Energy Quotient Platform Documentation

The Energy Quotient platform delivers continuous waveform intelligence for mission-critical power systems. This documentation covers the platform architecture, sensor hardware, data access, edge analytics, and deployment options.

Platform

EQ Wave — High-fidelity continuous waveform capture

• 7-channel Continuous Point-on-Wave (CPOW) measurement (3-phase voltage + 4 current) at 32 ksps, 24-bit resolution
• Dual data streams: raw continuous waveforms + aggregated power monitoring
• Rugged, fanless design with two-stage measurement isolation and built-in backup power (v1.2 and later)

EQ Coherence — Data substrate

• High-bandwidth, low-latency ingestion pipeline (~6 Mbps per sensor, gap-free)
• Local Parquet storage with automatic rotation and optional data lake sync
• REST API and WebSocket streaming for concurrent consumers

EQ Syntropy — Edge analytics and agentic AI framework

• Physics-informed analysis with waveform-level evidence
• Statistical detectors, anomaly detection, and pattern interpretation
• API for historical queries and live streaming

EQ Sight — Human–AI interface for exploration and action

• Real-time dashboards with role-specific context
• Embedded AI assistant powered by EQ Syntropy
• Event detection, alerting, and diagnostic workflows
• JupyterLab for advanced analysis

EQ Wave is the only required component; EQ Coherence and EQ Syntropy are optional.

Getting Started

1. Install EQ Wave sensors
2. Connect the gateway
3. Explore data in EQ Sight

Release Notes

Release notes for EQ Coherence, EQ Sight, and EQ Syntropy are published at changelog.eq.systems.

Service Status

Monitor EQ service availability at status.eq.systems. The status page covers core infrastructure, EQ Sight deployments, and documentation services, and auto-refreshes every 5 minutes.

EQ Wave

EQ Wave v1 installed with 3-phase voltage, current sensors, and POF fiber link

Purpose and Applications

The EQ Wave is a power system sensor that provides Continuous Point-on-Wave (CPOW) monitoring — uninterrupted, full-resolution waveform capture across all channels — alongside aggregated energy and power metrics (RMS, power, frequency, harmonics). Applications range from energy management and cost optimization through power quality investigation to controls and equipment protection.

• Two-stage measurement isolation with fiber optic networking and TCP data transport
• Fanless operation across -40°C to +70°C in a ruggedized enclosure
• Low-latency measurements: ~4 ms sensor (ADC-to-wire), ~8 ms to the gateway (~16 ms including gateway processing)
• Deterministic, gap-free streaming of waveform and metric data to EQ Coherence, EQ Syntropy, and EQ Sight

Primary Use Cases

Power Quality Analysis:

• CPOW capture for complete event investigation
• IEC 61000–4-30 Class S measurements for standards-based reporting
• High-resolution visibility into sags, swells, transients, and distortion
• Harmonic analysis and distortion characterization

Energy Management:

• Continuous energy monitoring and demand profiling
• Load analytics for optimization and forecasting
• Power factor monitoring and correction verification
• Cost allocation and billing verification

Equipment Protection & Diagnostics:

• Signal-level monitoring of voltage and current waveforms
• Threshold-based alerting for out-of-range conditions
• Historical trending for predictive maintenance
• Integration with facility monitoring systems
• Inputs suitable for closed-loop monitoring and future control applications

How It Works

The EQ Wave continuously samples all voltage and current channels at 32 ksps, providing continuous waveform monitoring without gaps or blind spots. CPOW data is streamed and stored as uninterrupted, lossless waveform samples without resampling, trigger dependence, or summarization. Data is streamed over TCP/IP in two formats:

• CPOW: Raw sample data streamed in 2 ms frames for detailed signal-level analysis
• Power monitoring (PMon): Aggregated metrics reported every 10 cycles (50 Hz grids) or 12 cycles (60 Hz grids), providing a consistent 5 Hz update rate for real-time dashboards and trending

The gateway runs EQ Coherence, which is storage-agnostic, supporting microSD via USB adapter, NVMe SSD, external SSD, or network-attached storage. With drives up to multiple terabytes, deployments can retain months or even years of continuous waveform history. EQ Syntropy adds physics-informed AI analytics and diagnostics. EQ Sight provides real-time visualization, event monitoring, and interactive investigation.

Application Areas

• Semiconductor Fabrication: Power quality monitoring for process control
• Medical Imaging: Equipment power validation and monitoring
• Data Centers: Power management and monitoring
• Industrial Processes: Real-time monitoring and control
• Grid Infrastructure: Distributed resource monitoring
• Energy Systems: Energy management and optimization

Key Features

Data Acquisition

• 32 ksps sampling with 24-bit ADC resolution (~16.5-bit effective number of bits (ENOB))
• Complete 3-phase monitoring: 3 voltage + 4 current channels
• Sub-cycle response time
• Global Navigation Satellite System (GNSS)/GPS time synchronization via optional onboard receiver (footprint present; not populated on current units)

Design Specifications

• Two-stage measurement isolation through fiber optics and capacitive-coupled digital communication
• Industrial temperature range: -40°C to +70°C, fanless operation
• Enclosure: UL94-V0 polycarbonate
• Universal power input: 85–528V AC or 4.5–36V DC (v1.2+); USB cable included for benchtop testing via DC input
• Internal power backup (v1.2 and later)
• Compact industrial form: 140 × 89 × 41 mm

Real-time Processing

• Dual network services:
  • CPOW reported in 2 ms frames (TCP port 1534)
  • PMon reported every 10/12 cycle (TCP port 1535)
• Gap-less signal-level streaming with onboard data storage
• Low-latency data delivery: ~4 ms sensor (ADC-to-wire), ~8 ms to the gateway (~16 ms including gateway processing)

Platform Integration

EQ Wave sensors stream data to an EQ Gateway running:

• EQ Coherence: Data collection, storage, REST API, WebSocket streaming, and facility integration (Modbus TCP, MQTT, DNP3 available upon request)
• EQ Sight: Real-time visualization, event monitoring, alerting, and reporting
• EQ Syntropy (optional): Physics-informed AI analytics and diagnostics

EQ Wave v2, the certified production version, is in development. See eq.systems/platform/eq-wave for the product roadmap.

EQ Wave Quick Start

What You Need

• EQ Wave unit
• Power source (4.5–36V DC for v1.2+ or 85–528V AC)
• Network connection
• Pre-configured EQ Gateway or other system running EQ Coherence software

Steps

1. Physical Installation

Mount the Device:

• Install on DIN rail or mount to panel

Connect Power:

• For DC power: Connect to VDC +/- terminals (4.5–36V DC for v1.2+)
• For AC power: Connect to L/L1 and L2/N terminals (85–528V AC line-to-neutral or line-to-line)

Connect Network:

• Connect POF (plastic optical fiber) cable between sensor and media converter. Each end has one transmit (red light visible) and one receive port; connect transmit to receive on each side.
• Connect media converter to gateway via RJ45 Ethernet
• Power the media converter via USB from the gateway or a separate USB power adapter

2. Network Configuration

The gateway software will communicate with EQ Wave using the following network settings by default:

• IP address: 192.168.10.10 (fixed)
• Subnet: 255.255.255.0

Future firmware updates will support automatic network configuration.

3. Verify Operation

Check LEDs on EQ Wave:

• On startup, both LEDs flash red briefly
• LINK turns solid green when fiber optic link is established with endpoint
• ACT turns solid green when actively transmitting data

Confirm in EQ Sight:

Open a web browser and navigate to the gateway’s LAN address or your pre-assigned subdomain (e.g., [site].pq.app). The sensor should appear as a connected device.

Next Steps

See Installation Guide for voltage and current sensor wiring, safety precautions, mounting specifications, and cable routing.

See Configuration Guide for network settings, power system parameters, and advanced features.

See Data Access Guide for protocol specifications, sample code, and API documentation.

Troubleshooting

• Both LEDs off — Verify power at the input terminals. Check polarity for DC inputs.
• LINK LED off — Verify the POF cable is fully seated at both the sensor and media converter. Check that the media converter has power.
• Sensor not in EQ Sight — Confirm the gateway’s sensor-facing interface is on the same subnet (192.168.10.x).

See the Troubleshooting Guide for further diagnostics.

EQ Wave Installation

Safety Requirements

Critical Safety Information

Qualified Personnel:

• Electrical contractor license (where required by local jurisdiction)
• Experience with 3-phase power systems up to 600V
• Knowledge of power measurement equipment
• Understanding of NFPA 70E arc flash and electrical safety standards

De-Energization Requirements:

• De-energize all circuits using proper lockout/tagout (LOTO) procedures before making electrical connections
• Follow NFPA 70E or equivalent electrical safety standards
• Use appropriate personal protective equipment (PPE):
  • Arc-rated clothing (minimum 8 cal/cm² where required)
  • Arc-rated face shield (if working on energized circuits)
  • Insulated gloves rated for working voltage
  • Safety glasses and hard hat (where required)

Voltage and Power Limits:

• Voltage measurement inputs (V0-V3): 600V RMS maximum (designed to IEC 61010-1 CAT III)
• AC power supply terminals (L/L1, N/L2): 528V AC maximum
• For systems above 600V line-to-ground, use appropriate potential transformers (PTs)
• For AC power above 528V, use external DC power supply

Installation Environment:

• Operating temperature: -40°C to +70°C
• Install in electrical enclosure or panel providing appropriate environmental protection
• Maintain clearances per local electrical code requirements

Package Contents

EQ Wave Device Unit:

• Device with serial number EQW-_________ (record for support reference)
• Mounting hardware (one option, as specified when ordering):
  • 3M Dual Lock adhesive strips (standard)
  • DIN rail clips (35mm rail, IEC 60715)
  • Magnets for ferrous surfaces
• USB power cable (for benchtop testing and configuration)

Communications:

• 2.2mm duplex POF (plastic optical fiber) cable (default 2m length; custom lengths available)
• Media converter (Firecomms FY-ENT-KSU)
• Media converter USB power cable
• Media converter USB power supply (5V, 0.5A)
• POF cutting tool
• Cat5e Ethernet patch cable (RJ45, for media converter to gateway connection)

Installation Materials:

• Self-adhesive cable identification labels
• Alcohol prep pad (surface preparation for adhesive mounting)

Accessories: Additional POF cable and media converter kits (FF-FYENT-KSU, includes media converter, USB power cable, USB power supply, POF cutting tool, and Cat5e patch cable) are available from FiberFin in North America.

Note: The EQ Wave requires a host computer for data collection and configuration. This can be the EQ Gateway (sold as a separate option) or a customer-supplied computer. See Deployment Options for details.

Connection Guide

Step 1: Mount the Device

Mounting Method (as supplied with your kit):

3M Dual Lock Adhesive (standard):

1. Clean mounting surface with alcohol prep pad; allow 15 seconds to dry
2. Peel protective backing from adhesive strips (pre-applied to sensor enclosure)
3. Position sensor and press firmly for 30 seconds

DIN Rail Clips:

1. Hook bottom edge of clip onto DIN rail (35mm per IEC 60715)
2. Lift the device upwards to compress the springs while bringing the top edge of the clip over the top of the rail
3. Release the device, then check that it is secure by pulling downwards and tilting side to side slightly

Magnetic Mounting:

1. Ensure ferrous mounting surface is clean and flat
2. Position sensor; magnets hold securely on contact
3. Verify mounting by attempting to slide sensor (should resist movement)

Clearance Requirements:

• Minimum 25mm clearance on all sides for convective cooling
• Do not obstruct optical fiber connector or terminal access
• Mounting surface temperature must remain within -40°C to +70°C range

Step 2: Connect Voltage Measurement Inputs

Voltage Input Terminals (Weidmüller Omnimate 4.0 push-in plug, 7.5 mm pitch, 12–20 AWG; header 8000078318, plug 8000078357):

Voltage Measurement Specifications:

• Maximum input voltage: 600V RMS line-to-ground (designed to IEC 61010-1 CAT III)
• Input impedance: 4 MΩ per channel (V1/V2/V3 referenced to V0)
• For systems above 600V: Use potential transformers (PTs) and connect sensor to PT secondaries
• Reference input V0: Connect to system neutral or ground reference for proper common-mode rejection

Connection Procedure:

1. Strip wire ends 7–8 mm
2. Push stripped wire into the connector until it clicks (no tools required, even for stranded wire without ferrules)
3. Confirm the green visual indicator has popped out on each terminal
4. To release a wire, lift the lever to open the contact
5. Apply terminal identification labels to facilitate future maintenance

Wire Routing:

• Route voltage wiring separately from current sensor wiring where practical

Voltage Connections by System Type

The datasheet shows typical connection diagrams. The following notes cover topology-specific wiring details.

4-Wire Wye Systems (3-Phase with Neutral)

• Voltage Connections: Connect neutral to V0, lines to V1, V2, V3
• Power Supply: Connect L/L1 to L1 terminal, N to N/L2 terminal
• Wiring Recommendation: Use separate wires for AC power supply and voltage measurement to avoid spurious readings from the ~3W power draw

3-Wire Delta Systems

• Voltage Connections: Connect L1 to V1, L2 to both V0 and V2, L3 to V3
• Firmware Configuration: Contact EQ to update firmware for L-L measurement reporting

Split-Phase Systems (2-Phase with Neutral)

• Voltage Connections: Connect neutral to V0, lines to V1 and V2 (leave V3 unused)
• Configuration: Similar to 4-wire Wye but with only two active phases
• Typical Applications: North American residential 240V/120V systems

Single-Phase Systems

• With Neutral Available: Connect neutral to V0, line to V1 (leave V2, V3 unused)
• Without Neutral (2-Wire): Connect L1 to V1, L2 to V0 (like Delta configuration without third wire)

Step 3: Connect Current Measurement Inputs

Current Input Terminals (Weidmüller Omnimate 4.0 push-in plug, 5.0 mm pitch, 14–20 AWG; header 8000072456, plug 2741750000):

Current Sensor Installation:

1. Install current sensors on conductors with arrow or directional marking pointing toward the source (or as indicated)
2. Verify current sensor ratio matches the expected load current (e.g., 100A sensor for 80A maximum load)
3. Secure current sensor to conductor per manufacturer specifications
4. Route current sensor cables separately from high-voltage wiring
5. Push sensor wires into connectors until they click (observing polarity)
6. Confirm the green visual indicator has popped out on each terminal

Polarity:

• Correct polarity is critical for accurate power measurement; reversed polarity will show negative power for consuming loads
• If polarity is reversed, physically swap the connections at the terminal

Wire Colors:

• Shown are typical for Socomec Accu-CT and similar transducers
• Always verify wiring per your specific current sensor manufacturer’s documentation

Step 4: Apply Power

Power Input Terminals (same connector as voltage inputs, 7.5 mm pitch push-in, 12–20 AWG):

AC Power Option (85–528V AC, 50/60 Hz):

• Reference-agnostic design supports line-to-neutral or line-to-line configurations
• Common voltages: 100V (Japan), 120V (North America), 230V (Europe), 277V (commercial L-N), 480V (industrial L-L)
• Maximum input: 528V AC
• For systems above 528V AC: Use external DC power supply
• Power consumption: 3.3W typical

DC Output (when AC powered, same connector as current inputs, 5.0 mm pitch push-in):

• Terminals VDC+/VDC- provide regulated DC output for auxiliary devices
• Default output: 12V DC at up to 1.6W (custom voltages available when specified at ordering)
• Typical use: Powering Rogowski coil signal conditioners, transmitters, or indicator lights

DC Power Option (4.5–36V DC, v1.2 only, same connector as current inputs, 5.0 mm pitch push-in):

• Wide input range (4.5–36V) accommodates automotive (12V) and industrial (24V) DC sources
• Reverse polarity protected
• Power consumption: 2.4W typical
• Recommended: Connect VDC- to earth ground for optimal signal quality and noise immunity

USB power cable connected to DC input for benchtop testing

The included USB power cable provides a convenient 5V DC source for benchtop testing and configuration without requiring AC mains or an external DC supply. The USB cable connects to the VDC+/VDC- terminals.

