Published March 30, 2026 © LGPL

Ultra-Low-Power Industrial Data Logger with 11-Year Battery

A comprehensive guide to deploying the Norvi EC-M12-BC-C6-C-A for remote fuel monitoring with cellular connectivity

IntermediateProtip2 hours193

Ultra-Low-Power Industrial Data Logger with 11-Year Battery

Things used in this project

Hardware components

NORVI EC-M12-BC-C6-C-A Industrial Data Logger

ER34615H (D) Lithium Battery 3.6V 19,000mAh

4-20mA Diesel Fuel Tank Level Sensor

ST-Link V2 Programmer

Software apps and online services

Arduino IDE

STMicroelectronics STM32CUBEPROG

MQTT

ThingsBoard IoT Platform

Hand tools and fabrication machines

Digilent Mastech MS8217 Autorange Digital Multimeter

Wire Stripper & Cutter, 18-10 AWG / 0.75-4mm² Capacity Wires

Multitool, Screwdriver

Story

Introduction

Remote industrial monitoring presents a critical challenge: achieving reliable data collection over extended periods without frequent maintenance interventions. The Norvi EC-M12-BC-C6-C-A addresses this challenge through intelligent power management, achieving 11.4 years of battery life while maintaining real-time cloud connectivity.

This project guide demonstrates a complete industrial deployment for diesel fuel tank monitoring, covering hardware setup, firmware configuration, cloud integration, and real-world optimisation strategies.

Application Overview

This implementation monitors diesel fuel levels in remote storage tanks using:

4-20mA analog sensor interface for industry-standard instrumentation
NB-IoT/LTE-M cellular connectivity for wide-area coverage
MQTT over 4G for efficient cloud communication
ThingsBoard dashboard for real-time visualization and alerts

Key Achievement: Through optimised duty cycling and ultra-low-power design, the system operates for over a decade on primary lithium batteries—eliminating costly maintenance visits to remote sites.

Why This Matters

Traditional IoT solutions struggle with battery life in remote deployments:

Conventional cellular devices: 6-12 months battery life
Frequent site visits for battery replacement: $150+ per visit
System downtime during maintenance
Safety concerns accessing tanks in hazardous locations

The EC-M12-BC-C6-C-A's architecture solves these issues through:

STM32L0 ultra-low-power microcontroller (1.05µA sleep current)
Intelligent peripheral power management
Optimized cellular communication protocols
Robust fault detection and diagnostics

How It Works

System Architecture

[Fuel Sensor] → [ADS1115 ADC] → [STM32L0 MCU] → [SIM7070 Modem] → [4G Network] → [ThingsBoard]

↓ ↓ ↓

4-20mA 38, 000mAh Battery Web Dashboard

Operating Cycle

The EC-M12-BC-C6-C-A operates in a power-optimized duty cycle:

1. Wake Up Phase (60 seconds)

RTC alarm triggers system wake from shutdown mode
STM32L0 initializes peripherals
SIM7070 modem powers up and registers with cellular network
GPRS/LTE-M data connection established

2. Data Acquisition (2.5 seconds)

Enable 12V booster for sensor power
ADS1115 ADC reads 4-20mA sensor current
Convert current to fuel level percentage and height
Read battery voltage for health monitoring
Perform sensor fault detection (open/short circuit)

3. Cloud Transmission (5 seconds)

Establish MQTT connection with ThingsBoard broker
Publish telemetry data (fuel level, battery status, faults)
Wait for broker acknowledgement.
Disconnect MQTT session

4. Sleep Phase (15-60 minutes)

Disable all peripherals (ADC, I2C, SPI, UART)
Power down SIM7070 modem
Disable 12V booster
Enter STM32L0 shutdown mode (only RTC active)

This duty-cycled approach reduces average current from 17mA (active) to 1.29mA (15-min cycle) or 380µA (1-hour cycle).

Power Consumption Profile

Detailed measurements using precision current probes:

|--------------------|---------|------------|-------------|----------------|

| Modem Initialization | 60 sec | 17.71 mA | 840 mA | 3.93 J |

| Data Acquisition | 2.5 sec | 11.24 mA | 120 mA | 0.14 J |

| Data Transmission | 5 sec | 11.24 mA | 443 mA | 0.17 J |

| Sleep Mode | 15 min | 1.05 µA | 2.85 mA | 0.0035 J |

Complete 15-minute cycle: 1.29mA average, 4.60 J total energy

Battery Life Projections

With 38, 000mAh total capacity (2× 19, 000mAh ER34615H cells):

|----------------------|----------------|--------------|----------------------------|

Key Insight: Doubling the sleep interval from 15 to 30 minutes doubles battery life. The relationship is nearly linear because sleep current (1.05µA) is negligible compared to active current (17mA).

