Missing ASIC Chips — Troubleshooting Guide

Diagnose and fix missing ASIC chips on your miner. Covers partial hashrate, incorrect chip counts, and specific chip position failures across Antminer, Whatsminer, and Avalon models.

Symptoms

You are experiencing one or more of the following:

Hash board reports fewer chips than expected (e.g., "72 of 76 chips" on an S19)
Hashrate is lower than normal but the board is still partially working
Web UI shows "XX" or "x" in specific chip positions
Specific chip positions consistently fail across reboots
Error messages like "Chain X: only found Y chips" in miner logs
One or more chips report abnormal frequency or temperature

Safety First: Hashboard repair involves working near high-current, low-voltage circuits. Always disconnect power and wait at least 60 seconds before handling hash boards. Wear an ESD wrist strap to prevent static damage to ASIC chips.

Expected Chip Counts by Model

Before troubleshooting, confirm how many chips your specific model should detect per hash board:

Bitmain Antminer

Model	ASIC Chip	Chips per Board	Boards	Total Chips
S9	BM1387	63	3	189
S17	BM1397	36	3	108
S17 Pro	BM1397	42	3	126
S19	BM1398	76	3	228
S19 Pro	BM1398	114	3	342
S19j Pro	BM1398	126	3	378
S19k Pro	BM1398	77	3	231
S19 XP	BM1366	110	3	330
S21	BM1368	156	3	468
S21 Pro	BM1370	120	3	360
T21	BM1368	117	3	351

MicroBT Whatsminer

Model	ASIC Chip	Chips per Board	Boards	Total Chips
M30S	—	78	3	234
M30S++	—	82	3	246
M50	—	84	3	252
M50S	—	86	3	258
M56S	—	78	3	234
M60	—	110	3	330
M60S	—	86	3	258

Canaan Avalon

Model	ASIC Chip	Chips per Board	Boards	Total Chips
A1246	A1246	78	3	234
A1346	A1346	80	4	320
A1366	A1366	82	4	328
A1466	A1466	88	4	352

Some chip counts vary between hardware revisions. Check your specific batch version if the numbers do not match exactly. The manufacturer's specification sheet for your serial number range is the authoritative source.

Quick Checks

How many chips are missing? Log into the web UI and note the exact chip count per chain. One or two missing chips is a different problem than 20 missing chips.
Is the pattern consistent? Reboot the miner 2-3 times and check if the same chip positions are always missing. Random variation suggests a loose connection; consistent missing positions suggest a hardware failure.
When did it start? If it started after shipping, the cause is likely a connector or physical damage issue. If it degraded gradually, a chip or regulator is failing.
Check temperatures. If the missing chips are in a region of the board that also shows high temperatures, thermal damage may be the cause.

Diagnostic Flowchart

Missing chips detected
│
├─ How many chips missing?
│  │
│  ├─ ALL chips on one board (0 ASIC)
│  │  └─ See "No Hashrate" guide → likely connector or board failure
│  │
│  ├─ Large block of chips missing (>10 consecutive)
│  │  └─ Likely a voltage domain failure
│  │     ├─ Measure voltage across the affected domain
│  │     ├─ 0V → Check voltage regulator and surrounding components
│  │     └─ Voltage present but chips missing → Signal chain break
│  │
│  ├─ Small cluster of chips missing (2-5 consecutive)
│  │  └─ One failed chip breaking the daisy chain in that segment
│  │     ├─ Identify the FIRST missing chip in the sequence
│  │     └─ That chip (or its connections) is the most likely failure
│  │
│  └─ Single chip missing or scattered individual chips
│     ├─ Chip-level failure (most common)
│     ├─ Bad solder joint on that specific chip
│     └─ Thermal damage to individual chip
│
├─ Is the pattern consistent across reboots?
│  ├─ YES → Hardware failure (chip, trace, or regulator)
│  └─ NO (varies each boot) → Loose connector or marginal solder joint
│
└─ Did this start suddenly or gradually?
   ├─ SUDDENLY → Physical damage, connector issue, or power event
   └─ GRADUALLY (losing more chips over time) → Progressive chip degradation

Causes

1. Failed ASIC Chip

Probability: Very High

Individual ASIC chips can fail due to thermal stress, electrostatic discharge, or manufacturing defects. This is the single most common cause of missing chips.