Grounding and Bonding

Chassis Ground:

• Connect sensor enclosure to facility earth ground per local electrical code
• Recommended wire: 12 AWG minimum copper conductor
• Verify continuity to main ground bus with multimeter
• Proper grounding ensures optimal signal quality and noise immunity

Signal Grounds:

• All measurement circuits are isolated from chassis
• VDC- terminal should be connected to earth ground when using DC power
• Avoid ground loops through current sensor shields
• Unlike other power quality meters, the EQ Wave has no internal ground leakage current
• Ground is limited to the isolated section after voltage measurement only

LED Status Indicators:

After applying power, observe the two LED indicators on the device enclosure:

• Startup: Both LINK and ACT LEDs flash red briefly
• LINK LED: Turns solid green when fiber optic link is established with endpoint
• ACT LED: Turns solid green when actively transmitting data
• Fault condition: LEDs remain red — see Troubleshooting Guide

Step 5: Connect Plastic Optical Fiber (POF)

POF Cable Preparation (if not supplied with pre-installed connectors):

1. Cut cable end: Use the supplied POF cutting tool to make a clean, perpendicular cut
2. Separate duplex fibers: Split the webbing 10–15mm from the cut end to separate the two fiber strands
3. Verify clean ends: Inspect fiber ends; cuts must be smooth and perpendicular, free from debris

Connect to EQ Wave Device:

1. Unlock OptoLock® connector (Firecomms): Pull retainer clip outward to open position
2. Insert fiber strands: Push both prepared fiber ends fully into the OptoLock receptacle until seated
3. Lock connector: Press retainer clip back to closed position to secure fibers
4. Verify insertion: Fibers should be firmly held and not pull out with gentle tugging

Route POF Cable:

• Route cable from sensor location (inside electrical panel) to media converter location (typically outside panel), up to 100 m
• Minimum bend radius: 25mm (do not create sharp bends or kinks)
• Avoid routing near sharp edges that could damage fiber
• Secure cable with tie-wraps or cable clamps every 300–500mm along routing path

Connect to Media Converter:

1. Repeat OptoLock connection procedure at media converter end
2. Match transmit/receive fibers (if devices are powered, illuminated fiber connects to dark port, dark fiber to illuminated port)

Power Media Converter:

• Recommended (simpler): Connect media converter USB power cable to EQ Gateway USB port
• Alternative: Connect to included USB power supply and plug into mains power

Verify Optical Link:

After connecting fiber and powering media converter:

• EQ Wave sensor: LINK LED turns solid green, ACT LED turns green when data transfer is active
• Media converter: Both amber and green LEDs illuminated
• If link does not establish, see Troubleshooting section

Close Electrical Panel:

Once LED status confirms successful link establishment, reinstall panel cover and secure per local electrical code requirements.

Data Access and Configuration

With EQ Coherence (Recommended)

EQ Coherence records all data locally and provides REST API and WebSocket access. EQ Sight provides real-time visualization of waveforms, power metrics, and harmonics. Data can be exported using the equser Python package. Additional protocol support for SCADA/BMS integration is available upon request.

Setup: Refer to the Gateway Overview for setup procedures.

Direct Network Access (Advanced Integration)

For custom applications requiring direct sensor communication:

Network Configuration:

• IP address: 192.168.10.10 (fixed)
• Subnet: 255.255.255.0
• Connect to sensor network via media converter and Ethernet

Data Protocols:

• CPOW: TCP port 1534 — real-time high-resolution waveform streaming
• PMon: TCP port 1535 — 10 or 12 cycle aggregated measurements

Documentation:

• API specifications and programming examples available from [email protected]
• See Data Access Guide for protocol details and sample code

Verification Checklist

After completing installation, verify system operation through EQ Sight or direct network access.

Physical Installation Verification

• Device securely mounted with adequate clearance (25mm minimum all sides)
• All voltage and current measurement wires fully seated in push-in connectors
• Power connection secure (AC or DC, not both)
• POF cable connected at both ends (sensor and media converter)
• POF cable routing has no sharp bends (25mm minimum radius)
• Terminal identification labels applied for future reference
• Electrical panel cover reinstalled and secured

LED Status Verification

• LINK LED solid green (optical link established)
• ACT LED solid green (sensor active and transmitting data)
• Media converter LEDs illuminated (both amber and green)

Voltage Measurement Verification

• All phase voltages displayed and within ±5% of nominal system voltage
• Phase-to-phase voltages balanced (within 2% of each other for symmetrical systems)
• Phase sequence correct (A→B→C rotation, 120° phase separation for 3-phase systems)
• Voltage reference (V0) reading appropriate for system configuration

Current Measurement Verification

• All phase currents displaying reasonable magnitudes for connected load
• Current magnitudes balanced (within expected range for load configuration)
• Active power (kW) positive for all phases (consuming load) or as expected for generation
• Power factor values reasonable for load type
• Current sensor polarity correct on all phases (reverse polarity if power shows negative for consuming loads)

Network Communication Verification

• Gateway can communicate with sensor (web interface shows live data)
• Data updating in real-time (waveforms refreshing, values changing with load)
• No communication errors or timeout messages
• Network link stable over 5–10 minute observation period

Troubleshooting

For comprehensive troubleshooting beyond installation issues, see the Troubleshooting Guide.

Common Installation Issues

Both LEDs Off - No Power

Symptoms: LEDs completely dark, no startup flash.

Diagnostic Steps:

1. Verify power source is energized (measure voltage at source)
2. Check terminal connections for proper seating (pull gently to confirm)
3. Verify correct power type (AC or DC, not both)
4. For AC power: Verify voltage within 85–528V range
5. For DC power: Verify voltage within 4.5–36V range and correct polarity

Solutions:

• Verify power supply voltage and re-check terminal connections
• Ensure power source circuit breaker is closed
• For DC power: Verify polarity (VDC+ positive, VDC- negative)
• If power verified at terminals but sensor not operating, contact [email protected]

LEDs Remain Red - Fault Condition

Symptoms: LEDs flash or remain solid red after startup, do not turn green.

Diagnostic Steps:

1. Check all measurement input connections for proper seating
2. Verify no short circuits or over-voltage conditions on measurement inputs
3. Cycle power (disconnect, wait 10 seconds, reconnect)
4. Check for visible damage to sensor enclosure or terminals

Solutions:

• Verify all terminal connections are fully seated
• Disconnect measurement inputs temporarily to isolate fault
• If fault persists with no connections, contact [email protected] for RMA

LEDs Not Green - No Optical Link

Symptoms: ACT LED solid green, but LINK LED not solid green.

Diagnostic Steps:

1. Verify media converter has power (LEDs illuminated)
2. Check POF cable connections at both ends (sensor and media converter)
3. Inspect fiber ends for contamination or damage
4. Verify cable routing has no sharp bends or pinch points

Solutions:

• Remove and reconnect POF cable at both ends, ensure full insertion
• Clean fiber ends with isopropyl alcohol if contaminated
• Re-cut fiber ends if damaged or poorly prepared
• Try swapping transmit/receive fibers at media converter
• If link cannot be established, refer to EQ Gateway troubleshooting documentation or contact [email protected]

Negative Power Readings - Current Sensor Polarity Reversed

Symptoms: Active power (kW) shows negative for consuming loads on one or more phases.

Solutions:

• Recommended: Correct polarity via EQ Gateway software configuration (no rewiring required)
• Alternative: Physically reverse current sensor connections for affected phase (swap + and - terminals)

Incorrect Phase Sequence

Symptoms: Phase rotation displays as A→C→B instead of A→B→C, or phases are misidentified.

Solutions:

• Recommended: Correct phase mapping via EQ Gateway software configuration
• Alternative: Verify and correct voltage input wiring to match actual system phases

Technical Specifications

For complete technical specifications, see the EQ Wave v1 Datasheet (PDF).

EQ Wave Configuration

This guide covers the configuration parameters for EQ Wave v1 sensors. Refer to the Installation Guide for hardware setup.

Network Settings

EQ Wave v1 ships with fixed network settings:

• IP Address: 192.168.10.10
• Subnet Mask: 255.255.255.0
• TCP Services: Port 1534 (CPOW waveform data), Port 1535 (PMon power metrics)

These settings are factory-configured and cannot be changed in the field with current firmware. The gateway’s sensor-facing interface must be configured on the same subnet (default: 192.168.10.2).

Future firmware updates will support automatic network configuration.

Power System Parameters

The following parameters are set at the factory based on your order specifications. Contact [email protected] to request changes.

System Type

• Number of phases: 1, 2, or 3
• Nominal frequency: 50 Hz or 60 Hz
• Nominal voltage: Typically 120V, 208V, 240V, 277V, or 480V

Current Sensor Setup

• Default configuration: Voltage-output current transducers (333 mV at full scale, e.g., Socomec Accu-CT)
• Alternative sensors (must be specified when ordering): Split-core CTs (5A secondary typical), Rogowski coils (100 mV/A output typical)
• CT ratio (e.g., 150:5 for 150A primary, 5A secondary)
• CT direction (ensure power flow polarity is correct)

Polarity Convention:

• CT Arrow: Points toward load
• Primary Current: Flows from source to load
• Secondary Connection: I+ connects to CT terminal marked with arrow
• Power Factor: Positive for inductive loads when properly connected

Voltage and Current Scaling

Voltage scaling, gain settings, and current sensor calibration are factory-configured based on your order specifications and installation environment. These parameters are optimized for your specific system type (Wye, Delta, split-phase, single-phase) and voltage level.

If readings do not match expected values or your installation requirements change, contact [email protected]. Future firmware releases will support user-adjustable scaling and gain settings.

For measurement accuracy specifications, see the EQ Wave v1 Datasheet (PDF).

Sampling Parameters

• PMon update period: 12 cycles at 60 Hz / 10 cycles at 50 Hz (200 ms)
• CPOW frame rate: 500 packets/second (64 samples per packet at 32 ksps)

Verifying Configuration

After installation, verify the factory settings are correct for your site:

1. Voltage readings: Compare displayed RMS voltage against a known reference meter. If readings are off by a fixed ratio, the voltage scaling may need adjustment.
2. Current readings: Apply a known load and verify current magnitude and sign. Negative power indicates reversed CT polarity.
3. Phase sequence: Confirm phase angles are approximately 120 degrees apart for 3-phase systems.
4. Frequency: Should match your local grid (50 Hz or 60 Hz).

If any readings are incorrect, contact [email protected] with your measurements and the expected values.

Next Steps

• See Data Access Guide for connecting to data streams
• See Operations & Maintenance for ongoing system monitoring

EQ Wave Data Access

Data Streams Overview

The EQ Wave provides two primary data streams via TCP sockets for monitoring:

1. PMon — Port 1535

Purpose: Processed power quality metrics

• Parameters: RMS voltage/current, power, frequency, total harmonic distortion (THD)
• Data Format: IEEE 754 float32 values
• Update Rate: Every 12 cycles at 60 Hz or 10 cycles at 50 Hz (200 ms = 5 Hz update rate)
• Bandwidth: ~1 kbps

Data includes:

• RMS voltage and current per phase (total and fundamental)
• Active power per phase (total and fundamental)
• Fundamental reactive power per phase
• Frequency
• Derived metrics (THD, apparent power, power factor) computed by the gateway

Integration Options:

• Python: Full support for power monitoring data
• Rust: Enhanced performance libraries available under NDA
• Other Languages: TCP socket compatibility

2. Waveform Data (CPOW) - Port 1534

Purpose: High-resolution continuous waveform data

• Sample Rate: 32 ksps per channel
• Channels: 7 channels per sample row
• Channel Order: IA, VA, IB, VB, IC, VC, IN (current-voltage interleaved)
• Data Format: 24-bit signed integers, little-endian
• Update Rate: 2ms packets (500 packets/second)
• Bandwidth: ~6 Mbps continuous (payload plus TCP/IP overhead)

Connection Protocol:

1. Connect to IP:1534
2. Send 0x01 to start streaming
3. Receive continuous binary data
4. Send 0x02 to stop streaming

Performance Requirements:

• Compiled Language Required: Due to timing and buffer resource constraints, waveform data access requires compiled languages
• Recommended: Rust (core libraries available under NDA)
• Alternative: C/C++ (possible but less feature-rich)
• Not Suitable: Python or other interpreted languages for real-time waveform processing

Data Format Specifications

Waveform Data Structure

Frame Size: 1344 bytes

• 7 channels × 64 samples × 3 bytes per sample
• 24-bit signed integers, little-endian
• Channel order per sample row: IA, VA, IB, VB, IC, VC, IN
• Sample rate: 32 ksps per channel

Power Monitoring Data Structure

Frame Size: 96 bytes

• 1 × uint32 configuration word + 23 × float32 values
• IEEE 754 float32 format, little-endian

Field order:

• Configuration (uint32): system config (number of phases)
• Frequency (1 float)
• Voltage RMS (VA, VB, VC) — 3 floats
• Current RMS (IA, IB, IC) — 3 floats
• Neutral Current RMS (IN) — 1 float
• Active Power (PA, PB, PC) — 3 floats
• Fundamental Voltage RMS (VA1, VB1, VC1) — 3 floats
• Fundamental Current RMS (IA1, IB1, IC1) — 3 floats
• Fundamental Active Power (PA1, PB1, PC1) — 3 floats
• Fundamental Reactive Power (QA1, QB1, QC1) — 3 floats

Integration

Power Monitoring (PMon) — Python

The recommended way to access power monitoring data is through the equser Python package (GitHub · PyPI), which is pre-installed on every gateway.