Schematics

Pin Configuration

EC-M12-BC-C6-C-A Key Pins:

STM32L0 Microcontroller: PA8 - Modem power key PA15 - 12V booster enable PB6 - I2C SCL (ADC) PB7 - I2C SDA (ADC) PA2 - UART TX (modem) PA3 - UART RX (modem)8-Pin Terminal Block: Pin 1 - 4-20mA analog input Pin 2 - Digital input 1 Pin 3 - Digital input 2 Pin 4 - GND Pin 5 - RS485 A Pin 6 - RS485 B Pin 7 - 12V output (sensor power) Pin 8 - GND

Hardware Connections

4-20mA Fuel Level Sensor:

Sensor Wire → EC-M12 Terminal

━━━━━━━━━━━━━━━━━━━━━━━━━━━━

Red (+) → Pin 7 (12V out)

Black (-) → Pin 8 (GND)

White (sig) → Pin 1 (4-20mA in)

ST-Link V2 Programming Interface:

ST-Link Pin → EC-M12 Header

━━━━━━━━━━━━━━━━━━━━━━━━━━━━

SWCLK → CLK

SWDIO → DIO

GND → GROUND

3.3V → (leave disconnected)

RST → RESET

Power Supply Architecture

The EC-M12-BC-C6-C-A incorporates sophisticated power management:

Battery Configuration:

2× ER34615H lithium thionyl chloride cells
Parallel connection: 3.6V @ 38, 000mAh
Pulse current capability: 600mA (2× 300mA per cell)
Operating temperature: -60°C to +85°C

Critical Design Elements:

1. 1000µF Tantalum Capacitor

Located near SIM7070 modem
Buffers 840mA current spikes during network registration
Prevents voltage sag below modem's 3.0V minimum

2. Boost Converter (12V Output)

Generates 12V for 4-20mA sensor power
Software-controlled enable (PA15)
Disabled during sleep to eliminate quiescent current
Max output current: 150mA

3. Low-Dropout Regulators

3.3V LDO for STM32L0 and peripherals
Ultra-low quiescent current (<5µA) for sleep mode
Thermal shutdown protection

Power Sequencing:

Wake → Enable LDO → Init STM32 → Power modem → Enable booster →

Read sensor → Transmit data → Disable booster → Power down modem → Sleep

Jumper Configuration

JP1 - Power Source Selection:

Position 1: Battery power (field deployment)
Position 2: USB power (programming and testing)

Important: Always use USB power (Position 2) during initial programming to preserve battery charge

Code

Complete Project Repository

GitHub Repository: NORVI EC-M12-BC-C6-C-A 4-20mA Sensor Application

Contains:

Complete Arduino sketch
Hardware configuration files
ThingsBoard dashboard templates
Battery life calculation spreadsheet
Power consumption analysis tools

Core Firmware Architecture

System Initialization

#include <STM32LowPower.h>
#include <SIM7070.h>
#include <Adafruit_ADS1X15.h>
#include <PubSubClient.h>
// Hardware pin definitions
#define MODEM_PWRKEY    PA8
#define BOOSTER_ENABLE  PA15
#define LED_STATUS      PC13
#define ADC_SDA         PB7
#define ADC_SCL         PB6
// Network configuration
#define APN             "iot.1nce.net"  // Adjust for your carrier
#define MQTT_SERVER     "thingsboard.cloud"
#define MQTT_PORT       1883
#define DEVICE_TOKEN    "YOUR_DEVICE_TOKEN"
// Timing configuration
#define SLEEP_MINUTES   15  // Adjust for desired battery life
// Global objects
SIM7070 modem;
Adafruit_ADS1115 ads;
WiFiClient netClient;
PubSubClient mqtt(netClient);
void setup() {
// Initialize serial for debugging
Serial.begin(115200);
// Configure GPIO pins
pinMode(MODEM_PWRKEY, OUTPUT);
pinMode(BOOSTER_ENABLE, OUTPUT);
pinMode(LED_STATUS, OUTPUT);
// Ensure peripherals start disabled
digitalWrite(BOOSTER_ENABLE, LOW);
digitalWrite(LED_STATUS, LOW);
// Initialize low-power framework
LowPower.begin();
// Initialize I2C ADC
Wire.begin();
if (!ads.begin(0x48)) {
Serial.println("ADS1115 not found!");
while(1);
}
// Configure ADC for 4-20mA measurement
// Gain: ±4.096V, Resolution: 0.125mV
ads.setGain(GAIN_ONE);
ads.setDataRate(RATE_ADS1115_128SPS);
// Initialize cellular modem
modem.begin();
// Connect to cellular network
connectNetwork();
// Setup MQTT connection
mqtt.setServer(MQTT_SERVER, MQTT_PORT);
mqtt.setCallback(mqttCallback);
}

Cellular Network Connection

bool connectNetwork() {
  Serial.println("Initializing modem...");
  
  // Power on modem
  digitalWrite(MODEM_PWRKEY, HIGH);
  delay(500);
  digitalWrite(MODEM_PWRKEY, LOW);
  delay(3000);  // Wait for modem boot
  
  // Wait for network registration
  int attempts = 0;
  while (!modem.isNetworkRegistered() && attempts < 60) {
    delay(1000);
    attempts++;
    Serial.print(".");
  }
  
  if (!modem.isNetworkRegistered()) {
    Serial.println("Network registration failed!");
    return false;
  }
  
  Serial.println("\nNetwork registered");
  