Symptoms specific to this cause:

One or a few consecutive chips missing in the chain
Always the same chip position(s) across reboots
Board hashrate reduced proportionally to the number of missing chips
Remaining chips operate normally

Diagnosis:

Identify the Failed Chip Position

From the web UI or miner logs, determine which chip number(s) are missing. ASIC chips are numbered sequentially along the daisy chain, starting from chip 0 (closest to the signal connector) to chip N (end of the chain).

SSH into the miner and check the chip status:

# View chip frequency and status
cat /tmp/freq.txt

# Detailed chain information
bitmain-soc --chips

Chips marked with "x" or showing 0 frequency are failed or undetected.

Use the WhatsMiner Tool or web UI Status page. Missing chips show as gaps in the chip status array. The tool displays chip voltages and frequencies per position.

Check CGMiner API output or the web dashboard. Missing chips appear as zeroed entries in the chip status report. Avalon miners report chip status per module.

Understand the Daisy Chain Impact

ASIC chips are wired in a daisy chain for communication. When a chip fails in certain ways, it can break the chain and cause all chips after it to also be undetected. This is why you may see a block of missing chips rather than just one.

Key insight: The first missing chip in the sequence is almost always the failed one. Chips after it may be perfectly fine but cannot communicate because the chain is broken.

Visual Inspection of the Failed Chip

Remove the hash board and locate the physical chip corresponding to the missing chip number. Chip numbering usually follows the board layout from the connector end. Look for:

Discoloration or burn marks on the chip package
Cracked or chipped package
Poor solder joints (especially on BGA packages)
Solder bridges between pads

Thermal Imaging (If Available)

Power the board briefly (30 seconds) and use a thermal camera or IR thermometer to check the suspected chip. A failed short-circuited chip will be significantly hotter than its neighbors. A failed open-circuit chip will be cooler (not drawing current).

Fix: Replace the failed ASIC chip. This requires:

Hot air rework station (recommended: Quick 861DW or equivalent)
Replacement chip (same model and revision)
Solder paste and flux
Stencil for BGA chips

See the Antminer S21 Hashboard Repair Guide for detailed chip replacement procedures.

ASIC chip replacement is an advanced repair. Incorrect rework temperature or technique can damage the PCB and surrounding components. If you are not experienced with BGA rework, consider sending the board to a professional repair service.

Prevention: Maintain proper cooling (clean fans, unrestricted airflow). Avoid overclocking beyond manufacturer specifications. Use surge protection on power input.

2. Voltage Domain Failure

Probability: High

Hash boards are divided into voltage domains -- groups of chips that share a common voltage regulator. When a domain's regulator fails, all chips in that domain lose power and disappear.

Symptoms specific to this cause:

A contiguous block of chips is missing (matching a domain boundary)
Missing chip count matches the chips-per-domain count for your model
The board's total hashrate drops by roughly 1/N (where N is the number of domains)
Voltage measurement across the affected domain shows 0V or abnormal voltage

Diagnosis:

Map the Missing Chips to a Domain

Using the chip positions reported as missing, determine which voltage domain they belong to. Each domain typically contains 7-13 chips depending on the model (refer to the chip counts table above and your model's domain map).

For example, on an S19 (BM1398) with 12 domains of ~7 chips each, if chips 22-28 are all missing, they likely share a voltage domain.

Measure Domain Voltage

With the board powered on (use extreme caution), measure the voltage across the domain's power rail:

Use a multimeter set to DC voltage, low range (200mV or 2V)
Place probes across the domain's input/output capacitors
Expected voltage: 0.25V-0.40V depending on model and chip generation

Readings:

0V = Regulator not switching, or shorted chip pulling domain down
Normal voltage but chips missing = Communication chain break, not power issue
Abnormally high voltage = Regulator feedback loop broken (dangerous, can destroy chips)

Inspect the Voltage Regulator

Locate the boost/buck converter IC and surrounding components (inductors, capacitors, MOSFETs) for the affected domain. Look for:

Burnt or discolored regulator IC
Cracked or shorted MOSFETs
Damaged inductors
Shorted or open sense resistors

For more on how voltage regulators work in miners, see How Hash Boards Work.

Fix: Replace the failed voltage regulator component. This typically involves replacing the buck converter IC, damaged MOSFETs, or blown sense resistors. Requires good soldering skills and the correct replacement components.