# Acquire PMon data from the sensor and write Parquet files (uses a config file
# for the sensor address, data directory, and options)
equser pmon acquire -c config.yaml

# Convert previously captured Avro files to Parquet
equser pmon convert data/*.avro

The equser package also provides a Python API for programmatic access. Run equser --help on the gateway or see the GitHub README for details.

Waveform Data (CPOW) — Compiled Binaries

Other Languages and Environments

Any environment with TCP socket support can connect to the PMon stream on port 1535 (e.g., MATLAB, LabVIEW, Node.js). See the data format specifications above for the 96-byte frame structure. The open-source equser package serves as a reference implementation for parsing and framing.

Platform Integration

For enterprise deployments, EQ Coherence provides:

• Long-term waveform and metrics storage
• REST API and WebSocket access for custom applications

EQ Sight provides real-time visualization, event monitoring, and interactive investigation.

See the EQ Coherence API Reference for REST and WebSocket endpoint details.

Data Quality Considerations

Network Requirements

• Minimum Bandwidth: ~6 Mbps for full waveform streaming
• Latency: <10ms recommended for real-time applications
• Packet Loss: <0.1% for optimal data quality
• Buffer Management: Implement application-level buffering for network variations

Troubleshooting Data Integration

Connection Issues

• Verify network connectivity to device IP
• Ensure device is powered and LINK and ACT LEDs are solid green
• Try connecting with simple TCP client (telnet, netcat)

Data Quality Issues

• Monitor for missing data packets
• Check network bandwidth and congestion
• Verify data parsing matches protocol specification
• Check for endianness issues in multi-byte values

Performance Issues

• Increase TCP buffer sizes if possible
• Implement application-level buffering
• Use multiple threads for concurrent data streams
• Consider data reduction techniques if bandwidth is limited

EQ Wave Operations and Maintenance

This guide covers routine operation verification and maintenance procedures for the EQ Wave sensor.

Normal Operation

Expected LED and data status:

• Both LINK and ACT LEDs solid green
• EQ Sight shows live voltage and current updating
• Voltage readings within ±5% of nominal system voltage
• Current readings reasonable for monitored load
• Power readings (kW, kVAR, kVA) consistent with load

Startup behavior:

1. Both LEDs flash red briefly after power is applied
2. LINK LED turns solid green when fiber optic link is established with endpoint (typically media converter)
3. ACT LED turns solid green when actively transmitting data
4. Data appears in EQ Sight shortly after ACT LED is green

If LEDs are not solid green or data is not updating in EQ Sight, see the Troubleshooting Guide for LED status reference and diagnostic procedures.

Measurement Verification

Voltage Readings:

• Should match known system voltages within ±0.5%
• 3-phase systems: Phases should be balanced within ±2% in symmetrical systems
• Phase angles approximately 120° apart for 3-phase

Current Readings:

• Should correlate with monitored load
• Verify CT polarity: Active power positive for consuming loads, negative for generation
• Phase angles should align with voltage (near 0° for resistive loads, leading/lagging for reactive loads)

Power Readings:

• Real power (kW) should match load nameplates and expected consumption
• Reactive power (kVAR) should correlate with load type (motors, transformers produce reactive power)
• Power factor should be reasonable for load type (0.9–1.0 for resistive, 0.7–0.9 for inductive)

If Readings Incorrect:

1. Verify CT ratios configured correctly in gateway
2. Check CT polarity (reverse if power negative for consuming load)
3. Verify voltage scaling matches system voltage
4. Compare with portable reference meter to confirm sensor accuracy
5. See Configuration Guide for CT ratio and polarity configuration

Routine Operation Checks

Check during facility inspections:

• LEDs: Both solid green
• Data: Updating normally in EQ Sight
• Readings: Consistent with expected load

Frequency: When accessing electrical panel for other maintenance (no dedicated inspection required).

No active monitoring required: The sensor operates continuously without user intervention. EQ Sight can be configured to send alerts if communication is lost or measurements are outside the expected range.

Maintenance Procedures

The EQ Wave sensor is a solid-state device with no moving parts, batteries, or user-serviceable components. Maintenance requirements are minimal.

Fiber Optic Connection Care

Normal Operation:

• When disconnecting POF cable, install red dust cover on OptoLock® connector to prevent contamination
• No routine fiber cleaning is necessary

Troubleshooting Only: If optical link fails after all other troubleshooting (LEDs not solid green):

1. De-energize sensor per lockout/tagout procedures
2. Disconnect POF cable from OptoLock connector
3. Inspect fiber ends for contamination or wear
4. If contamination suspected: Blow out OptoLock connector with low-pressure dry compressed air
5. If fiber ends worn or damaged: Re-cut fiber ends using POF cutting tool (clean perpendicular cut required)
6. Reconnect cable ensuring full insertion into OptoLock receptacle
7. Install red dust cover when connector not in use
8. Re-energize and verify LINK LED turns solid green once fiber is connected and ACT LED turns solid green when data transfer resumes

Note: Fiber optic connection issues are rare. If problems persist, see Troubleshooting Guide.

Terminal Connection Inspection

When required: During annual facility electrical inspection.

Procedure:

1. De-energize circuits per lockout/tagout procedures
2. Visually inspect all push-in connectors for:
   • Corrosion or oxidation
   • Signs of overheating (discoloration)
   • Wire insulation degradation
3. Verify all green visual indicators are extended (wire fully seated)
4. Replace any degraded wiring
5. Re-energize and verify normal operation (LEDs solid green, data updating in EQ Sight)

Frequency: Annually or per facility maintenance schedule.

Mounting Hardware Verification

3M Dual Lock Adhesive:

• Inspect adhesive bond for peeling or degradation
• Verify sensor remains securely attached
• If bond fails: Clean surface with isopropyl alcohol and apply new adhesive strips

DIN Rail Clips:

• Verify clips fully engaged on rail
• Ensure sensor does not move when pulled downward
• If loose: Remove and re-install clips

Magnetic Mount:

• Verify magnets hold sensor firmly to mounting surface
• Check for corrosion between magnet and surface
• Ensure sensor does not slide when subjected to normal vibration

Frequency: Annually or if sensor subjected to unusual vibration or mechanical shock.

Firmware Updates

v1 Devices: Require return to factory for firmware updates. Contact [email protected] to initiate RMA process. Typical turnaround: 7–10 business days.

Update Notifications: Firmware updates distributed via email to registered users. Updates are optional unless addressing security vulnerabilities or critical issues.

Device Service Life and Replacement

Expected Service Life: 10 years of continuous operation under normal conditions.

Design and Components:

• Solid-state design with no moving parts or wear mechanisms
• Electrolytic capacitors in AC/DC power supply (if AC powered) and supercapacitors have typical 10-year service life
• All other components are solid-state electronics with indefinite service life under normal conditions

Replace sensor if:

• Physical damage to enclosure or terminals
• Lightning strike or electrical surge damage
• LEDs remain red or off after power cycling
• Measurements drift beyond acceptable limits (verify with known reference first)
• Approximately 10 years of continuous operation

Replacement procedure: See Installation Guide for new sensor installation. Existing wiring and fiber optic cable can typically be reused.

Calibration

EQ Wave v1 uses theoretical scaling ratios for voltage and current measurement. No per-unit factory calibration is performed. The ADC and analog front-end provide high intrinsic accuracy (see the EQ Wave v1 Datasheet (PDF)), so theoretical ratios are sufficient for most applications.

Verification: Compare sensor readings with a portable reference meter annually or per facility requirements.

If readings are outside expected accuracy:

1. Verify correct CT ratios and voltage scaling are configured
2. Verify voltage and current sensor connections
3. Check for loose terminal connections
4. Compare with multiple reference instruments to rule out reference meter error
5. Contact [email protected] if sensor fault is suspected

For environmental specifications (temperature, humidity, vibration ratings), see the EQ Wave v1 Datasheet (PDF). For environment-related troubleshooting, see the Troubleshooting Guide.

EQ Wave Troubleshooting

LED Status Indicators

The EQ Wave has two status LEDs that provide quick diagnostic information:

Quick Diagnostic Checklist

Before starting detailed troubleshooting:

• Check LED status using table above
• Verify media converter LEDs (both amber and green illuminated)
• Test network connectivity: ping sensor at 192.168.10.10 or gateway at 192.168.10.2
• Review data streams via EQ Sight or direct TCP connection

Power and System Issues

No Power - Both LEDs Off

Symptoms:

• LINK and ACT LEDs completely dark (no red startup flash)
• Device unresponsive
• No network connectivity

Troubleshooting Steps:

1. Check Power Source:

   • Verify power supply voltage (4.5–36V DC or 85–528V AC)
   • Measure voltage at power input terminals
   • Verify power supply can deliver at least 2.4W DC or 3.3W AC
   • Try alternative power input method

2. Check Connections:

   • Verify power terminal connections are tight
   • Check for corrosion or damage
   • Ensure proper wire sizing (12–20 AWG)
   • Verify polarity for DC inputs

3. Test with Alternative Power:

   • Try USB power for testing (limited functionality)
   • Use known good power supply
   • Check for blown fuses in external power supply

Expected Resolution: LINK and ACT LEDs should both illuminate solid green

Intermittent Power Issues

Symptoms:

• LEDs flicker or go off intermittently
• System restarts unexpectedly
• Data interruptions

Troubleshooting Steps:

1. Check Power Quality:

   • Monitor input voltage stability
   • Look for voltage sags or surges
   • Check for loose connections
   • Verify power supply capacity under load

2. Environmental Factors:

   • Check operating temperature (-40°C to +70°C)
   • Ensure adequate ventilation
   • Look for vibration or shock
   • Check for moisture ingress

3. Load Analysis:

   • Verify power consumption 3.3W AC / 2.4W DC typical
   • Check for excessive network load
   • Monitor internal temperature
   • Review system diagnostics

Network Connectivity Issues

No Network Connection

Symptoms:

• Cannot ping device IP address
• LINK LED off
• No data streaming

Troubleshooting Steps:

1. Physical Connection:

   • Verify fiber optic cable connections
   • Check media converter power and status LEDs
   • Test fiber cable with optical power meter
   • Try known good fiber cable

2. Fiber Optic Issues:

   • Clean fiber connectors with lint-free wipes
   • Inspect connector end faces for damage
   • Ensure proper connector seating
   • Check for cable bending radius violations

3. Network Configuration:

   • Verify IP address settings
   • Check subnet and gateway configuration
   • Default IP is 192.168.10.10 (fixed)

4. Media Converter Issues:

   • Verify media converter power
   • Check Ethernet connection to switch
   • Test with different media converter
   • Verify 100Base-FX compatibility

Expected Resolution: LINK LED solid green, ACT LED solid green, device responds to ping

Cannot Access Data Ports

Symptoms:

• Network connectivity works (ping successful)
• TCP connections to ports 1534/1535 fail or timeout

Troubleshooting Steps:

1. Network Verification:

   • Verify device responds to ping at 192.168.10.10
   • Test TCP connection with telnet or netcat (ports 1534, 1535)
   • Verify computer IP address is in same subnet (192.168.10.x)

2. Device Issues:

   • Power cycle the device
   • Verify both LINK and ACT LEDs are solid green
   • Test from a different computer

Data Connection Timeouts

Symptoms:

• Ping works but data streaming fails
• Intermittent data loss
• Application cannot connect to data ports

Troubleshooting Steps:

1. Port Accessibility:

   • Verify TCP ports 1534, 1535 are open
   • Test with telnet or netcat
   • Monitor network statistics

2. Bandwidth Issues:

   • Check network bandwidth availability
   • Monitor system diagnostics for buffer overflows
   • Reduce data streaming rate if necessary
   • Optimize network infrastructure

3. Application Issues:

   • Increase application TCP buffer sizes
   • Implement proper error handling
   • Check application processing speed
   • Monitor client-side resources

Measurement Issues

Inaccurate Voltage Readings

Symptoms:

• Voltage readings significantly different from reference
• Inconsistent measurements
• Wrong scaling

Troubleshooting Steps:

1. Connection Verification:

   • Verify voltage reference connections (V0 terminal)
   • Check voltage divider scaling configuration
   • Ensure proper grounding of measurement circuit
   • Verify input voltage is within specified range

2. Configuration Check:

   • Verify system voltage configuration
   • Check PGA gain settings
   • Confirm voltage divider ratios
   • Review calibration factors

3. Signal Quality:

   • Check for EMI/RFI interference
   • Verify cable shielding integrity
   • Look for ground loops
   • Monitor signal-to-noise ratio

Acceptance Criteria: ±0.2% accuracy for RMS voltage (sensor only; field measurements may vary with wiring and connections)

Inaccurate Current Readings

Symptoms:

• Current readings don’t match applied current
• Wrong power factor calculations
• Phase angle errors

Troubleshooting Steps:

1. CT/Rogowski Setup:

   • Check CT polarity and direction
   • Verify CT ratio configuration matches actual CT
   • Ensure CT secondary is properly terminated
   • Check for proper CT burden resistance

2. Wiring Verification:

   • Verify CT primary direction (arrow toward load)
   • Check secondary connections (I+ and I-)
   • Ensure CT secondary never left open
   • Verify consistent CT orientations

3. Configuration:

   • Check CT ratio settings in configuration
   • Verify Rogowski coil sensitivity (if applicable)
   • Review current sensor calibration
   • Check input range settings

Acceptance Criteria: ±0.5% magnitude, ±0.2° phase accuracy (includes typical external sensor contribution)

Phase Angle Errors

Symptoms:

• Incorrect power factor readings
• Wrong phase sequence
• Inconsistent phase relationships

Troubleshooting Steps:

1. Wiring Check:

   • Verify all voltage and current connections
   • Check for consistent CT orientations
   • Ensure common voltage reference (V0) connection
   • Verify timing synchronization between channels

2. System Configuration:

   • Check system type setting (Delta, Wye, etc.)
   • Verify phase sequence (A-B-C rotation)
   • Review nominal frequency setting
   • Check sampling synchronization

3. Signal Quality:

   • Monitor for noise and interference
   • Check cable lengths and routing
   • Verify grounding practices
   • Look for timing synchronization issues

Expected Results:

• Phase sequence: A→B→C positive rotation
• Phase angles: ∠VA = 0°, ∠VB = -120°, ∠VC = +120°

Data Quality Issues

Missing Data Packets

Symptoms:

• Intermittent data gaps
• Application reports missing data
• Inconsistent update rates

Troubleshooting Steps:

1. Network Analysis:

   • Check network bandwidth and congestion
   • Monitor packet loss statistics
   • Verify network switch performance
   • Test network infrastructure

2. System Health:

   • Check system status via EQ Sight
   • Monitor for network congestion
   • Review recent configuration changes
   • Contact support if persistent

3. Application Optimization:

   • Increase TCP buffer sizes if possible
   • Implement application-level buffering
   • Optimize data processing speed
   • Use multiple threads for data handling

Data Corruption

Symptoms:

• Invalid data values
• Checksum errors
• Parsing failures

Troubleshooting Steps:

1. Protocol Verification:

   • Verify data parsing matches protocol specification
   • Check for endianness issues in multi-byte values
   • Validate data structure alignment
   • Review protocol documentation

2. Network Issues:

   • Monitor for network errors
   • Check cable integrity
   • Look for EMI/RFI interference
   • Test with different network equipment

3. If Issues Persist:

   • Document specific error patterns
   • Note when corruption occurs (specific data types, times)
   • Contact support with error examples
   • Provide network configuration details

Environmental Issues

Temperature-Related Problems

Symptoms:

• Performance degradation in extreme temperatures
• Thermal shutdown events
• Measurement drift

Troubleshooting Steps:

1. Temperature Monitoring:

   • Check internal temperature readings
   • Verify operating range (-40°C to +70°C)
   • Monitor ambient temperature
   • Check for heat sources nearby

2. Ventilation:

   • Ensure adequate clearance around device
   • Check for blocked ventilation
   • Verify mounting orientation
   • Consider additional cooling if needed

3. Environmental Protection:

   • Check for direct sunlight exposure
   • Verify enclosure IP rating
   • Look for heat-generating equipment nearby
   • Consider thermal insulation if needed

Vibration and Shock Issues

Symptoms:

• Intermittent connection problems
• Mechanical damage
• Mounting hardware loosening

Troubleshooting Steps:

1. Mounting Inspection:

   • Check mounting hardware tightness
   • Verify DIN rail or panel mount security
   • Look for mechanical stress
   • Check for proper mounting orientation

2. Vibration Sources:

   • Identify vibration sources
   • Consider vibration isolation
   • Check equipment mounting
   • Monitor acceleration levels

3. Connection Security:

   • Verify all electrical connections
   • Check for wire fatigue
   • Ensure strain relief adequate
   • Consider flexible connections

System Recovery Procedures

Power Cycle

If the device behaves erratically or becomes unresponsive:

1. Disconnect power
2. Wait 10 seconds
3. Reconnect power
4. Verify both LEDs flash red briefly, then turn solid green

Firmware Issues

v1 devices require return to factory for firmware updates. Contact support to initiate the RMA process if:

• Device behaves unexpectedly after a firmware change
• System is unresponsive and power cycling does not resolve the issue

Getting Additional Help

Information to Collect

Before contacting support, gather:

• Device serial number and firmware version
• LED status patterns (refer to table above)
• Complete error description: when it started, how often it occurs, what changed
• Network configuration: IP address, gateway, subnet
• Installation details: power source, sensor types, environmental conditions
• Recent changes: configuration updates, firmware updates, wiring changes

EQ Wave v1

The EQ Wave v1 is a CPOW power quality sensor designed for permanent installation in electrical panels. It captures gapless, high-resolution voltage and current waveforms and streams them to an EQ Gateway for storage and analysis.

Key Characteristics

• Measurement: 3-phase voltage (up to 600V RMS) + 4 current channels via external transducers or Rogowski coils
• Resolution: 24-bit, 32 ksps, 7-channel simultaneous-sampling delta-sigma ADC
• Two-stage isolation: Fiber optic networking + internal isolation barrier (3 kV)
• Self-powered: AC input (85–528V) or DC input (4.5–36V, v1.2)
• Industrial: -40°C to +70°C, fanless, rugged polycarbonate enclosure
• Data streams: Continuous CPOW waveforms (port 1534) + power metrics (port 1535) over TCP
• Network: 100Base-FX over plastic optical fiber (POF) via Firecomms OptoLock® connector
• Processor: ARM Cortex-M4 with FPU, real-time operating system with TCP/IP networking

Platform Integration

The sensor connects to an EQ Gateway running EQ Coherence for gapless data recording, long-term storage, APIs, and real-time visualization. EQ Syntropy adds optional AI analytics.

Hardware Versions

• v1.0 (2023): Initial prototype with core measurement and streaming
• v1.1 (2024): Dual high-voltage resistor divider for improved safety; timing-optimized GNSS receiver footprint
• v1.2 (2025): Current pilot version. Expanded voltage range (600V), supercapacitor backup power (~8 seconds), improved analog conditioning, DC power input (4.5–36V)

All v1 hardware versions share the same firmware and network interface.

Optional Hardware Features

Factory-configurable options — order with the matching suffix; see the datasheet for capability detail and lead time:

• GNSS receiver (-G) — Precision time synchronization. Not populated on standard units.
• Onboard storage (-SD) — microSD-based local buffering of CPOW and power metrics through gateway or network interruptions, with backfill on reconnect.

Core hardware (ADC, MCU, 100Base-FX optical networking, isolation architecture, power supply options) is covered in the datasheet.

Current Limitations (v1 Firmware)

• Fixed IP: 192.168.10.10, no DHCP
• No remote firmware update: v1 requires physical access (factory RMA) for firmware changes

EQ Wave v2, the certified production version, is in development. See eq.systems/platform/eq-wave, or contact [email protected] to discuss upcoming capabilities.

Component Sourcing

• Firecomms OptoLock® series (POF connector)
• POF cable: Mitsubishi ESKA® 2.2 mm duplex (FF-GHCP-4002), available from FiberFin (custom lengths and kit-bundled) or Radwell (bulk reels) in North America
• Media converter kit (FF-FYENT-KSU): includes media converter, USB power cable, USB power supply, POF cutting tool, and Cat5e patch cable; also available from FiberFin

More Information

• Installation Guide — wiring, safety, and setup procedures
• Data Access Guide — protocol specifications and integration
• Datasheet (PDF) — full electrical, mechanical, and communication specifications
• Operations & Maintenance — ongoing system monitoring

EQ Coherence

EQ Coherence is the data foundation of the Energy Quotient platform. It runs high-bandwidth, low-latency pipelines to ingest Continuous Point-on-Wave (CPOW) and power monitoring (PMon) data from EQ Wave sensors, gap-free, with minimal memory and CPU requirements. All data is recorded locally and served to any number of concurrent consumers through flexible REST and WebSocket APIs with on-demand processing.

EQ Syntropy and EQ Sight both build on EQ Coherence. Every gateway deployment includes it.

Data Pipeline

EQ Coherence captures two concurrent data streams from each EQ Wave sensor:

• Continuous Point-on-Wave (CPOW): 32 ksps, 7-channel waveform data streamed at ~2 ms intervals with sub-cycle latency. No triggering, no gaps.
• Power Monitoring (PMon): Aggregated RMS voltage, current, power, power factor, frequency, harmonics, and unbalance updated every 10/12 cycles (~200 ms at 50/60 Hz).

Both streams are validated, timestamped, and written to storage in real time.

Storage

• Format: Apache Parquet columnar files, optimized for time-series analytics and direct access from Python, Rust, and other tools
• Capacity: Approximately 50 GB/day for a typical 3-phase deployment (losslessly compressed; varies with channel count)
• Flexible media: Removable USB storage (microSD via USB adapter, USB flash drives, external SSDs) for easy capacity expansion and field data transfer
• Automatic rotation: Oldest data is reclaimed when storage fills, ensuring uninterrupted recording
• Data lake sync: Optional synchronization to the EQ data lake for centralized access and long-term archival (subscription service)

See Storage Media for capacity planning, recommended hardware, and swap procedures.

APIs

• REST API: Query months or years of waveform history with sub-cycle precision. Supports time-range selection, downsampling, and bulk export.
• WebSocket streaming: Live waveform and spectral data pushed to connected clients as it arrives. Multiple consumers can subscribe concurrently without impacting recording performance.

See the API Reference for REST endpoints, WebSocket streaming, and integration examples.

Integration

• EQ Wave sensors: Primary data source via fiber optic network (100Base-FX POF)
• Facility systems: REST API for custom integrations; additional protocol support (Modbus TCP, MQTT, DNP3) available upon request
• Enterprise tools: Data export in Parquet format for external analysis platforms and data pipelines
• Remote support: VPN connectivity for authorized remote access when enabled

Supported Gateway Hardware

See Deployment Options for the full list of supported hardware, including Compulab industrial gateways, Raspberry Pi (lab/demo), and customer-provided Linux computers.

Logs

EQ Coherence writes a plain-text activity log alongside the CPOW data at cpow/cpow_daq.log. It records sensor connection events and service start/stop, useful for verifying that recording is active or diagnosing connectivity issues.

Getting Started

• Gateway Overview — hardware setup and connections
• Storage Media — capacity planning and media setup
• API Reference — REST endpoints, WebSocket streaming, and integration methods

EQ Coherence API Reference

EQ Coherence exposes REST and WebSocket APIs for programmatic access to power quality data. All API endpoints are served from the gateway.

Access Paths

There are two ways to reach a gateway’s API, and they use different hosts and ports. Choose based on where your client runs.

Notes:

• Port 8080 is the gateway’s direct backend port. It is reachable only over the LAN, on the gateway’s own IP. It is not exposed on the eqg-*.pq.app address. Do not append :8080 to a pq.app host.
• The WAN path routes through Cloudflare to the gateway and is protected by single sign-on (CF Access / Keycloak). Remote requests must carry a valid login session, so this path is easiest from a browser-authenticated context. For quick one-off scripting, prefer the LAN path.

The endpoint paths (/api/v1/..., /api/ws/...) are identical on both. Only the host and port differ.

Integration Methods

Additional protocol support (Modbus TCP, MQTT, DNP3) is available upon request for SCADA/BMS integration. Contact [email protected] for details.

REST API

Base URL (LAN): http://[gateway-ip]:8080/api/v1/ Base URL (WAN): https://eqg-{serial}.pq.app/api/v1/

Data query parameters: start_time (ISO 8601), end_time (ISO 8601), metrics (comma-separated), limit.

Data endpoints return Apache Arrow IPC binary format for efficient transfer. Use Arrow libraries in Python, JavaScript, Rust, or other languages to deserialize.

Device Endpoints

Data Endpoints

Event Endpoints

System Endpoints

Example: Python with Arrow

import pyarrow.ipc as ipc
import requests

# Fetch power monitoring data
url = "http://192.168.1.100:8080/api/v1/devices/wave-001/pmon/data"
resp = requests.get(url, params={"start_time": "2025-06-15T12:00:00Z"})
reader = ipc.open_stream(resp.content)
table = reader.read_all()
df = table.to_pandas()

WebSocket Endpoints

For live streaming data from EQ Coherence.

Prefix with the host from the access path you are using:

• LAN: ws://[gateway-ip]:8080/api/ws/cpow_stream
• WAN: wss://eqg-{serial}.pq.app/api/ws/cpow_stream

Spectral WebSocket parameters: channels (e.g. VA,IA), mode (cycle_aligned or fixed), cycles (default 12, cycle_aligned mode), fft_size (default 4096, fixed mode), freq_min (default 0), freq_max (default 3000), include_phase (default false). The stream is a broadcast consumer of the live CPOW feed (not per-device); each binary message is an Arrow IPC window with per-window metadata on the schema.

Custom Integration

For integration requirements beyond what is documented here, contact [email protected].

EQ Sight

EQ Sight is the web-based interface for viewing and exploring power system data collected by EQ Wave sensors.

Accessing EQ Sight

EQ Sight is available through:

• Remote access (recommended): Via your assigned subdomain at [site].pq.app. This connection is secured with industry-standard TLS encryption and requires authentication. Contact [email protected] to add or manage users.
• Local network: Directly from the gateway at http://[gateway-ip] when connected to the same facility network. This uses an unencrypted HTTP connection, which is standard for LAN-only devices on trusted networks. No login is required for local access.

Contact your system administrator for your specific access URL.

Navigation

The sidebar on the left provides access to the main sections:

Click any icon to switch sections. The active section is highlighted in the sidebar.

EQ Sight — Dashboard

The Dashboard is the landing page in EQ Sight, accessible via the EQ icon in the sidebar.

Summary Cards

The top of the dashboard shows key performance indicators:

• Connected Devices: Number of active EQ Wave sensors
• PQ Events Today: Count of power quality events detected
• Uptime / Downtime: System availability over the last 30 days

Overview Tab

The default view provides two modes, toggled at the top:

• Device View: A facility map showing sensor locations and status. Hover over a device to see its current metrics. Click a device for quick actions (live waveform, device detail, or event analysis).
• Power Network: A single-line schematic of the electrical distribution system showing equipment status, cable loading, and power flow. Hover over any node to see voltage, current, power factor, THD, and impedance.

Device and equipment status is color-coded: green (normal), yellow (warning), red (alarm), gray (offline).

Events Tab

Shows recent power quality events organized by assignment and priority:

• Assigned To Me: Events assigned for your review, with status tracking (Pending, In Review, Analyzed, Completed)
• Saved PQ Events: Bookmarked events for follow-up
• Generated Reports: Previously generated event reports with download links

Reports Tab

A table of all generated reports with date, time, event type, and download actions.

Live Instruments

EQ Sight includes two real-time instruments that operate directly in your browser, streaming continuous waveform data from EQ Wave sensors at 32 ksps.

Accessing Live Instruments

From the device page, click Live Instruments. The page opens with the Oscilloscope view by default.