  // Activate GPRS/LTE data
  if (!modem.activateDataConnection(APN)) {
    Serial.println("Data activation failed!");
    return false;
  }
  
  Serial.println("Data connection active");
  return true;
}

4-20mA Sensor Reading

struct SensorReading {
  float current;        // mA
  float fuelPercent;    // 0-100%
  float fuelHeight;     // meters
  bool fault;           // true if sensor error
  String faultType;     // "open", "short", or "none"
};

SensorReading readFuelSensor() {
  SensorReading reading;
  
  // Enable 12V booster for sensor power
  digitalWrite(BOOSTER_ENABLE, HIGH);
  delay(500);  // Stabilization time
  
  // Read ADC (differential input for current sensing)
  int16_t adcValue = ads.readADC_Differential_0_1();
  
  // Convert ADC reading to voltage
  // ADS1115: 16-bit, ±4.096V range = 0.125mV/bit
  float voltage = adcValue * 0.000125;  // Volts
  
  // Convert voltage to current using sense resistor
  // 249Ω sense resistor (1% tolerance)
  reading.current = (voltage / 249.0) * 1000.0;  // mA
  
  // Detect sensor faults
  if (reading.current < 3.8) {
    reading.fault = true;
    reading.faultType = "open";
    reading.fuelPercent = -1;
    reading.fuelHeight = -1;
  } 
  else if (reading.current > 20.5) {
    reading.fault = true;
    reading.faultType = "short";
    reading.fuelPercent = -1;
    reading.fuelHeight = -1;
  } 
  else {
    reading.fault = false;
    reading.faultType = "none";
    
    // Convert 4-20mA to 0-100% fuel level
    // 4mA = 0%, 20mA = 100%
    reading.fuelPercent = ((reading.current - 4.0) / 16.0) * 100.0;
    
    // Calculate fuel height (assuming 2-meter tank)
    reading.fuelHeight = (reading.fuelPercent / 100.0) * 2.0;
    
    // Clamp values to valid range
    reading.fuelPercent = constrain(reading.fuelPercent, 0, 100);
    reading.fuelHeight = constrain(reading.fuelHeight, 0, 2.0);
  }
  
  // Disable booster to save power
  digitalWrite(BOOSTER_ENABLE, LOW);
  
  // Debug output
  Serial.print("Current: "); Serial.print(reading.current); Serial.println(" mA");
  Serial.print("Fuel: "); Serial.print(reading.fuelPercent); Serial.println(" %");
  Serial.print("Height: "); Serial.print(reading.fuelHeight); Serial.println(" m");
  
  return reading;
}

Battery Voltage Monitoring

float readBatteryVoltage() {
  // Battery voltage is connected to ADC channel 3 via 2:1 divider
  int16_t adcValue = ads.readADC_SingleEnded(3);
  
  // Convert to actual voltage
  float voltage = (adcValue * 0.000125) * 2.0;  // Account for divider
  
  Serial.print("Battery: "); Serial.print(voltage); Serial.println(" V");
  
  return voltage;
}

MQTT Data Publishing

bool publishTelemetry() {
  // Ensure MQTT connected
  if (!mqtt.connected()) {
    Serial.println("Connecting to MQTT...");
    if (!mqtt.connect("EC-M12-Device", DEVICE_TOKEN, NULL)) {
      Serial.println("MQTT connection failed!");
      return false;
    }
  }
  
  // Read sensors
  SensorReading fuel = readFuelSensor();
  float battery = readBatteryVoltage();
  
  // Construct JSON payload
  String payload = "{";
  payload += "\"fuel_level\":" + String(fuel.fuelPercent, 1) + ",";
  payload += "\"fuel_height\":" + String(fuel.fuelHeight, 2) + ",";
  payload += "\"sensor_current\":" + String(fuel.current, 2) + ",";
  payload += "\"battery_voltage\":" + String(battery, 2) + ",";
  payload += "\"sensor_fault\":" + String(fuel.fault ? "true" : "false") + ",";
  payload += "\"fault_type\":\"" + fuel.faultType + "\"";
  payload += "}";
  
  Serial.println("Publishing: " + payload);
  
  // Publish to ThingsBoard
  bool success = mqtt.publish("v1/devices/me/telemetry", payload.c_str());
  
  if (success) {
    Serial.println("Data published successfully");
    digitalWrite(LED_STATUS, HIGH);
    delay(100);
    digitalWrite(LED_STATUS, LOW);
  } else {
    Serial.println("Publish failed!");
  }
  
  // Allow time for acknowledgment
  mqtt.loop();
  delay(1000);
  
  return success;
}

void mqttCallback(char* topic, byte* payload, unsigned int length) {
  // Handle incoming MQTT messages (if needed for remote control)
  Serial.print("Message received [");
  Serial.print(topic);
  Serial.print("]: ");
  for (int i = 0; i < length; i++) {
    Serial.print((char)payload[i]);
  }
  Serial.println();
}

Ultra-Low-Power Sleep Mode

void enterDeepSleep(uint32_t minutes) {
  Serial.print("Entering sleep for ");
  Serial.print(minutes);
  Serial.println(" minutes...");
  delay(100);  // Flush serial
  
  // Disconnect MQTT
  mqtt.disconnect();
  
