Abnormal Chip Voltage — Troubleshooting Guide

Diagnose and fix voltage-related errors on ASIC mining hashboards, including out-of-range chip voltage, failed buck converters, shorted ASICs, and PSU voltage droop across Antminer, Whatsminer, and Avalon models.

Overview

Abnormal chip voltage is one of the most frequent hashboard failures in ASIC miners. Every ASIC chip on a hashboard operates within an extremely narrow voltage window — typically between 0.25V and 0.40V depending on the model. When voltage drifts outside this range, chips malfunction, hashrate drops, or the board shuts down entirely.

This guide walks through a systematic approach to diagnosing and resolving voltage-related issues across all major miner brands.

Symptoms

You are likely dealing with a voltage problem if you observe any of the following:

"Voltage error" or "Vol error" messages in the miner's web dashboard or kernel logs
A voltage domain reporting 0V (complete power loss to a group of chips)
Chips underperforming — hashrate lower than expected without thermal throttling
Domain voltage significantly higher or lower than specification
Miner shutting down a hashboard with a voltage-related fault code
Antminer: "Voltage Error" in kernel log, "Vol Err" on status page
Whatsminer: Fault codes 302, 303, or 350 referencing voltage
Avalon: "Voltage out of range" in AUC log

Quick Checks

Before diving into detailed diagnostics, perform these rapid checks that resolve the most common causes:

Check PSU Output Voltage

Measure the PSU output at the connector with a multimeter. You should see:

12V models (S19, M30S, A1246): 11.8V–12.2V under load
48V models (S21, M60S): 47V–52V under load
14V models (S19 XP): 13.6V–14.4V under load

If the PSU voltage is off, the problem is upstream — not on the hashboard.

Visual Inspection for Burnt Components

Remove the hashboard and inspect both sides under good lighting. Look for:

Darkened or charred areas around voltage regulators or MOSFETs
Bulging or cracked capacitors near voltage domains
Solder bridges or debris between pads (especially after previous repairs)
Corroded or greenish connector pins

A single visually damaged component can explain the entire voltage failure.

Reseat Power Connectors

Disconnect and firmly reseat all power connectors between the PSU and hashboard. Oxidation or a loose pin can cause intermittent voltage droop under load.

Diagnostic Flowchart

Follow this decision tree to isolate the root cause:

Voltage error reported
│
├─ Is the PSU output voltage within spec?
│  ├─ NO → Replace or repair PSU. See Power Supply Troubleshooting.
│  └─ YES ↓
│
├─ Is the error on one domain or all domains?
│  ├─ ALL DOMAINS → Check input power connector, main bus traces,
│  │               or input-side bulk capacitor.
│  └─ ONE DOMAIN ↓
│
├─ Is the domain voltage 0V or just out of range?
│  ├─ 0V → Buck converter failure or shorted ASIC.
│  │       Measure resistance across domain (see below).
│  └─ OUT OF RANGE ↓
│
├─ Is the voltage too HIGH or too LOW?
│  ├─ TOO HIGH → Buck converter feedback resistor open, or
│  │            converter running unregulated. Replace converter.
│  └─ TOO LOW → Check for shorted chip (draws excess current),
│               damaged trace (resistance in power path),
│               or corroded connector reducing current delivery.
│
└─ After fixing, does the error persist?
   ├─ YES → Possible control board issue or firmware
   │        voltage calibration error. Try reflashing.
   └─ NO → Issue resolved. Monitor for 24 hours.

Understanding Voltage Domains

A hashboard is divided into voltage domains — groups of ASIC chips that share a single power rail and voltage regulator. Each domain is an independent power delivery unit.

When one domain fails, only the chips in that domain are affected. The rest of the board may continue to operate at reduced hashrate, or the miner firmware may shut down the entire board as a safety measure.

Expected Voltages by Model

Model	ASIC Chip	Core Voltage (VDD)	I/O Voltage (VDDIO)	Domains	Chips/Domain
S9	BM1387	0.40V	1.8V	6	10–11
S17	BM1397	0.34V	1.8V	4	19
S19	BM1398	0.31V	1.8V	12	7
S19 Pro	BM1398	0.31V	1.8V	12	7
S19 XP	BM1366	0.28V	1.8V	12	11
S19k Pro	BM1366	0.28V	1.8V	12	11
S21	BM1368	0.30V	1.8V	12	13
S21 Pro	BM1370	0.29V	1.8V	16	11
T21	BM1368	0.30V	1.8V	12	13

Tolerance: VDD is typically ±0.02V. VDDIO is ±0.1V.