The header bar provides:

Oscilloscope

The oscilloscope displays live voltage and current waveforms with zero-crossing triggered capture.

Triggered mode locks the display to rising zero crossings on the highest-priority voltage channel, producing a stable, oscilloscope-like view. The number of displayed cycles is adjustable via mouse wheel scroll on the chart.

Free-running mode shows a continuous rolling window of raw waveform data.

The frequency readout (e.g., “59.98 Hz (VA)”) is measured from interpolated zero crossings on the trigger channel, giving sub-Hz accuracy from the 32 ksps sample clock.

Channel selection

Click channel buttons to toggle individual phases on or off. Multiple channels can be active simultaneously. Voltage channels appear in the upper chart; current channels in the lower chart.

A high-pass filter (7 Hz corner frequency) automatically removes sensor DC offset from current channels.

Spectrum Analyzer

The spectrum analyzer provides a real-time FFT spectrogram with a live spectrum line chart.

The display has three sections:

• Color bar (left) — Magnitude scale in dB
• Waterfall spectrogram (center) — Rolling time-frequency heatmap. Time flows from left (25 seconds ago) to right (now). Color indicates magnitude.
• Spectrum line (right) — Current FFT snapshot at the leading edge of the waterfall, showing magnitude vs. frequency

Frequency range

Scroll the mouse wheel over the spectrogram to zoom the frequency axis:

• Scroll up — Zoom in (narrower range, more detail on low-frequency harmonics)
• Scroll down — Zoom out (up to 8 kHz, the anti-alias corner frequency)

The frequency range is anchored at 0 Hz (bottom).

Resolution

Click the resolution indicator in the header bar to cycle through available settings:

Channel selection

In spectrum analyzer mode, channel buttons switch to single-select. Only one channel is analyzed at a time. The WebSocket connection updates to stream only the selected channel.

Side-by-Side Operation

Both instruments can run simultaneously in separate browser windows or tabs, streaming from the same device.

URL State

All instrument settings are preserved in the URL for bookmarking and sharing:

• mode=triggered / mode=free-running / mode=spectrogram
• channels=VA,IA (multi-select) or channels=VA (single in spectrum mode)
• fftSize=4096 and freqMax=8000 (spectrum analyzer settings)
• cycles=5 (triggered mode cycle count)

EQ Sight — PQ Events

The PQ Events section, accessible via the triangle icon in the sidebar, provides three views for monitoring power quality across your facility.

Facility Map

The default view shows a map of your facility with sensors displayed at their installed locations. Each device is color-coded by status:

• Green: Healthy, no active events
• Yellow: Warning-level events detected
• Red: Critical events requiring attention

Hover over a device to see a summary of its current metrics. Click a device for quick actions: view live waveforms, open the device detail page, or jump to event analysis.

List View

Displays devices sorted by event priority, with the most critical issues at the top. Each device card shows:

• Device name and location
• Count of events by severity (Critical, High, Medium, Low)
• Quick action buttons for live waveform, device detail, and event analysis

Expand a device card to see individual events with type, priority, and timestamp.

Power Network

An interactive single-line schematic of the electrical distribution system. Shows:

• Equipment nodes (utility feeds, transformers, switchgear, MCCs, panels, loads)
• Cable connections with color-coded loading (green < 70%, yellow 70-90%, red > 90%)
• Animated power flow direction (toggleable)

Hover over any node for detailed metrics: voltage, current, power factor, THD, impedance, and status.

Event Reports

Click on an individual event to open the Event Report page. From there you can:

• Review event details (type, priority, device, timestamp)
• Add context about what was happening at the time
• Generate an analysis report with technical findings and recommendations

EQ Sight — Devices

The Devices section, accessible via the plug icon in the sidebar, shows connected EQ Wave sensors and provides access to historical data.

Device Grid

The main view displays a card for each connected sensor. Each card shows:

• Device name
• Power quality status indicator (green = compliant, yellow/red = events detected)
• A thumbnail chart of the selected metric over time

Use the metric selector at the top to choose which measurement is displayed on the cards (e.g., voltage RMS, current RMS, active power, frequency).

Click any device card to open the device detail page.

Device Detail

The detail page provides interactive plotting of historical data from an individual sensor.

Data Modes

• PMon: Power monitoring metrics recorded every 200 ms (voltage RMS, current RMS, active/reactive/apparent power, frequency, THD, power factor)
• CPOW: Raw continuous point-on-wave data at 32 ksps for waveform-level analysis

Chart Types

For PMon data:

• Line, Area, and Stepped charts for time-series trends
• Histogram for metric distribution analysis

For CPOW data:

• Waveform: Time-domain voltage and current waveforms
• FFT/Spectral: Frequency-domain analysis showing harmonics up to the 50th (3 kHz at 60 Hz)

Controls

• Metric selector (left panel): Choose which measurements to plot; multiple metrics can be displayed simultaneously
• Chart type selector: Switch between visualization modes
• Phase toggles (CPOW mode): Select which phases (A, B, C) to display for voltage and current

Live Waveform

From either the device grid or event views, you can open a live waveform page for any sensor. This streams CPOW data in real time over WebSocket and offers two display modes:

• Triggered: Captures and queues individual cycles (1-50 cycles per capture) synchronized to zero crossings. When no zero crossings are detected (DC signal or noise floor), the display falls back to time-based capture at the nominal cycle duration.
• Free-running: Continuous scrolling waveform display

A spectrogram view is also available, showing frequency content over time with selectable phase and FFT parameters.

Channel Selection

The toolbar shows toggle buttons for each channel: VA VB VC · IN · IA IB IC. Click a channel to enable or disable it. The default is VA + IA, which minimizes WebSocket bandwidth. Only enabled channels are streamed from the gateway; disabled channels are not transmitted.

Channel colors follow a consistent scheme:

• White: Phase A (VA, IA)
• Red: Phase B (VB, IB)
• Blue: Phase C (VC, IC)
• Purple: Neutral (IN)

The channel selection is saved in the URL (e.g., ?channels=VA,IA,VB) so you can share or bookmark a specific view. The waveform display automatically triggers on the first enabled channel that has zero crossings, using the priority order VA, VB, VC, IN, IA, IB, IC.

Single-Device Mode

When the gateway is connected to a single sensor, the Devices section shows that sensor directly with quick-action buttons for historical plotting, live waveform, and event list.

EQ Sight — JupyterLab

The JupyterLab view, accessible via the Erlenmeyer flask icon in the sidebar, provides an interactive Python environment for advanced analysis of power quality data. On EQ Gateways, JupyterLab ships as the optional eq-lab package (sudo apt install eq-lab), which provides JupyterLab plus equser, an open-source Python package for working with EQ Wave data. The flask icon appears in the sidebar only when eq-lab is installed.

Overview

JupyterLab provides an interactive analysis environment for users who need to:

• Perform custom calculations on waveform data
• Create specialized visualizations
• Develop automated analysis workflows
• Export data in custom formats

Getting Started

When you open the JupyterLab tab:

1. A new notebook session starts automatically
2. The equser package is installed in the eq-lab environment and ready to use
3. Sample notebooks in tutorials/, analysis/, and tools/ demonstrate common analysis patterns

You can also install equser on any computer with pip install equser to work with exported data or connect to a gateway remotely.

Working with CPOW Waveform Data

Load continuous point-on-wave (CPOW) parquet files directly from the gateway’s storage:

from equser.data import load_cpow_scaled

# Load a CPOW parquet file (32 kHz, 7 channels)
data = load_cpow_scaled('/var/lib/eq-coherence/data/cpow/20250615_120000.parquet')

# Scaled voltage and current arrays are ready to use
print(f"Phase A voltage range: {data['VA'].min():.1f} to {data['VA'].max():.1f} V")
print(f"Start time: {data['start_time']}")
print(f"Sample rate: {data['sample_rate']} Hz")
print(f"Samples: {len(data['VA']):,}")

The load_cpow_scaled() function automatically handles raw int32-to-float scaling using the vscale/iscale metadata embedded in each parquet file.

Querying Data via the REST API

The gateway exposes a REST API at port 8080. See the API Reference for endpoint details.

Power Monitoring Data (Arrow IPC)

from equser.api import GatewayClient

# The gateway proxies /api on the normal HTTP port, so no port is needed.
client = GatewayClient('http://localhost')

# Fetch recent PMon data for a device (returns a pyarrow.Table)
table = client.get_pmon_data(
    'wave-001',
    start_time='2025-06-15T12:00:00Z',
    end_time='2025-06-15T13:00:00Z',
)

df = table.to_pandas()
print(df.columns.tolist())

SQL Queries

# SELECT-only SQL against the gateway's DuckDB store; returns a pyarrow.Table.
table = client.query_sql(
    "SELECT time_us, AVRMS, BVRMS, CVRMS FROM pmon_data ORDER BY time_us DESC",
    limit=100,
)
df = table.to_pandas()

Live Streaming via WebSocket

from equser.api import connect_spectral_stream

# The spectral stream is a broadcast consumer of the live CPOW feed. Each
# binary window is a pyarrow.Table of FFT magnitudes; text messages are gap
# markers (dicts).
for item in connect_spectral_stream(channels=['VA', 'IA'], cycles=12):
    if isinstance(item, dict):
        continue  # gap marker
    print(f"Spectral window: {item.num_rows} bins, channels {item.column_names}")
    break  # Remove to stream continuously

Plotting Tools

The equser package includes plotting utilities for common visualizations:

from equser.plotting import PowerMonitorPlotter, WaveformPlotter

Waveform analysis helpers such as find_zero_crossings(), extract_complete_cycles(), and plot_extracted_cycles() are available for detailed cycle-level inspection.

Saving Work

Your notebooks are automatically saved to your user workspace. You can also:

• Download notebooks to your local computer
• Export results as CSV, PNG, or PDF

Resources

• Sample Notebooks: Browse the tutorials/, analysis/, and tools/ directories in the file browser
• API Documentation: Built-in help via help(equser)
• equser Package: See from equser import pmon, plotting, analysis, data, api for available modules

Performance Notes

JupyterLab runs on the gateway alongside EQ Coherence and EQ Sight. For large data analyses:

• Use time filters to limit data volume
• Consider downsampling for trend analysis
• Export large datasets for processing on dedicated workstations

EQ Syntropy

EQ Syntropy is a cyber-physical AI (CPAI) framework — AI whose reasoning is grounded in continuous physical measurement from EQ Wave sensors via EQ Coherence. It runs analyses continuously (event detection, compliance monitoring) and on demand (natural language queries, root cause investigation, report generation).

Event Detection and Classification

EQ Syntropy analyzes the waveform and metric streams that EQ Coherence records:

• Voltage events: Sags, swells, and transients detected per-phase with severity scoring based on depth, duration, and recovery characteristics
• Harmonic analysis: FFT-based harmonic extraction (orders 1–50+), THD computation, and distortion trending
• Frequency deviation: Grid frequency monitoring against nominal (50/60 Hz)
• Severity scoring: Events are scored on a continuous scale incorporating magnitude, duration, and impact, not just binary threshold alerts

Detected events are automatically classified, timestamped, and annotated with contextual metadata for investigation.

Standards Compliance

EQ Syntropy checks power quality against industry standards:

• IEEE 519: Voltage and current harmonic distortion limits
• SEMI F47: Semiconductor fabrication voltage immunity requirements
• ITIC/CBEMA: Voltage tolerance envelope for IT equipment (upper and lower bounds by event duration)

IEC 61000-4-30 compliance evaluation is planned.

Agentic Investigation

EQ Syntropy runs an agentic AI system that investigates power quality conditions rather than just reporting metrics:

• Semantic router: Incoming queries are analyzed and directed to the most appropriate processing pathway, whether that is a SQL analysis, a physical model, a standards lookup, or a multi-step AI investigation
• Tool execution: The AI orchestrates tools including waveform queries, statistical analysis, domain knowledge lookups, and hypothesis generation to build a complete picture
• Knowledge integration: Investigations draw on facility context (single-line diagrams, equipment specifications, maintenance history) alongside waveform evidence
• Waveform similarity search: Vector embeddings of waveform events enable “find events like this one” queries across months or years of history
• Report generation: Structured incident reports and compliance documentation with supporting waveform evidence

All AI reasoning is grounded in electrical engineering principles. Answers include specific evidence from the underlying measurements, not summaries or statistics alone.

How It Works

When you ask a question through EQ Sight or the API:

1. The semantic router classifies your intent and selects the appropriate tools and models
2. Relevant waveform data, events, and facility context are retrieved
3. Analysis tools run against the data: SQL queries, compliance checks, trend analysis, similarity search
4. The AI synthesizes findings into a response with supporting evidence from the raw waveforms

This enables questions like:

• “What caused the voltage sag at 3:47 PM yesterday?”
• “Are there any signs of equipment degradation in Building A?”
• “How does the harmonic distortion compare to last month?”
• “What’s the likely root cause of the nuisance trips on Line 3?”
• “Show me events similar to the transient on Phase B last Tuesday”
• “Are we in compliance with SEMI F47 on the fab power feed?”

Local AI Inference

EQ Syntropy runs AI models locally on the gateway or edge AI hardware. No data leaves the facility unless EQ-hosted or data lake sync is enabled. Local inference supports:

• Natural language query processing
• Event classification and enrichment
• Hypothesis generation for root cause analysis
• Waveform embedding and similarity search

EQ Syntropy vs. EQ Coherence

EQ Syntropy builds on EQ Coherence. EQ Coherence runs on all deployments and is required.

Deployment Options

EQ Syntropy can run on:

• Edge AI Hardware: NVIDIA AGX Orin or Thor platforms for local processing
• EQ-Hosted: Energy Quotient managed servers for deployments without local GPU hardware
• Customer-Hosted: On-premises deployment for enterprise requirements

See Deployment Options for details on choosing the right configuration.

Beyond Power Quality

EQ Syntropy is initially focused on power quality and power monitoring data from EQ Wave, but its data pipeline, analytics engine, and agentic AI framework are adaptable to other continuous time-series domains — environmental monitoring, vibration analysis, process control, and other sensor-rich systems. Contact [email protected] to discuss ingestion and integration needs.

Security Architecture

Energy Quotient deployments are designed for environments where data sovereignty, network isolation, and auditability are non-negotiable. This page describes the security architecture across the platform stack.

Network Security

Zero-Trust Mesh Networking

All device-to-device communication uses a certificate-based, zero-trust mesh network built on Nebula (Apache 2.0, proven at scale). Every gateway, sensor node, and administrative device is authenticated by a CA-signed certificate that cryptographically defines its identity and network permissions.