  // Power down modem
  modem.powerDown();
  delay(1000);
  
  // Disable all peripherals
  digitalWrite(BOOSTER_ENABLE, LOW);
  digitalWrite(LED_STATUS, LOW);
  
  // Close I2C, SPI, UART
  Wire.end();
  SPI.end();
  Serial.end();
  
  // Configure RTC wakeup timer
  uint32_t sleepMillis = minutes * 60 * 1000;
  
  // Enter STM32L0 shutdown mode
  // Current consumption: ~1µA
  // Only RTC remains active
  LowPower.shutdown(sleepMillis);
  
  // After wake, execution restarts from setup()
}

Main Loop

void loop() {
  // Ensure network connectivity
  if (!modem.isNetworkRegistered()) {
    Serial.println("Network lost, reconnecting...");
    connectNetwork();
  }
  
  // Publish sensor data to cloud
  publishTelemetry();
  
  // Enter deep sleep to conserve battery
  enterDeepSleep(SLEEP_MINUTES);
  
  // Note: After wake, system restarts from setup()
  // This loop will not execute again in same session
}

Build Instructions

Step 1: Initial Hardware Setup

1.1 Power Configuration

Locate jumper JP1 on EC-M12-BC-C6-C-A board
Set to Position 2 (USB power) for programming
Do not insert batteries yet

1.2 ST-Link Connection

Connect the ST-Link V2 programmer to EC-M12 programming header:

ST-Link → EC-M12━━━━━━━━━━━━━━━━━SWCLK → CLKSWDIO → DIOGND → GROUNDRST → RESET

Connect ST-Link to computer via USB
Connect EC-M12 to computer via USB (for serial monitoring)

1.3 Sensor Wiring

Connect 4-20mA fuel sensor to 8-pin terminal block:

Sensor → Terminal━━━━━━━━━━━━━━━━━Red → Pin 7 (12V)Black → Pin 8 (GND)White → Pin 1 (4-20mA input)

Step 2: Arduino IDE Configuration

2.1 Install STM32 Board Support

Open Arduino IDE
Go to File → Preferences
Add to "Additional Boards Manager URLs"
Open Tools → Board → Boards Manager
Search "STM32" and install "STM32 MCU-based boards" by STMicroelectronics

2.2 Install Required Libraries

Via Library Manager (Sketch → Include Library → Manage Libraries):

STM32LowPower (by STMicroelectronics)
Adafruit ADS1X15 (by Adafruit)
PubSubClient (by Nick O'Leary)

Manual Installation for SIM7070 Library:

Download from NORVI GitHub repository
Extract to Arduino/libraries folder
Restart Arduino IDE

2.3 Board Configuration

Tools → Board → STM32 Boards → Generic STM32L0 Series
Tools → Board Part Number → STM32L072RBTx
Tools → Upload Method → STM32CubeProgrammer (SWD)
Tools → USB Support → "CDC (generic 'Serial' supersede U(S)ART)"
Tools → U(S)ART Support → "Enabled (generic 'Serial')"
Tools → Optimize → "Smallest (-Os default)"

Step 3: ThingsBoard Cloud Setup

3.1 Create ThingsBoard Account

Navigate to https://thingsboard.cloud
Sign up for free tier account
Verify email and log in

3.2 Add Device

Go to Devices → Add Device (+)
Name: "Fuel Tank Monitor 01"
Device Profile: "default"
Click "Add"
Copy the Device Access Token (you'll need this)

3.3 Configure Dashboard

Create a new dashboard with widgets:

Fuel Level Gauge:

Widget type: Analog Gauges → Radial Gauge
Data key: fuel_level
Min: 0, Max: 100
Units: %
Color ranges: 0-20 (red), 20-50 (yellow), 50-100 (green)

Fuel Height Display:

Widget type: Cards → Simple Card
Data key: fuel_height
Units: meters
Decimals: 2

Battery Voltage Chart:

Widget type: Charts → Timeseries Line Chart
Data key: battery_voltage
Time window: Last 7 days
Y-axis: 2.5 - 4.0V

Sensor Fault Indicator:

Widget type: Alarm Widgets → Alarms Table
Filter: sensor_fault = true
Severity: Critical

Step 4: Firmware Configuration

4.1 Download Project Code

Download

4.2 Update Configuration

Edit the main sketch file:

// Network Configuration
#define APN "iot.1nce.net"  // Replace with your carrier APN

// ThingsBoard Configuration
#define MQTT_SERVER "thingsboard.cloud"
#define DEVICE_TOKEN "YOUR_DEVICE_ACCESS_TOKEN_HERE"  // Paste token from Step 3.2

// Sleep Configuration
#define SLEEP_MINUTES 15  // Adjust for desired battery life

APN Configuration Guide:

| Carrier | APN |

|---------------|----------------|

| 1NCE (Global IoT) | iot.1nce.net |

| Hologram | hologram |

| AT&T IoT | m2m.com.attz |

| Verizon IoT | vzwinternet |

| Vodafone IoT | iot.vodafone.com |

Step 5: Upload Firmware

5.1 Compile Sketch

Open sketch in Arduino IDE
Click Verify (✓) to check for errors
If compilation successful, click Upload (→)

STM32CubeProgrammer Method:

Open STM32CubeProgrammer
Connect via ST-Link (click "Connect")
Download → Browse for.bin file
Start Address: 0x08000000
Click "Start Programming"

Step 6: Initial Testing

6.1 Serial Monitor Verification

Open Arduino IDE Serial Monitor (Tools → Serial Monitor)
Set baud rate to 115200
Press reset button on EC-M12
You should see:

Initializing modem...Network registeredData connection activeCurrent: 12.35 mAFuel: 51.9 %Height: 1.04 mBattery: 3.58 VPublishing: {...}Data published successfullyEntering sleep for 15 minutes...