Prevention: Avoid voltage and frequency overclocking. Ensure adequate cooling -- voltage regulators generate significant heat.

3. Broken Signal Trace or Via

Probability: Medium

The communication signals (CLK, CI/CO, RI/RO) travel through PCB traces and vias between chips. A cracked trace or failed via breaks the daisy chain at that point.

Symptoms specific to this cause:

A block of chips is missing starting from a specific position
Voltage domains in the affected area measure normal voltage
Problem appeared after physical stress (shipping, dropping, or flex from thermal cycling)
Board may work intermittently if the crack is marginal

Diagnosis:

Identify the Break Point

The daisy chain break is at or just before the first missing chip. All chips before the break communicate normally; all chips after the break are invisible.

Visual Inspection Under Magnification

Use a microscope or high-magnification loupe to inspect:

The signal traces between the last detected chip and the first missing chip
Vias in the signal path (look for cracked or disconnected via barrels)
Solder joints on the last detected and first missing chip

Multilayer hash board PCBs are susceptible to via fractures from thermal cycling.

Continuity Test

With the board powered off, use a multimeter in continuity mode to test signal traces from the output of the last detected chip to the input of the first missing chip. Test CLK, CI/CO, and RI/RO lines. No continuity confirms a broken trace or via.

Fix: Repair broken traces by scraping off solder mask and soldering a fine wire jumper across the break. Failed vias can sometimes be bypassed with a wire through the board. This is precision work that requires a microscope and fine soldering skills.

Prevention: Avoid flexing hash boards. Handle them by the edges, not the middle. Proper thermal management reduces thermal cycling stress on vias.

4. Bad Solder Joint

Probability: Medium

Solder joints can degrade over time due to thermal cycling (repeated heating and cooling). This is especially common on BGA (Ball Grid Array) packages used by most modern ASIC chips.

Symptoms specific to this cause:

Missing chips vary between reboots (intermittent)
Tapping or flexing the board slightly changes which chips are detected
Problem worsens over time as the joint degrades further
Issue may be temperature-dependent (works when cold, fails when hot, or vice versa)

Diagnosis:

Reboot Test

Reboot the miner 5+ times and log which chips are missing each time. If the same chip position appears missing in most but not all boots, a marginal solder joint on that chip is the likely cause.

Thermal Test

Run the miner and note when chips drop out. If chips disappear as the board warms up, the solder joint is expanding and losing contact. If chips appear only after warmup, the joint is making contact as it expands. Both patterns indicate a solder joint problem.

Reflow Attempt

A controlled reflow of the suspected chip using a hot air station can sometimes restore a bad solder joint. Apply appropriate flux, heat the chip to reflow temperature (typically 220-240C for lead-free solder), and allow it to cool slowly. This is a temporary fix -- the joint may fail again.

Fix: The proper fix is to remove the chip, reball it with fresh solder balls, and resolder it. Alternatively, replace the chip entirely with a new one.

Prevention: Maintain consistent operating temperatures. Avoid frequent power cycling that causes thermal stress.

5. Connector or Cable Issue

Probability: Medium-Low

A partially damaged connector or cable can cause some but not all chips to be detected. This happens when some signal pins make contact but others do not.

Symptoms specific to this cause:

Missing chip count changes when the cable is wiggled
Problem appeared after maintenance or moving the miner
Connector pins appear bent or corroded
Multiple non-adjacent chips missing (unusual pattern)

Diagnosis:

Inspect the signal connector where the hash board connects to the control board. Look for:

Bent or recessed pins
Corrosion or discoloration on contacts
Damaged ribbon cable (creases, tears, or burned spots)
Loose connector latch

Antminer ribbon cables carry multiple signal lines. A partially inserted cable may allow CLK to pass but not CI/RI, causing partial detection. Remove and firmly reseat the ribbon cable, ensuring the connector latch clicks into place.

Whatsminer 10-pin signal connectors can develop bent pins from improper insertion. Inspect each pin under magnification. A single bent pin can cause a subset of chips to not be addressed properly.

Avalon module connectors can accumulate dust or develop oxidation. Clean contacts with isopropyl alcohol and a lint-free wipe. Ensure the connector is fully seated.

Fix: Clean or replace the connector/cable. For ribbon cables, replacement is inexpensive. For board-mounted connectors, resoldering or replacement may be needed.