Key properties:

• Mutual authentication on every connection — no implicit trust based on network location
• Certificate-based identity — each device has a unique certificate with embedded group memberships and an IP assigned from a private overlay range (100.64.0.0/16)
• No open inbound ports — gateways initiate outbound UDP connections with NAT traversal; no listening ports exposed to the public internet
• Cryptographic client isolation — certificate groups enforce network segmentation so one customer’s devices cannot reach another’s
• Geographically redundant lighthouses — distributed peer discovery with no single point of failure
• Open-source and auditable — no proprietary VPN components; full source available for review

Firewall

Each gateway runs UFW (Uncomplicated Firewall) with a default-deny incoming policy. Only explicitly allowed traffic is permitted:

SSH is not accessible from the public internet or the facility LAN unless a site administrator explicitly adds a firewall rule for their subnet. See SSH Access for details.

Intrusion Detection

fail2ban monitors authentication logs and automatically blocks IP addresses after repeated failed login attempts. VPN and sensor subnets are whitelisted since they are already authenticated at the certificate layer.

Physical Isolation

EQ Wave sensors connect to the gateway through plastic optical fiber (100Base-FX), providing complete electrical isolation between monitored power systems and the data network. There is no conductive path between the sensor’s measurement circuits and the gateway or facility LAN.

This two-stage isolation architecture means:

• Electrical faults on the monitored system cannot propagate to the network
• The sensor network is physically air-gapped from the facility LAN
• No routing exists between the sensor subnet (192.168.10.0/24) and the LAN interface

Authentication and Access Control

User Account Model

Each gateway provisions three user accounts with distinct privilege levels:

SSH Hardening

• Key-only authentication for vendor accounts (password authentication disabled)
• AllowUsers directive restricts which accounts can log in via SSH
• Root login disabled — no direct root SSH access
• Per-device authorization — vendor SSH keys are baked into the provisioning manifest; no shared credentials across sites

Scoped Sudo

The customer admin account has scoped sudo access limited to:

• EQ service management (systemctl for EQ services)
• Firewall rule management (ufw)
• User and group management
• Network configuration (nmcli)

Full system-level sudo is reserved for the vendor account.

Data Sovereignty

Local-First Architecture

All waveform data is recorded and stored locally on the gateway by default. No data leaves the site unless explicitly configured:

• CPOW waveform data is written to local storage (NVMe or microSD depending on platform)
• Analytics and event detection execute on the gateway — no cloud round-trips required
• Data export is an explicit action via REST API, Parquet file transfer, or configured rsync
• Remote support access traverses the encrypted Nebula overlay and can be disabled

Air-Gap Capability

The platform operates with no cloud dependencies. For environments that require it:

• Gateways function fully without internet connectivity
• Sensor data collection, storage, and local visualization continue offline
• Nebula overlay and remote support can be disabled entirely
• All software updates can be applied via local media

Audit Trail

Three independent audit layers provide traceability:

1. Certificate identity — which device connected, verified by CA-signed certificate
2. Authentication logs — who logged in, tracked by SSH and system authentication
3. Application logs — what was done, recorded by EQ Coherence and system services

All logs are stored locally and accessible to the site administrator.

Supply Chain

• Nebula (networking): Apache 2.0, open-source, maintained by the Defined Networking team (originally developed at Slack)
• EQ Coherence (data collection): Rust-based server with deterministic builds
• EQ Sight (visualization): TypeScript/React frontend served locally
• SBOM publication and formal supply chain attestation are on the 2026 roadmap

Standards Alignment

The platform’s security architecture draws on principles from these standards and frameworks. Formal certification is on our roadmap; today these reflect design choices, not compliance claims.

• NIST SP 800-82 — Guide to OT Security (network segmentation, access control, monitoring)
• IEC 62351 — Power systems communication security (authentication, access control)
• NERC CIP — Critical infrastructure protection concepts (electronic security perimeters, access management)
• Zero Trust Architecture (NIST SP 800-207) — No implicit trust; verify every connection

Contact

For security questions, vulnerability reports, or to request documentation for your compliance review, contact [email protected].

Deployment Options

System Topology

Every EQ deployment places components at different points in the system:

● included | ○ optional

EQ Wave sensors connect to the gateway via a fiber optic media converter (included with the EQ Wave package). EQ Coherence and EQ Sight run on every gateway. EQ Syntropy is optional and can run locally on GPU-equipped gateway hardware, on your own servers, or on EQ-hosted infrastructure.

EQ Sight access. EQ Coherence serves EQ Sight two ways by default: locally over the facility LAN for on-site browser access, and as the API backend for cloud-hosted EQ Sight at the [site].pq.app subdomain. Both become available when the gateway is connected to the facility LAN. This is a gateway role common to every deployment, independent of the hardware or AI option.

Gateway Hardware

The gateway runs EQ Coherence and EQ Sight. Any of the following can serve as the gateway:

Managed vs. Flex:

• Managed (EQ-provided hardware): Pre-configured and ready to connect. Includes VPN connectivity for remote support. Recommended for most deployments.
• Flex (customer-provided hardware): You install EQ Coherence on a Debian-based Linux computer. You manage the hardware and software updates. See Flex Deployment for setup.

Aside: The open-source equser Python package can capture PMon data (aggregated PQ metrics at 200 ms intervals) on any OS with Python, without the full EQ Coherence software. This is useful for lightweight integration and research but does not provide Continuous Point-on-Wave (CPOW) recording, the REST API, or EQ Sight. See the equser documentation for details.

Data Storage

EQ Coherence is storage-agnostic, supporting microSD via USB adapter, NVMe SSD, USB SSD, or network-attached storage. Optional remote storage provides long-term retention and cross-site access.

See Data Storage for local and remote storage options, and Storage Media for capacity planning and hardware recommendations.

EQ Syntropy (AI Capabilities)

EQ Syntropy is optional and independent of the gateway hardware choice:

When AI is enabled, EQ Syntropy provides physics-informed analytics, anomaly detection, and AI-assisted event analysis. Results appear in EQ Sight.

Setup

All deployments share these setup procedures:

• LAN Configuration
• Wi-Fi Configuration
• SSH Access

EQ Gateway

The EQ Gateway is a computer running EQ Coherence and EQ Sight — continuous power quality data collection with built-in visualization and analysis. It connects to EQ Wave sensors via a dedicated Ethernet subnet, records CPOW and power monitoring (PMon) data to local storage, and serves the data through REST and WebSocket APIs.

Hardware Requirements

• Minimum 4 GB RAM
• Minimum 4 GB free space on root filesystem
• Two Ethernet ports (one for sensor subnet, one for facility LAN)
• Debian-based Linux OS (Debian, Ubuntu, etc.)
• Removable storage media for CPOW data rated for sustained high-volume writes. Do not write CPOW data to the root filesystem; use dedicated removable media. See Data Storage for rates, retention, and Storage Media for hardware guidance.

The gateway also requires an EQ Wave sensor with media converter to collect data. These are included with the EQ Wave package; see EQ Wave Installation for details.

Network Architecture

The gateway has two Ethernet ports: one connects to your facility LAN, and the other connects to the EQ Wave sensor subnet via a media converter. The media converter bridges between Cat 5e Ethernet (100Base-TX) and plastic optical fiber (100Base-FX), providing complete electrical isolation between the sensor and gateway.

Sensor Network

• EQ Wave sensor: 192.168.10.10 (fixed IP)
• Gateway sensor interface: 192.168.10.2 (fixed IP)
• Subnet mask: 255.255.255.0
• Isolation: No routing to facility network (fiber optic isolation maintained)

Facility Network

• Gateway LAN interface: DHCP or static IP (configurable via LAN Configuration)
• EQ Coherence web interface: HTTP on port 80 (proxied to backend on port 8080)
• EQ Coherence REST API: Port 8080 (see API Reference)

System connection diagram

Example system setup

Power Supply

Each device in the system (gateway, media converter, EQ Wave sensor) requires exactly one power source. Do not connect multiple power sources to the same device.

EQ Wave sensor: Self-powered from the monitored AC mains (85–528V) or DC input (4.5–36V on v1.2).

Media converter (use one):

• USB power from the gateway
• Dedicated DC adapter

Gateway power options vary by hardware. See your platform-specific page for details.

Software

All gateways run the same EQ Coherence software stack regardless of hardware platform. Capabilities include:

• REST API: HTTP-based data access (JSON and Arrow IPC formats)
• WebSocket: Live waveform and spectral streaming
• Additional protocols: Modbus TCP, MQTT, DNP3 available upon request
• Remote Support: VPN tunnel for authorized technical support (enabled by default, can be disabled)
• Network Isolation: Sensor network is physically isolated from facility LAN via fiber optic connection

Supported Platforms

• Compulab IOT-GATE / IOT-DIN — Current production industrial gateway
• Toradex Verdin iMX95 + Ivy — Next-generation production gateway (available Q2 2026)
• Raspberry Pi 5 (Lab Gateway) — Lab, workbench, demonstrations, and smaller deployments

Platform Not Listed?

EQ Coherence runs on most Debian-based platforms (e.g. Debian or Ubuntu) that meet the hardware requirements above. If your hardware is not in the table, we can confirm whether Coherence can be installed and which Ethernet ports it should bind. Run the following and send us the output:

# CPU architecture (determines which package build applies)
uname -m

# Platform model — one of these will return a value for your hardware
cat /sys/firmware/devicetree/base/model 2>/dev/null   # ARM boards (Raspberry Pi, Toradex, NVIDIA Jetson/DGX, etc.)
cat /sys/class/dmi/id/sys_vendor /sys/class/dmi/id/product_name 2>/dev/null  # x86 boards (Advantech, industrial PCs, etc.)

# Network interfaces (so we can identify the sensor and LAN ports)
ip -br link

The model string identifies the platform; the interface list and CPU architecture let us determine package compatibility and map the correct sensor and facility-LAN ports. Send all of the output, including any blank lines — an empty result for one command is itself informative (for example, the device-tree command returns nothing on x86 hardware).

Industrial Gateway: Compulab IOT-GATE / IOT-DIN

Compulab IOT-GATE-iMX8PLUS

EQ Gateway DIN-rail mounted in instrumentation panel

Models

• IOT-GATE-iMX8PLUS — Panel mount
• IOT-DIN-iMX8PLUS — DIN-rail form factor

Technical Specifications

• Processor: NXP i.MX8M-Plus quad-core ARM Cortex-A53
• Power: 8–36V DC (standard) or PoE 802.3af/at (IOT-GATE only), 5–12W typical
• Operating Temperature: -40°C to +80°C (industrial grade)
• Network: Dual Gigabit Ethernet, 802.11ac dual-band WiFi, optional cellular
• Boot media: eMMC (internal)
• Data storage: microSD via USB adapter (see Storage Media)

Included Storage

• SanDisk Ultra 1.5 TB microSDXC UHS-I Memory Card — approximately one month of retention for a typical 3-phase deployment
• SanDisk MobileMate USB 3.0 microSD Reader (SDDR-B531-AN6NN)

Power Supply Options

Gateway power options (use one):

• PoE 802.3af/at via the LAN port (IOT-GATE panel-mount model only; not available on IOT-DIN)
• Wall power adapter to DC (Compulab IOTG-ACC-PSU)
• Existing DC supply (8–36V, 36W)

The left diagram below shows all possible power paths for each device. It is a reference — do not use all paths simultaneously. The right diagram shows a recommended configuration using PoE for the gateway with USB passthrough to the media converter.

All power options by device (use one per device)

Recommended: PoE gateway + USB-powered media converter

Ethernet Ports

The Compulab gateway has two Ethernet ports with two different roles:

• Sensor port — Connects to the EQ Wave sensor via the media converter. Fixed at 192.168.10.2/24.
• LAN port — Connects to your facility network. Configurable via LAN Configuration. PoE power input, if used, connects here.

For other gateway hardware (Raspberry Pi, flex deployments), port assignments differ. See Port Assignments for the complete table.

LED Status Indicators

The gateway uses two bi-color (green/red) LEDs to show service status:

• Power Monitor (pmon) Service:
  • Green: Blinks at ~2.5Hz (every 10/12 cycles) during normal operation
  • Red: Lights up to indicate error conditions
• Waveform (wave) Service — typically the 2nd LED:
  • Green: Blinks at ~2.5Hz during normal operation
  • Red: Lights up to indicate error conditions

The gateway’s RJ45 Ethernet ports have amber LEDs for link and green LEDs for activity.

The gateway also has a green LED on the power button, to indicate the device is powered up.

For EQ Wave sensor and media converter LED indicators, see EQ Wave Installation.

Lab Gateway: Raspberry Pi

EQ Lab Gateway (Raspberry Pi 5, 16 GB RAM, 1 TB NVMe SSD)

The Lab Gateway is built on Raspberry Pi 5. It runs the same EQ Coherence software stack as the industrial gateways but is suited for lab environments, workbench testing, demonstrations, and smaller deployments where industrial hardening is not required.

Technical Specifications

• Platform: Raspberry Pi 5, 16 GB RAM
• Processor: Broadcom BCM2712 quad-core ARM Cortex-A76
• Storage: 1 TB NVMe SSD (boot + data)
• Power: 5V/3A USB-C, 10–15W typical
• Operating Temperature: 0°C to 50°C
• Enclosure: Branded steel case with fan cooling
• Network: Built-in Gigabit Ethernet + USB 3.0 Gigabit Ethernet adapter (included)

Network Configuration

The Lab Gateway ships with dual Ethernet pre-configured:

• eth0 (built-in): Sensor network (EQ Wave via media converter)
• USB Ethernet adapter: LAN / facility network

This is the opposite of Compulab gateways where eth1 is typically the sensor interface.

Display & Kiosk Mode

Lab Gateways with a connected monitor can operate in two display modes:

• Kiosk mode (demo user): Chromium opens fullscreen showing EQ Sight. No desktop, no taskbar. Ideal for wall-mounted displays and demonstration units.
• Desktop mode (eqadmin or admin user): Standard desktop environment with full access to terminal, file manager, and browser.

User accounts

The gateway auto-logs in to whichever user is configured in setup.yml under users.autologin (default: eqadmin).