6.2 ThingsBoard Verification

Log in to ThingsBoard
Navigate to Devices → Fuel Tank Monitor 01
Check "Latest Telemetry" tab
You should see fresh data within 1-2 minutes
Open dashboard to view visualizations

Step 7: Field Deployment

7.1 Battery Installation

Disconnect USB and ST-Link
Move jumper JP1 to Position 1 (Battery power)
Insert two ER34615H batteries into holders
Verify polarity: + terminal to positive contact

7.2 Enclosure Installation

Mount EC-M12 in weatherproof enclosure (IP65 rated)
Route sensor cables through cable glands
Include desiccant pack to prevent condensation
Seal all openings with thread sealant

7.3 Antenna Placement

SIM7070 modem requires LTE antenna
Keep antenna vertical for best signal
Avoid metal obstacles within 10cm
For underground tanks, mount antenna above ground

7.4 Final System Check

Verify data appears on ThingsBoard within 2 minutes
Monitor battery voltage (should be 3.5-3.6V)
Check sensor_fault status (should be false)
Confirm fuel level readings match expected values

Testing & Calibration

Power Consumption Verification

Equipment Required:

Digital multimeter with mA/µA range
10Ω current sense resistor (optional for oscilloscope)
USB power supply (5V, 2A minimum)

Procedure:

1. Measure Sleep Current

Setup: Battery (+) → Multimeter (µA mode) → EC-M12 (+)
       EC-M12 (-) → Battery (-)

Expected: 1-2 µA average
          2-3 mA occasional spikes (RTC activity)

2. Measure Active Current

Setup: Same as above, but multimeter in mA mode

Expected during data transmission:
  - 15-20 mA baseline
  - 100-200 mA during MQTT publish
  - 400-800 mA during modem network registration

3. Capture Power Profile (Optional)

Using oscilloscope and current probe:

Probe: Current probe on battery positive rail
Timebase: 1 second/division
Trigger: Rising edge, 10mA threshold

You should see:
  - 60-second modem init phase (17mA avg, 800mA peaks)
  - 7-second data transmission phase (11mA avg)
  - Drop to near-zero during sleep

Sensor Calibration

The 4-20mA sensor outputs current proportional to fuel level:

Standard Calibration Points:

Fuel Level | Current | ADC Reading | Expected Output
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
0% (Empty) | 4.00 mA | ~12,900     | 0.0%
25%        | 8.00 mA | ~25,800     | 25.0%
50% (Half) | 12.00 mA| ~38,700     | 50.0%
75%        | 16.00 mA| ~51,600     | 75.0%
100% (Full)| 20.00 mA| ~64,500     | 100.0%

Calibration Procedure:

1. Empty Tank Calibration

Drain tank to known empty state
Record ADC reading
Should be approximately 12, 900 (4mA)
If different, adjust offset in code:

float offsetCurrent = 4.0; // Adjust to match actual empty readingreading.fuelPercent = ((reading.current - offsetCurrent) / 16.0) * 100.0;

2. Full Tank Calibration

Fill tank to known full state
Record ADC reading
Should be approximately 64, 500 (20mA)
If different, adjust span in code:

float spanCurrent = 16.0; // Adjust (full - empty) current rangereading.fuelPercent = ((reading.current - 4.0) / spanCurrent) * 100.0;

3. Verification

Test at multiple intermediate levels (25%, 50%, 75%)
Compare sensor reading to manual dipstick measurement
Accuracy should be within ±2% of full scale

Troubleshooting Calibration Issues:

 | Problem           | Possible Cause           | Solution                   |
|------------------|------------------------|----------------------------|
| Reading stuck at 0%   | Open circuit           | Check sensor wiring        |
| Reading stuck at 100% | Short circuit          | Check for damaged cable    |
| Erratic readings      | Poor ground connection | Verify GND continuity      |
| Readings drift        | Temperature effect     | Use sensor with temp compensation |
| Offset error          | Sense resistor tolerance | Adjust offset in code     |

Network Performance Testing

Signal Strength Check:

// Add to setup() function for testing
void printNetworkInfo() {
int rssi = modem.getSignalQuality();
String networkType = modem.getNetworkType();
Serial.print("Signal Strength (RSSI): ");
Serial.println(rssi);
Serial.print("Network Type: ");
Serial.println(networkType);
if (rssi < 10) {
Serial.println("WARNING: Weak signal - battery life will be reduced");
}
}

Expected Signal Levels:

RSSI Value | Signal Quality | Impact on Battery Life
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
0-9        | Poor          | +40% power consumption
10-14      | Fair          | +20% power consumption
15-19      | Good          | Normal operation
20-31      | Excellent     | Optimal power efficiency

Network Registration Time Test:

Good signal: 30-45 seconds
Fair signal: 45-75 seconds
Poor signal: 75-120 seconds (or failure)

Poor signal areas may require repositioning the antenna or choosing a different carrier.