Prevention: Handle connectors carefully. Do not force cables. Keep the miner environment clean and dry to prevent corrosion.

6. PIC Microcontroller Communication Failure

Probability: Low

Each hash board contains a PIC microcontroller that manages board identification, temperature sensor reading, and voltage control. If the PIC loses communication, the control board may fail to properly initialize the ASIC chain.

Symptoms specific to this cause:

Board is detected but chip count is wrong or zero
Temperature readings are absent or stuck at a fixed value
Board EEPROM cannot be read (serial number / model info missing)
Problem persists even after reflashing firmware

Diagnosis:

The PIC communicates over I2C with the control board. PIC failures can be caused by:

Corrupted PIC firmware
Failed I2C pull-up resistors on the hash board
PIC chip itself is damaged

PIC microcontroller issues are relatively rare but can be confusing because the symptoms mimic other failures. If you have ruled out all other causes, the PIC is worth investigating. See PIC Microcontrollers for more information.

Fix: Re-programming the PIC requires specialized tools (PICkit programmer) and the correct firmware image for your hash board model and revision. This is typically a professional repair task.

Prevention: PIC failures are usually caused by electrical events (surges, ESD). Use proper grounding and surge protection.

Binary Search Method for Finding Failed Chips

When you have access to a test fixture or the ability to isolate chip segments, you can use a binary search approach to quickly locate a failed chip in a long chain:

Start with the Full Chain

Note the total number of chips detected vs. expected. For example, 60 of 76 chips detected means 16 chips are missing.

Identify the Last Detected Chip

From the chip status log, find the chip number where detection stops. All chips after this point are in the "missing" zone.

Focus on the Boundary

The failure is most likely at or near the boundary between detected and undetected chips. Examine the last detected chip (its output signals) and the first undetected chip (its input signals and power).

Inspect and Test

Using a multimeter and visual inspection, check the boundary chips for:

Proper power supply voltage
Signal continuity from the last good chip to the first missing chip
Physical damage or bad solder joints

The first chip that fails these checks is your culprit.

When to Seek Professional Help

Contact a qualified repair technician if:

More than one voltage domain has failed (indicates a systemic issue)
You do not have BGA rework equipment for chip replacement
The missing chips pattern does not follow any logical grouping (may indicate control board issues)
The board has visible liquid damage or heavy corrosion
You have attempted a chip reflow/replacement and the problem persists

Frequently Asked Questions

Q: Is it worth repairing a board with only 1-2 missing chips?

It depends on the economics. One missing chip on an S21 (156 chips per board) reduces that board's hashrate by less than 1%. Many operators run boards with a few missing chips rather than paying for repair. However, a single missing chip can sometimes indicate an impending larger failure, so monitor the chip count over time.

Q: Can overclocking cause missing chips?

Yes. Overclocking increases chip voltage, current, and temperature, all of which accelerate chip degradation. If you are running overclocked firmware and see missing chips appearing over time, reduce the frequency/voltage to stock settings and monitor.

Q: My miner shows the correct chip count but some chips have very low hashrate. Is that the same problem?

Not exactly. A chip that is detected but hashing slowly may have a partial internal failure, thermal throttling, or a marginal power supply. Check the individual chip frequency and temperature in the logs. See Low Hashrate for this specific scenario.

Q: After replacing a chip, do I need to recalibrate anything?

Replacement chips need to be the same model and revision as the originals. After replacement, the miner's firmware will automatically detect and configure the new chip during initialization. No manual calibration is typically required, but a full power cycle (not just a reboot) is recommended.

Q: Can humidity cause missing chips?

Yes. High humidity can cause condensation on the board, leading to corrosion of solder joints and traces over time. In extreme cases, condensation can create short circuits that destroy chips. Operate miners in environments with 20-80% relative humidity and avoid rapid temperature changes that cause condensation.

Miner Shows No Hashrate -- When all chips are missing
Abnormal Chip Voltage -- Voltage domain diagnostics
How Hash Boards Work -- Understanding chip daisy chains and voltage domains
Antminer S21 Hashboard Repair Guide -- Step-by-step repair procedures
BM1366 ASIC Chip -- Chip-specific documentation
BM1398 ASIC Chip -- Chip-specific documentation

Missing ASIC Chips — Troubleshooting Guide

On this page