Switching between modes

Using raspi-config

To change which user auto-logs in:

sudo raspi-config

Navigate to System Options > Boot / Auto Login and select:

• Desktop Autologin — logs in as the current user (shown in the menu)
• Desktop — requires manual login at the greeter (pick any user)

After changing, reboot for it to take effect:

sudo reboot

Using setup.yml (during provisioning)

Set the users.autologin field before running sudo eq system provision:

users:
  autologin: "demo"     # Kiosk mode (Chromium fullscreen)
  # autologin: "eqadmin"  # Desktop mode (default)

Exiting kiosk mode

If the gateway is in kiosk mode and you need temporary access to the desktop:

Chromium will relaunch in kiosk mode on the next reboot.

To permanently exit kiosk mode, switch the autologin user to eqadmin using raspi-config as described above, then reboot.

Kiosk configuration in setup.yml

Kiosk mode is configured during provisioning via the kiosk section:

kiosk:
  enabled: false                     # Set true for display deployments
  hostname: "{site}.pq.app"         # Site hostname for EQ Sight
  hide_cursor: true                  # Hide mouse cursor when idle
  disable_blanking: true             # Prevent screen from blanking

When enabled, provisioning will:

1. Install Chromium and unclutter (cursor hider)
2. Add a localhost entry in /etc/hosts so the hostname resolves locally, bypassing Cloudflare Access
3. Create autostart entries in the demo user’s home directory
4. Disable screen blanking and low-voltage warnings for a clean display

Lab Kit

The Lab Gateway is available as a complete lab kit including an EQ Wave sensor, media converter, POF cable, and all accessories needed for benchtop or panel deployment. Contact [email protected] for details.

Data Storage

Every EQ deployment needs local storage for waveform and metrics data. Optional remote storage provides long-term retention and cross-site access. EQ Coherence is storage-agnostic; choose the media and architecture that fits your facility.

Local Storage

Local storage is required for all deployments. It buffers incoming data and provides near-term access through EQ Sight. Larger media extends local retention and reduces write wear.

CPOW data generates approximately 30 GB per day for a typical 3-phase deployment (losslessly compressed int32 waveform data; lighter configurations use less). Industrial gateways ship with 1.5 TB storage, providing approximately 7 weeks of retention. Lab gateways ship with 1 TB, providing over 5 weeks.

See Storage Media for capacity planning, recommended hardware, formatting, and swap procedures.

Remote Storage

Remote storage replicates data from EQ Coherence to a remote server for long-term retention, backup, and cross-site access (typical sync latency ~1 min). It is optional for all deployments except EQ-hosted AI, which requires it. Contact us for details on remote storage options.

When remote storage is enabled, EQ Coherence retains local data for near-term access while the server handles archival. Both copies are accessible through EQ Sight.

Storage Media

This guide covers storage hardware selection, capacity planning, endurance, and swap procedures. See Data Storage for storage rates and retention estimates.

Choice of Storage Hardware

Basic Requirements

• USB storage drive, SD card, or other removable media
• USB port on gateway (USB 3.0 required for optimal performance)
• Minimum write speed of 10 MB/s for CPOW recording

Recommended Features

• Industrial temperature rating (-25°C to 85°C) for harsh environments
• Published endurance ratings (terabytes written, TBW)
• UHS-I or better for optimal performance

Performance Considerations

• Interface speed affects data retrieval and visualization:
  • Use the blue USB 3.0 port on front panel for best performance
  • Back panel USB 2.0 ports will limit transfer speeds
• Sustained write speed affects recording reliability:
  • Minimum required: 10 MB/s (e.g., V10 speed class for SD cards)
  • Recommended: 30 MB/s or better (e.g., UHS-I for SD cards)
• Although CPOW data averages less than 1 MB/s, data is written in batches and derived data may be added during postprocessing

Storage Rotation

EQ Coherence automatically manages disk space by removing the oldest CPOW data files when free space drops below 4 GB (configurable). If remote storage is enabled, already-replicated files are removed first. Unsynced files are only removed when necessary.

With typical data rates (see Data Storage), 1.5 TB provides approximately 7 weeks of local retention before rotation begins. If your deployment does not include remote data storage, back up any important data before it is overwritten or in case of media failure.

Storage Endurance

• Cards/drives with published endurance ratings (TBW) are strongly recommended for continuous recording
  • TBW is the product of rated writes per cell and capacity
  • However, actual endurance varies significantly between manufacturers and models
• With a modest 3000 write cycles per cell, endurance varies by capacity:
  • A 64 GB card would last about 17 years (64 GB × 3000 ÷ 30 GB/day)
  • A 1 TB card would last about 270 years (1000 GB × 3000 ÷ 30 GB/day)

Included Hardware

Industrial gateways ship with 1.5 TB microSD storage; lab gateways include 1 TB NVMe. For replacements or self-sourced media:

• SanDisk MobileMate USB 3.0 Card Reader (SanDisk part number SDDR-B531-AN6NN)
• SanDisk 1.5 TB Ultra UHS-I microSDXC Memory Card (or 1 TB for lab gateways):
  • Operating temperature -25°C to 85°C
  • V10 speed class (10 MB/s minimum)
  • UHS-I interface
  • 150 MB/s sequential read
  • 10-year manufacturer warranty
  • Write endurance not rated

SanDisk Ultra UHS-I microSDXC Memory Card and MobileMate USB 3.0 Card Reader

Extended Storage with SSDs

For higher endurance and capacity than a microSD card, a 2.5“ SSD is the better choice for long-term continuous recording. SSDs connect over USB 3.0 either as a native USB drive or as a SATA 2.5“ SSD paired with a SATA-to-USB 3.0 adapter, and they support up to 4 TB (see Data Storage). Their key advantage is a published write-endurance rating: the bundled microSD card is not endurance-rated, whereas a NAS-class SSD is rated for hundreds to thousands of TBW.

A representative option is the WD Red SA500 NAS SATA SSD, designed for 24/7 write workloads:

• Capacities from 500 GB to 4 TB
• Endurance from 350 TBW (500 GB) to 2,500 TBW (4 TB)
• 560 MB/s read, 530 MB/s write, 5-year warranty
• Operating temperature 0°C to 70°C

DIN-rail mounting. In a control cabinet, a 2.5“ SSD can be secured with a DIN-rail SSD frame (a bracket that holds the drive on standard 35 mm DIN rail) so the drive is mounted alongside the gateway rather than left loose on the USB cable.

If you are unsure which media fits your deployment, capacity target, or environment, contact us for guidance.

Swapping Storage Media

Prerequisites

• New storage media (pre-formatted from Energy Quotient, or bring your own)
• SSH access to gateway if you are using your own storage media

Steps

Picture of gateway including power button, power LED, and SD card with adapter

1. To safely shut down the gateway:

   • Press and hold the power button for 1 second, then release
   • Wait for the LED to turn off (about 2–3 seconds) before proceeding

2. Remove the storage media:

   • For SD cards: Remove card from adapter (may be easier to remove adapter from USB first)
   • For USB devices: Unplug from the USB port

3. Insert the new storage media:

   • For SD cards: Insert new card into adapter and reinsert into USB port
   • For USB devices: Plug directly into USB port

4. Press and hold the power button for 1 second to turn the gateway back on.

5. If you are using your own storage media, follow the Storage Media Setup instructions to format and prepare the media.

Storage Media Setup

These instructions are for users who purchase their own storage media rather than using pre-formatted media from Energy Quotient. This process will prepare the media for use with EQ Coherence.

1. Check Device

   # Insert SD card into adapter and adapter into gateway's USB port
   ls /dev/sda*
   
   # Confirm that /dev/sda1 exists
   # If not, try:
   # sudo mknod /dev/sda1 b 8 1
   
   # Useful commands for device verification:
   lsblk -o name,uuid,label
   sudo blkid
   

2. Partition and Format

   # Create a new partition table and single partition using all space
   sudo parted -s /dev/sda mklabel gpt
   sudo parted -s /dev/sda mkpart primary ext4 0% 100%
   
   # Format the card with no percentage-based reserved space
   sudo mkfs.ext4 /dev/sda1 -L eqdata -m 0
   
   # Reserve 100MB (25600 4K blocks) for log management
   sudo tune2fs -r 25600 /dev/sda1
   
   # Mount and set up permissions
   sudo mount -a
   sudo rm -r /mnt/eqdata/lost+found/
   
   # Set correct ownership and permissions for synapse group
   sudo chown synapse:synapse /mnt/eqdata
   sudo chmod 2775 /mnt/eqdata
   

3. Verify

   # Check mount points, sizes, and device info
   df
   sudo blkid
   
   # Check filesystem integrity (unmount first)
   sudo umount /dev/sda1
   sudo e2fsck -f /dev/sda1
   sudo mount /dev/sda1 /mnt/eqdata
   

4. Remove or start using the new media

   # To remove the media:
   sudo umount /dev/sda1
   
   # Or to restart and use the new media:
   sudo shutdown -r now
   

AI Compute

EQ Syntropy is optional and requires GPU-accelerated hardware. Every gateway runs EQ Coherence and EQ Sight without it. When you want Syntropy’s physics-informed analytics, event detection, and AI-assisted diagnostics, choose where to run it:

All options deliver the same Syntropy capabilities through EQ Sight. The choice depends on your network environment, data policies, and existing infrastructure.

With edge AI or customer-hosted deployments, AI inference runs entirely on local hardware. There are no per-query fees, no usage limits, and no dependency on external services.

EQ-hosted deployments provide the same Syntropy capabilities with EQ managing the AI infrastructure on your behalf.

Edge AI platforms

Three on-site GPU platforms, from hardened industrial edge to a high-capacity developer node:

Orin and Thor are hardened, always-on industrial edge units. DGX Spark is a developer/evaluation-class desktop (the AI-compute analogue of the Lab Gateway) — use it for prototyping, benchmarking, and high-capacity on-prem evaluation, not hardened field deployment.

Software stack

Edge AI deployments run the same EQ stack regardless of platform:

AI Gateway: NVIDIA AGX Orin

The Neousys NRU-220S with NVIDIA Jetson AGX Orin provides edge AI capabilities for EQ deployments requiring on-premises intelligence. The fanless industrial system combines gateway functionality with GPU-accelerated processing.

When to Choose AGX Orin

• On-premises AI: GPU-accelerated analytics without cloud dependency
• Multi-sensor gateway: Up to 5 EQ Wave sensors per unit via direct Ethernet connection (6 ports; 1 reserved for facility LAN). Multi-sensor configurations available upon request.
• Air-gapped environments: Defense, critical infrastructure, or facilities with strict network isolation

Platform Specifications (Neousys NRU-220S)

Network Configuration

The NRU-220S has six Ethernet ports with screw-lock connectors:

For installations with more sensors than available ports, use a managed switch on the sensor subnet.

See LAN Configuration for detailed network setup.

AI Capabilities

When EQ Syntropy is enabled on the AGX Orin, the GPU provides hardware acceleration for physics-informed analytics, event detection, and diagnostic models. AI inference runs locally with no per-query costs, no usage limits, and no external connectivity required. AI capabilities are accessed through EQ Sight.

See EQ Syntropy for details on AI features.

Software Stack

The AGX Orin runs JetPack (Ubuntu-based) with CUDA/TensorRT. The EQ stack (Coherence, Syntropy, Sight) is the same across all edge AI platforms — see the AI Compute overview.

Storage Requirements

CPOW data storage scales with sensor count and retention period:

Figures assume a typical 3-phase deployment. See Data Storage for details. NVMe SSD recommended.

Installation

EQ provides pre-configured AGX Orin systems with all software installed. On-site installation involves:

1. Mount the NRU-220S and connect power
2. Connect facility LAN to Port 1
3. Connect EQ Wave sensors via media converters to the GbE PoE+ ports
4. Access EQ Sight via web browser

Contact [email protected] for deployment assistance.

AI Gateway: NVIDIA Thor

The Advantech MIC-742-AT with NVIDIA Jetson Thor provides edge AI compute for large-scale EQ deployments. The 100GbE QSFP28 interface supports high-density sensor networks via managed switching, making it suitable for campus-scale monitoring.

When to Choose Thor

• Large-scale monitoring: High sensor density across a facility or campus
• Maximum AI throughput: 2,070 FP4 TFLOPs for demanding inference workloads
• High-bandwidth aggregation: QSFP28 with 4x 25GbE breakout for sensor network fabric
• Industrial environments: Extended operating temperature range (-10°C to 60°C)

Platform Specifications

Network Configuration

The QSFP28 port supports 4x 25GbE lanes via standard breakout cables. Each lane connects to a managed switch serving multiple EQ Wave sensors via media converters.

For how Thor compares to the other edge AI platforms (Orin, DGX Spark, and the no-AI gateway), see the platform table in the AI Compute overview.

Current Status

EQ is actively developing the Thor deployment option. Hardware (MIC-743-AT) is in hand for integration and validation.

AI Compute: NVIDIA DGX Spark

The NVIDIA DGX Spark is a compact desktop “AI supercomputer” built on the GB10 Grace Blackwell Superchip. It provides high-capacity on-premises AI compute for EQ Syntropy, with enough unified memory to run large reasoning models locally.

When to Choose DGX Spark

• Developer / data-science workstation: a workstation for building, fine-tuning, and benchmarking Syntropy models locally, with 128 GB of unified memory for models up to ~200B parameters (up to ~405B across two linked units).
• High-capacity on-prem AI in a non-industrial setting: a climate-controlled lab, office, or server room where you want data-local inference and large local models without the industrial hardening of an edge unit or the rack-mounting of a GPU server.
• On-prem, no cloud: data and inference stay local, with no per-query fees or usage limits.

To evaluate Syntropy before committing to hardware, the simpler options are EQ-Hosted (cloud) or Customer-Hosted (your existing GPU). For hardened field or edge deployment, use AGX Orin or Thor.

Platform Specifications

Network Configuration

Like any gateway, DGX Spark connects to your facility LAN and to the EQ Wave sensor subnet via fiber media converters (the LAN side serves EQ Sight — see Deployment Options). A single sensor can connect directly; for several sensors, add a switch on the isolated sensor subnet — an unmanaged switch is sufficient, since the subnet uses fixed addressing. Its ConnectX-7 networking far exceeds what the sensor fabric requires.

AI Capabilities

With EQ Syntropy enabled, the Blackwell GPU accelerates physics-informed analytics, event detection, and diagnostic models, accessed through EQ Sight. Inference runs locally with no external connectivity required. See EQ Syntropy for AI feature details, and the AI Compute overview for the shared software stack and a platform comparison.