Optimization Strategies

Battery Life Optimization

1. Adjust Transmission Interval

Match your interval to application requirements:

// High-priority monitoring (fuel deliveries expected)
#define SLEEP_MINUTES 15  // 3.36 years battery life

// Standard monitoring (stable consumption)
#define SLEEP_MINUTES 30  // 6.5 years battery life

// Low-priority monitoring (backup tank)
#define SLEEP_MINUTES 60  // 11.4 years battery life

2. Implement Adaptive Intervals

Only transmit when fuel level changes significantly:

float lastFuelLevel = -1;
#define CHANGE_THRESHOLD 5.0  // 5% change triggers transmission

void loop() {
  SensorReading fuel = readFuelSensor();
  
  // Only transmit if significant change
  if (lastFuelLevel < 0 || abs(fuel.fuelPercent - lastFuelLevel) > CHANGE_THRESHOLD) {
    publishTelemetry();
    lastFuelLevel = fuel.fuelPercent;
  } else {
    Serial.println("No significant change, skipping transmission");
  }
  
  enterDeepSleep(SLEEP_MINUTES);
}

This can extend battery life by 2-3× in stable tank applications.

3. Enable Power-Saving Mode

Configure SIM7070 for maximum efficiency:

void setupModemPowerSaving() {
// Enable PSM (Power Saving Mode)
modem.setPowerSavingMode(true);
// Enable eDRX (Extended Discontinuous Reception)
modem.seteDRX(true, 5); // 5-second cycle
// Reduce TX power if signal is strong
int rssi = modem.getSignalQuality();
if (rssi > 20) {
modem.setTxPower(5); // Minimum power
}
}

4. Optimize MQTT Payload

Reduce data size to minimize transmission time:

// Verbose payload: 156 bytes
{"fuel_level_percentage":87.5,"fuel_height_meters":1.75,"battery_voltage":3.58,"sensor_fault_status":false}
// Optimized payload: 45 bytes
{"fl":87.5,"fh":1.75,"bv":3.58,"sf":0}
// Savings: 71% reduction in data size// Verbose payload: 156 bytes
{"fuel_level_percentage":87.5,"fuel_height_meters":1.75,"battery_voltage":3.58,"sensor_fault_status":false}
// Optimized payload: 45 bytes
{"fl":87.5,"fh":1.75,"bv":3.58,"sf":0}

Smaller payloads = faster transmission = less power consumption.

Environmental Considerations

Temperature Compensation

Lithium batteries lose capacity in extreme cold:

float getTemperatureDerating() {
  float temp = readTemperature();  // Add temperature sensor
  
  if (temp > 20) return 1.0;       // Full capacity
  if (temp > 0) return 0.95;       // 5% reduction
  if (temp > -10) return 0.85;     // 15% reduction
  if (temp > -20) return 0.75;     // 25% reduction
  return 0.60;                     // 40% reduction below -20°C
}

// Adjust battery life estimate
float effectiveCapacity = 38000 * getTemperatureDerating();

Weatherproofing Best Practices

1. Enclosure Selection

Minimum IP65 rating (dust-tight, water jet resistant)
Polycarbonate or ABS material
UV-resistant for outdoor installations
Vented to prevent condensation

2. Cable Entry

Use PG glands with O-rings
Apply thread sealant (not Teflon tape)
Route cables downward to prevent water ingress
Leave drip loops before entry points

3. Antenna Mounting

External antenna for metal enclosures
Vertical orientation for omnidirectional coverage
Lightning arrestor for exposed installations
Keep clear of metal structures (>10cm clearance)

Troubleshooting Guide

Common Issues and Solutions

Problem: Device not connecting to cellular network

Symptoms:

Modem initialization exceeds 60 seconds
Serial output shows "Network registration failed"
No data appears on ThingsBoard

Solutions:

Verify SIM card is activated and has data plan
Check APN settings match your carrier
Test cellular coverage with smartphone at same location
Try different network mode (NB-IoT vs LTE-M):

modem.setPreferredMode(NBIOT_MODE); // or LTE_M_MODE

Increase connection timeout to 120 seconds
Check antenna connection (SMA connector tight)

Problem: Current consumption higher than expected

Symptoms:

Sleep current >10µA
Battery drains in months instead of years
Device feels warm during sleep

Diagnostic Steps:

// Add debug code to verify peripheral shutdown
void verifyPowerDown() {
  Serial.println("Checking power-down sequence:");
  
  Serial.print("Booster EN: ");
  Serial.println(digitalRead(BOOSTER_ENABLE) == LOW ? "OFF" : "ON");
  
  Serial.print("Modem power: ");
  Serial.println(modem.isPoweredOn() ? "ON" : "OFF");
  
