly. Replace fans at the first sign of bearing wear rather than waiting for complete failure. And when a repair exceeds your skill level or the cost-benefit equation no longer works, know when to call in professional help.

At D-Central Technologies, we have been working with miners like you since 2016 — from small home setups to large-scale operations. Whether you need a repair estimate, replacement parts, or just an honest opinion on whether your A4 is worth fixing, our team is here to help. We are Bitcoin Mining Hackers, and keeping old hardware alive and hashing is exactly what we do.

Questions about your A4? Call us at 1-855-753-9997 or visit d-central.tech/asic-repair.

Interactive Hashboard Schematic

Explore the INNOSILICON A4 hashboard layout below. Toggle layers to isolate voltage domains, signal chains, test points, key components, and thermal zones. Hover over any region for quick specs — click for detailed diagnostics, failure modes, and repair guidance.