Model	ASIC Chip	Core Voltage (VDD)	Domains	Chips/Domain
M30S	—	0.36V	10	7–8
M30S++	—	0.34V	10	8
M50	—	0.30V	12	8
M50S	—	0.30V	12	8
M56S	—	0.28V	12	10
M60	—	0.27V	16	7
M60S	—	0.27V	16	7

Note: Whatsminer boards use a different voltage regulator topology. Voltages are measured at the domain test points on the PCB edge.

Model	ASIC Chip	Core Voltage (VDD)	Domains	Chips/Domain
A1246	A3210	0.35V	8	9
A1366	A3212	0.30V	10	10
A1466	A3214	0.28V	12	9

Note: Avalon boards often expose test pads for each domain. Refer to the silkscreen markings for pad locations.

How to Measure Voltage Domains with a Multimeter

Safety Warning: Voltage domain measurements must be taken with the board powered on, which means lethal currents are present at the PSU connector. Never touch the 12V/48V input bus while measuring. Use insulated probes and keep one hand behind your back when probing near high-current paths. See the Repair Safety Guide for complete safety procedures.

Set Up Your Multimeter

Set your multimeter to DC Voltage, lowest range that covers 1V (typically the 2V or 4V range). Use fine-tip probes or probe tip adapters to reach small test points.

Locate Domain Test Points

Each voltage domain has test points on the hashboard PCB:

Antminer S19/S21 series: Look for small circular pads along the edge of the board, labeled with domain numbers or silkscreened with "V1", "V2", etc.
Whatsminer: Test points are typically along one edge, sometimes marked with "TP" followed by a number.
Avalon: Pads are usually exposed on the top side near each voltage regulator.

If you cannot locate test points, measure directly across the VDD pins of any ASIC chip in the domain (pin 1 and the ground pad).

Measure Each Domain Sequentially

With the board installed and the miner powered on:

Place the black probe on a known ground point (any large ground pad or heatsink mounting hole)
Place the red probe on domain 1's test point
Record the reading
Move to domain 2 and repeat through all domains
Compare all readings to the expected values in the tables above

What to look for:

All domains should read within ±0.02V of each other
A domain reading 0V indicates total power failure for that domain
A domain reading significantly low (e.g., 0.15V instead of 0.30V) suggests a partially shorted chip
A domain reading higher than expected suggests a feedback circuit failure in the buck converter

Measure Domain Resistance (Power Off)

For a domain reading 0V, disconnect the board and measure resistance across the domain (VDD to ground at the test points):

Normal: 5–50 ohms (varies by number of chips in domain)
Shorted chip: Less than 1 ohm — a chip in the domain has an internal short
Open circuit: Greater than 1M ohm — the power trace is broken or the buck converter output is open

If you find a short, narrow it down by measuring resistance across individual chip VDD/GND pins, working from both ends of the domain toward the center.

Common Causes (Ordered by Probability)

1. Failed Buck Converter / Voltage Regulator

Probability: ~35%

The buck converter (or linear regulator on older models) that powers the domain has failed. This is the most common cause because voltage regulators handle the highest thermal stress after the ASIC chips themselves.

Indicators:

Domain reads 0V or significantly off-spec
The regulator IC is physically hot to the touch (even when the domain is not producing hashrate)
Visible damage to the converter IC or surrounding components

Fix: Replace the voltage regulator IC and verify surrounding passives (inductor, capacitors, feedback resistors). See your model-specific repair guide for the exact part number and replacement procedure.

2. Shorted ASIC Chip

Probability: ~25%

An ASIC chip with an internal die failure can present as a short circuit across VDD and GND. This pulls the entire domain voltage down because the regulator cannot supply enough current.