Current Status

EQ-Hosted AI

EQ-hosted AI runs EQ Syntropy on a dedicated EQ server. Your gateway syncs waveform and metrics data to the server, where Syntropy models run and return results to EQ Sight.

How It Works

1. Gateway collects data from EQ Wave sensors and stores it locally
2. Data syncs to the EQ server over your network connection
3. Syntropy runs physics-informed analytics on the server
4. Results are available through EQ Sight on the gateway and remotely via your [site].pq.app subdomain

Requirements

• Network connectivity from the gateway to the EQ server
• Remote data storage enabled on the gateway (the AI models run on the same server that stores the data)

Advantages

• No GPU hardware needed at your facility
• No AI software to install or maintain on-site
• Cross-site analytics available when multiple facilities sync to the same server
• Same Syntropy capabilities as edge AI deployments
• AI infrastructure managed by EQ; no per-query fees or usage limits

When to Choose EQ-Hosted

• Your facility has reliable network connectivity
• You prefer EQ to manage the AI infrastructure
• You want cross-site visibility across multiple deployments
• You do not have strict data residency requirements

For facilities that require all data to stay on-premises, see Edge AI or Self-Hosted options.

Customer-Hosted AI

For organizations with existing data center infrastructure, EQ Syntropy can be deployed on customer servers with full AI capabilities.

Overview

Customer-hosted deployment is designed for:

• Enterprise environments with established IT infrastructure
• Organizations with strict data residency requirements
• Air-gapped or restricted network environments
• Multi-site deployments requiring centralized management

AI inference runs locally with no per-query fees, no usage limits, and no external dependencies.

Requirements

Minimum Server Specifications

• CPU: 8+ cores, x86-64 architecture
• RAM: 32GB minimum, 64GB recommended
• Storage: 1TB NVMe SSD for application and recent data
• Network: Gigabit Ethernet connectivity to gateway devices

Recommended GPU (for full AI capabilities)

• NVIDIA GPU: RTX 4090 or A6000 class
• VRAM: 24GB minimum
• Driver: CUDA 12.x compatible

Operating System

• Debian 12 (Bookworm) or Ubuntu 22.04 LTS
• Docker and Docker Compose for containerized deployment
• Systemd for service management

Architecture

Customer-hosted deployments typically include:

1. Application Server: EQ Syntropy core services (Rust backend, Python tools)
2. Data Storage: Structured parquet files for waveform data, DuckDB for indexed queries
3. GPU Server: Optional, for AI inference workloads

Installation

Customer-hosted deployments use the same gateway software package as flex gateway deployments, installed on customer infrastructure.

Contact [email protected] for installation guidance and enterprise deployment assistance.

Setup

After the gateway and sensors are physically installed and powered, the remaining setup configures how the gateway connects to your network and how it is accessed for administration.

This section covers the facility-network and access configuration of the gateway. The sensor subnet itself uses fixed addressing and needs no configuration (see the Gateway overview for the sensor network details).

• LAN Configuration — connect the gateway to your facility network with DHCP or a static IP.
• Wi-Fi Configuration — optionally connect the gateway over wireless instead of, or in addition to, wired Ethernet.
• SSH Access — enable and secure remote shell access for administration.

LAN Configuration

The gateway’s LAN port can be configured to match your network requirements. By default, it is set to obtain an IP address automatically via DHCP.

Port Assignments

The gateway has two Ethernet roles:

• Sensor port — Connects to the EQ Wave sensor via the media converter. Fixed at 192.168.10.2/24. Not user-configurable.
• LAN port — Connects to your facility network. Configurable (DHCP by default).

Which physical port or network interface fills each role depends on the gateway hardware:

The gateway software creates two NetworkManager connection profiles that abstract away these differences: eq-sensor (sensor port) and eq-lan (LAN port). The nmcli commands in this guide use these profile names and work on all platforms.

Prerequisites

• SSH access to the gateway
• Choice and details of the network configuration:

  1. DHCP configuration:

     • Network CIDR from your administrator (e.g., 192.168.1.0/24)

  2. Static IP configuration:

     • Network details from your administrator:
       • IP address and subnet mask (e.g., 192.168.1.100/24)
       • Gateway address
       • DNS server addresses (optional)

  3. Direct PC connection:

     • Choose a static IP/subnet (<static-ip>/<subnet>) for the gateway. Any valid IP/subnet will work, but we recommend 192.168.1.2/24 because:
       • It’s a commonly used private network range
       • It allows direct connection while keeping the sensor connected

Steps

1. View Current Configuration

# List all network connections
nmcli connection show

# Show details for LAN connection
nmcli connection show "eq-lan"

# Show active connections (green indicates active)
nmcli -p connection show

2. Implement the Chosen IP Configuration

Choose one of the following configuration methods:

Option A: DHCP Configuration (Default)

To use automatic IP address configuration:

sudo nmcli connection modify "eq-lan" ipv4.method auto

Option B: Static IP Configuration

You might want to configure a static IP address in these scenarios:

1. Network policy requires static IP addresses
2. Direct connection to a PC/laptop (when not connecting to a LAN)

For LAN integration with static IP requirements:

sudo nmcli connection modify "eq-lan" \
    ipv4.method manual \
    ipv4.addresses "<ip-address>/<subnet-mask>" \
    ipv4.gateway "<gateway-address>" \
    ipv4.dns "<dns-servers>"

Example using typical values (coordinate with your network administrator):

sudo nmcli connection modify "eq-lan" \
    ipv4.method manual \
    ipv4.addresses "192.168.1.100/24" \
    ipv4.gateway "192.168.1.1" \
    ipv4.dns "8.8.8.8,8.8.4.4"

For direct PC connection:

sudo nmcli connection modify "eq-lan" \
    ipv4.method manual \
    ipv4.addresses "<static-ip>/<subnet>"  # Use your chosen static IP and subnet from prerequisites

3. Activate the New Configuration

sudo nmcli connection down "eq-lan"
sudo nmcli connection up "eq-lan"

4. Verify Configuration

After applying the configuration:

1. For LAN connection:

   • Connect the gateway’s LAN port to your network
   • Check the gateway’s IP address:

     nmcli connection show "eq-lan" | grep IP4.ADDRESS
     

2. For direct PC connection:

   • Set your PC’s network adapter to a static IP in the same subnet as your chosen static IP (but different from the gateway’s IP)
   • For example, with gateway IP 192.168.1.2/24:
     • Configure your PC as 192.168.1.3/24
   • Connect an Ethernet cable between your PC and the gateway’s LAN port
   • Your gateway’s IP address will be your chosen static IP

5. Allow SSH Access

The gateway’s firewall only allows SSH from the sensor subnet (192.168.10.0/24) and the VPN (100.64.0.0/16) by default. To access the gateway via your LAN, add a firewall rule for your subnet:

# Replace <network-cidr> with your subnet
# For example: 192.168.2.0/24 or 10.0.1.0/24
sudo ufw allow from <network-cidr> to any port ssh

6. Verify SSH Access

Test the new configuration by connecting to the gateway following SSH Access

Troubleshooting

If you lose network connectivity or cannot connect:

• Check network cable connections
• Verify IP address configuration with nmcli connection show "eq-lan"
• Review firewall rules with sudo ufw status
• Test network connectivity

You can always access the gateway through the sensor port if needed.

Wi-Fi Configuration

If your gateway is equipped with a Wi-Fi transceiver, you can follow the steps below to connect to Wi-Fi networks in range. Raspberry Pi gateways have built-in Wi-Fi and do not require external antennas.

Steps

1. Connect Antennas (Compulab only)

Compulab IOT-GATE gateways require external Wi-Fi antennas. Connect the antennas to the WLAN-A/BT and WLAN-B ports. Tighten securely but not aggressively.

• Keep antennas at least 20cm away from large metal objects
• Maintain clear line of sight to your Wi-Fi router when possible
• Both antennas must be connected for optimal MIMO performance

2. Add Networks

Your gateway can be pre-configured with Wi-Fi network settings during ordering. If you need to modify or add networks later, use the manual configuration steps below.

Manual Configuration

# Enable Wi-Fi radio (if disabled)
sudo nmcli radio wifi on

# List available networks
sudo nmcli device wifi

# Connect to a network
sudo nmcli device wifi connect "<AP name>" password "<password>"

Managing Saved Networks

# List saved connections
nmcli connection show

# Delete a saved network
nmcli connection delete "<connection_name>"

# Modify connection priority
nmcli connection modify "<connection_name>" connection.autoconnect-priority <number>

3. Verify Connection

# Check connection status
nmcli connection show
nmcli device status

# Test internet connectivity
ping -c 4 8.8.8.8

# Check signal strength
nmcli device wifi list --rescan yes

4. SSH Access Over Wi-Fi (optional)

To enable SSH access over Wi-Fi:

1. Find the Wi-Fi network CIDR:

# Get Wi-Fi IP and subnet information
ip addr show wlan0 | grep "inet "

2. Allow SSH access from the Wi-Fi network:

# Replace <network-cidr> with your Wi-Fi network
# For example: sudo ufw allow from 192.168.1.0/24 to any port ssh
sudo ufw allow from <network-cidr> to any port ssh

3. Note your gateway’s current IP address for future access

   ip -brief addr show wlan0
   

4. Test the new configuration by connecting to the gateway following SSH Access

Troubleshooting

Basic Diagnostics

# Show detailed device information
sudo nmcli device show wlan0

# Restart Wi-Fi interface
sudo nmcli radio wifi off
sudo nmcli radio wifi on

Common Issues

Poor Signal Strength

• Verify antenna installation
• Check antenna orientation
• Reduce distance to Wi-Fi router
• Remove metal obstacles

Connection Drops

• Check for interference from other devices
• Verify router settings (band, channel)
• Check system logs:

  journalctl -u NetworkManager | tail -n 100
  

Authentication Failures

• Double-check password
• Verify network security type (WPA2, WPA3)
• Ensure device time is synchronized
• Clear existing connection and retry:

  sudo nmcli connection delete "<connection_name>"
  sudo nmcli device wifi connect "<AP name>" password "<password>"
  

Advanced Configuration

Static IP Configuration

nmcli connection modify "<connection_name>" \
    ipv4.method manual \
    ipv4.addresses "192.168.1.200/24" \
    ipv4.gateway "192.168.1.1" \
    ipv4.dns "8.8.8.8,8.8.4.4"

Multiple Network Setup

# Add secondary network (same connection command as above)
sudo nmcli device wifi connect "<AP name>" password "<password>"

# Set connection priorities (higher number = higher priority)
nmcli connection modify "<connection_name 1>" connection.autoconnect-priority 100
nmcli connection modify "<connection_name 2>" connection.autoconnect-priority 50

Additional Resources

• nmcli Documentation (Manual Page)

SSH Access

The gateway has two Ethernet ports: the sensor port and the LAN port. We recommend starting setup through the sensor port, which always works with a direct Ethernet cable:

1. Initial SSH access — Connect directly to the sensor port (fixed at 192.168.10.2).
2. Configure LAN — Via SSH, follow LAN Configuration.
3. Ongoing access — SSH over the LAN port via your facility network or Wi-Fi.

Alternatively, if the gateway’s LAN port is connected to a DHCP-served network with internet access, and your service includes remote management (as pilot gateways do), our support team can assist remotely via VPN — including sharing the gateway’s LAN IP address and opening the firewall for local SSH access on your behalf. Contact [email protected] to arrange this.

If you prefer to set up access yourself, you can also SSH in via the LAN port if you know the DHCP-assigned address (e.g., from your router’s client list) — but you will first need to add your LAN subnet to the firewall via the sensor port (see Allow SSH Access).

Firewall

The gateway runs UFW (Uncomplicated Firewall) with a default-deny incoming policy. These ports are open after provisioning:

SSH is only allowed from the sensor subnet and the VPN. To access the gateway via your LAN, you must add a firewall rule for your subnet during LAN Configuration.

Prerequisites

For all access methods:

• Admin username and password from system documentation
• SSH client:
  • Windows: PowerShell (built-in), MobaXTerm (recommended), or PuTTY
  • macOS/Linux: Built-in terminal

For sensor port access:

• Ethernet cable
• Ability to configure PC network settings

For LAN port access:

• Network connection
• Gateway’s IP address (otherwise some options for finding it are suggested below)

Initial Access via Sensor Port

The sensor port provides a reliable way to access the gateway with its fixed configuration:

1. Connect your computer directly to the gateway’s sensor port using an Ethernet cable

2. Configure your computer’s Ethernet adapter with a static IP address:

   • IP Address: 192.168.10.3 (or any address in 192.168.10.0/24 except 192.168.10.2)
   • Subnet Mask: 255.255.255.0

3. Connect using your chosen SSH client with these parameters:

   • Host: 192.168.10.2
   • Username: admin
   • Port: 22
   • Connection type: SSH

   If using a terminal, the command is:

   ssh [email protected]
   

After establishing this initial connection, you should:

1. Configure the LAN port for your network following LAN Configuration
2. If supported and desired, connect to your wireless network following Wi-Fi Configuration
3. Test SSH access through the LAN port and/or Wi-Fi
4. Disconnect your computer and reconnect the sensor to the sensor port

The LAN port will then become your primary means of accessing the gateway, with Wi-Fi and the sensor port remaining available if needed.

Network Access (LAN/Wi-Fi)

1. Find your gateway’s IP address using one of the following methods

   Option A: Use the IP address from your network configuration

   • For LAN: See LAN Configuration
   • For Wi-Fi: See Wi-Fi Configuration

   Option B: Ask your network administrator

   Option C: Use your router’s admin interface

   • Access your router’s web interface (typically at 192.168.1.1 or similar)
   • Look for “Connected Devices” or “DHCP Clients”
   • Find the device named EQG-XXXX where XXXX is your gateway’s identifier (capitalized hex)

   Option D: Scan your network

   • Windows: Use Advanced IP Scanner (download)
   • Linux/macOS: Use nmap

     # Adjust this to match your network:
     sudo nmap -sn 192.168.1.0/24
     

2. Connect using your chosen SSH client with these parameters:

   • Host: <gateway-ip-address>
   • Username: admin
   • Port: 22
   • Connection type: SSH

   If using a terminal, the command is:

   ssh admin@<gateway-ip-address>
   

3. Enter the admin password when prompted

Need Help?

If you are unable to establish a connection, contact [email protected]. If the gateway has internet access via the LAN port, we can typically resolve access issues remotely via VPN. Otherwise, we can provide a USB debug cable and instructions for direct console access.