Indicators:

Domain voltage is present but low (e.g., 0.15V instead of 0.30V)
Domain resistance is abnormally low (below 2 ohms)
One chip runs noticeably hotter than its neighbors

Fix: Identify the shorted chip via resistance measurement (narrow down by halving the domain) and replace it. See How Hash Boards Work for chip daisy-chain details.

3. PSU Voltage Droop Under Load

Probability: ~15%

The PSU cannot maintain its rated voltage when all three hashboards are drawing full power. This causes all domains across all boards to read slightly low.

Indicators:

All three boards show voltage warnings simultaneously
PSU output measures below spec under load
Problem disappears when running only one or two boards
PSU fan runs at maximum speed

Fix: Replace the PSU. If using a secondary market PSU, ensure it meets the wattage requirement for your miner model with at least 10% headroom.

4. Damaged PCB Traces

Probability: ~12%

Physical damage to the copper traces that carry power to a domain. This can result from impact damage during shipping, improper handling, or corrosion.

Indicators:

Domain voltage is slightly low, especially under hashing load
Visible scratch or crack on the PCB crossing the power trace path
Problem is intermittent and position-dependent (changes if you flex the board slightly)

Fix: Repair the trace with a copper jumper wire soldered across the break. For internal layer damage, a jumper wire on the surface connecting the two sides of the break is the standard repair approach.

5. Corroded or Damaged Connector

Probability: ~10%

The power connector between the hashboard and the control board or PSU has corroded pins, bent pins, or has loosened over time from thermal cycling.

Indicators:

Intermittent voltage errors that come and go
Problem resolves temporarily after reseating the connector
Visible green/white oxidation on connector pins
Melted or discolored plastic on the connector housing

Fix: Clean connector pins with isopropyl alcohol and a fine brush. If pins are bent or the housing is damaged, replace the connector. In severe cases, replace the cable.

6. Feedback Resistor Failure (Overvoltage)

Probability: ~3%

If a voltage domain reads higher than expected, the voltage feedback network (resistor divider) that tells the buck converter what voltage to produce has failed — typically an open resistor.

Indicators:

One domain reads 0.4V+ when it should be 0.30V
Chips in that domain run extremely hot or have already been damaged by overvoltage

Critical: An overvoltage domain can permanently destroy every ASIC chip in the domain within seconds. If you detect overvoltage, shut down the miner immediately.

Fix: Identify and replace the open feedback resistor. The exact value depends on the model's voltage regulator design. Consult the schematic for your board or measure a known-good board.

When to Seek Professional Help

Consider professional repair if:

Multiple voltage domains on the same board are failed — this may indicate a systemic issue requiring schematic-level analysis
The board has suffered liquid damage with visible corrosion across multiple areas
You do not have the tools to replace BGA ASIC chips (hot air rework station)
The voltage regulator is a custom or uncommon IC that you cannot source
Overvoltage has occurred and you suspect multiple chips are damaged

Frequently Asked Questions

Can I run a miner with one voltage domain disabled?

Some firmware (Braiins OS, VNish) allows running with reduced domains. Stock firmware typically shuts down the entire board if any domain reports an error. Running with a failed domain reduces hashrate proportionally but does not damage the board.

Why did my voltage regulator fail?

Voltage regulators fail primarily from thermal stress (cumulative heat damage over thousands of hours), input voltage spikes from the PSU, or secondary failure caused by a shorted ASIC chip in their domain that forced them to deliver excessive current.

Is it safe to run a board with slightly low voltage?

If voltage is within the ±0.02V tolerance, this is normal manufacturing variation. If voltage is more than 0.03V below spec but the board is hashing without errors, it is generally safe but may indicate early-stage regulator degradation. Monitor closely.

How do I tell the difference between a PSU issue and a hashboard issue?

Measure the PSU output under load. If the PSU output is within spec but domain voltages are wrong, the problem is on the hashboard. If the PSU output itself is low, the PSU is the issue. A simple test: swap the suspected board into a known-good miner. If the error follows the board, it is a board issue.

What causes voltage errors after a board has been working fine for months?

Thermal cycling (repeated heating and cooling) fatigues solder joints and degrades components over time. This is normal wear. Voltage regulators and their surrounding capacitors are the most common wear-out failure after 12–24 months of continuous operation.

Abnormal Chip Voltage — Troubleshooting Guide

On this page