  // Should all be OFF/false before sleep
}

Common Causes:

Booster not disabled - Verify digitalWrite(BOOSTER_ENABLE, LOW) called
Modem still powered - Call modem.powerDown() before sleep
Peripherals still active - Call Wire.end(), Serial.end(), etc.
Poor solder joint - Check for current leakage on PCB
Incorrect sleep mode - Use LowPower.shutdown() not LowPower.sleep()

Measurement Procedure:

Remove EC-M12 from enclosure
Disconnect sensor (to isolate device current)
Connect ammeter in series with battery positive
Wait through complete cycle (wake → transmit → sleep)
Record minimum current during sleep phase
Should be <2µA (excluding brief RTC spikes)

Problem: Sensor readings incorrect or erratic

Symptoms:

Fuel level stuck at 0% or 100%
Readings fluctuate wildly
"Sensor fault" error constantly triggered

Diagnosis Table:

| Symptom        | Reading        | Likely Cause        | Solution                     |
|----------------|---------------|-------------------|------------------------------|
| Stuck at 0%    | <4mA           | Open circuit       | Check wiring continuity      |
| Stuck at 100%  | >20mA          | Short circuit      | Inspect cable for damage     |
| Erratic ±10%   | Varying        | Loose connection   | Tighten terminal screws      |
| Slow drift     | Gradual change | Temperature effect | Add temp compensation        |
| Noisy readings | ±2% fluctuation| EMI interference  | Add ferrite bead to cable    |

Verification Steps:

Measure sensor current with multimeter in series
Should be 4-20mA proportional to level
Disconnect sensor and measure ADC voltage directly
Should be 0.996V to 4.98V (for 249Ω sense resistor)
Check sense resistor value (should be 249Ω ±1%)

Problem: MQTT connection fails

Symptoms:

Serial output: "MQTT connection failed!"
Modem connected to network but no data on ThingsBoard
Publishing returns false

Checklist:

[ ] ThingsBoard device access token correct in code
[ ] MQTT server address correct ("thingsboard.cloud" or custom)
[ ] Port 1883 open (check firewall if using VPN)
[ ] Device exists in ThingsBoard and is active
[ ] Network time synchronized (some MQTT servers require accurate time)

Debug MQTT Connection:

void debugMQTT() {
  Serial.println("MQTT Debug:");
  Serial.print("Server: ");
  Serial.println(MQTT_SERVER);
  Serial.print("Port: ");
  Serial.println(MQTT_PORT);
  Serial.print("Client ID: EC-M12-Device");
  Serial.print("Token: ");
  Serial.println(DEVICE_TOKEN);
  
  int result = mqtt.connect("EC-M12-Device", DEVICE_TOKEN, NULL);
  Serial.print("Connection result: ");
  Serial.println(result);  // 0=success, negative=error code
}

MQTT Error Codes:

0: Success
-1: Connection refused - incorrect protocol version
-2: Connection refused - identifier rejected
-3: Connection refused - server unavailable
-4: Connection refused - bad username/password (token)
-5: Connection refused - not authorized

Problem: Battery voltage dropping faster than expected

Symptoms:

Battery reads <3.4V after 6 months
Voltage drops >0.1V per month
Device resets unexpectedly

Diagnostic:

void batteryHealthCheck() {
  float voltage = readBatteryVoltage();
  
  Serial.print("Battery voltage: ");
  Serial.print(voltage);
  Serial.println(" V");
  
  if (voltage < 3.0) {
    Serial.println("CRITICAL: Replace batteries immediately!");
  } else if (voltage < 3.2) {
    Serial.println("WARNING: Low battery - replace soon");
  } else if (voltage < 3.4) {
    Serial.println("NOTICE: Battery at 50% capacity");
  } else {
    Serial.println("Battery healthy");
  }
}

Causes:

High transmission frequency - Reduce from 15min to 60min intervals
Poor cellular signal - Modem uses more power in weak signal areas
Sensor draws high current - Verify 4-20mA sensor not exceeding 20mA
Battery manufacturing defect - Replace with fresh ER34615H cells
Temperature extremes - Cold weather reduces effective capacity

Battery Life Calculation Tool: Use actual measured currents to predict remaining life:

void estimateRemainingLife() {
  float batteryVoltage = readBatteryVoltage();
  float remainingCapacity = 38000 * ((batteryVoltage - 2.8) / (3.6 - 2.8));
  float avgCurrent = 1.29;  // mA for 15-min interval
  
  float remainingHours = remainingCapacity / avgCurrent;
  float remainingDays = remainingHours / 24;
  
  Serial.print("Estimated remaining: ");
  Serial.print(remainingDays);
  Serial.println(" days");
}

Real-World Deployment Case Studies

Case Study 1: Remote Construction Site Fuel Monitoring

Application: 15 diesel fuel tanks across construction sites in rural areas

Configuration:

Transmission interval: 30 minutes
4-20mA float-style level sensors
NB-IoT connectivity (better penetration in remote areas)
Weatherproof NEMA 4X enclosures

Results After 18 Months:

Zero maintenance visits required
Battery voltage: 3.42V (95% original capacity)
Projected total life: 6+ years
Prevented 3 fuel theft incidents (real-time alerts)
Saved 45 manual inspection trips ($6, 750 labor cost)

Lessons Learned:

Solar panel not needed for 6+ year target
NB-IoT more reliable than LTE-M in rural areas
30-minute interval provides good balance of data freshness and battery life
Antenna placement critical—above-ground mounting required for underground tanks

Case Study 2: Industrial Chemical Storage Compliance

Application: Hazardous chemical level monitoring for regulatory compliance

Configuration:

Transmission interval: 15 minutes (regulatory requirement)
Explosion-proof enclosure (Class I, Div 1)
Redundant 4-20mA pressure transducers
LTE-M connectivity
SMS alerts for fault conditions

Results After 12 Months:

100% uptime for compliance reporting
Battery voltage: 3.48V
Projected total life: 3.5 years
Avoided 2 overflow incidents ($40, 000+ cleanup costs)
Passed safety audit with continuous monitoring evidence

Lessons Learned:

Fault detection critical for safety applications
Redundant sensors worth the added complexity
More frequent transmission (15 min) still achieves multi-year life
Explosion-proof enclosure adds significant cost but required for hazmat

Case Study 3: Agricultural Water Tank Network

Application: 50 livestock water tanks across 5, 000-acre ranch

Configuration:

Transmission interval: 2 hours (water consumption is gradual)
Ultrasonic level sensors (non-contact, easier installation)
Mix of NB-IoT and LTE-M based on coverage
Solar panel supplement for indefinite operation

Results After 24 Months:

48 of 50 units still operational (2 lightning strikes)
Battery voltage: 3.55V (solar keeps batteries topped off)
Detected 7 tank leaks before animals were affected
Ranch hands save 15 hours/week on manual tank checks

Lessons Learned:

2-hour interval more than sufficient for slow-changing levels
Solar panel + battery = indefinite operation for accessible locations
Lightning protection essential for open-field installations
LoRaWAN considered but cellular coverage adequate

Advanced Features & Extensions

Over-the-Air (OTA) Firmware Updates

While not implemented in base firmware, OTA updates are possible:

// Conceptual OTA implementation
void checkForUpdates() {
  if (mqtt.connected()) {
    // Subscribe to firmware update topic
    mqtt.subscribe("v1/devices/me/firmware");
    
    // Download new firmware to flash memory
    // Verify checksum
    // Write to bootloader area
    // Set flag to trigger update on next reset
  }
}

Considerations:

Requires bootloader support (STM32 built-in bootloader compatible)
Large downloads consume significant battery (plan for battery life impact)
Failed updates can brick device (implement rollback mechanism)
Test thoroughly before deploying to production units

Multi-Sensor Support

EC-M12-BC-C6-C-A has multiple input channels:

// Read multiple tanks from one device
struct MultiTankReading {
  float tank1_level;
  float tank2_level;
  float tank3_level;
  float tank4_level;
};

MultiTankReading readAllTanks() {
  MultiTankReading tanks;
  
  // ADC has 4 single-ended inputs
  tanks.tank1_level = readFuelLevel(0);  // ADC channel 0
  tanks.tank2_level = readFuelLevel(1);  // ADC channel 1
  tanks.tank3_level = readFuelLevel(2);  // ADC channel 2
  tanks.tank4_level = readFuelLevel(3);  // ADC channel 3 (or battery)
  
  return tanks;
}

Predictive Maintenance

Implement trend analysis for proactive alerts:

// Store last 7 days of readings
float fuelHistory[168];  // 7 days * 24 hours
int historyIndex = 0;

void analyzeTrend() {
  // Calculate consumption rate
  float dailyChange = (fuelHistory[0] - fuelHistory[24]) / 24.0;
  
  // Predict days until empty
  float daysUntilEmpty = fuelHistory[0] / abs(dailyChange);
  
  if (daysUntilEmpty < 7) {
    // Send alert for refill
    String alert = "{\"alert\":\"Fuel low - refill needed in " + 
                   String(daysUntilEmpty) + " days\"}";
    mqtt.publish("v1/devices/me/attributes", alert.c_str());
  }
}

Integration with Enterprise Systems

Connect to existing infrastructure:

REST API Integration:

// HTTP POST instead of MQTT
void publishViaHTTP() {
  HTTPClient http;
  http.begin("https://your-erp-system.com/api/fuel-levels");
  http.addHeader("Content-Type", "application/json");
  http.addHeader("Authorization", "Bearer YOUR_TOKEN");
  
  String payload = "{\"tank_id\":\"TANK-001\",\"level\":" + String(fuelLevel) + "}";
  int httpCode = http.POST(payload);
  
  http.end();
}

Database Direct Insert:

// Some cloud MQTT brokers can auto-insert to PostgreSQL/MySQL
// Configure ThingsBoard rule chain to forward data to your database

Conclusion

The Norvi EC-M12-BC-C6-C-A demonstrates that ultra-low-power industrial IoT is not just theoretically possible—it's practical, economical, and reliable. By achieving 11.4 years of battery life with hourly data transmission, this platform solves the fundamental challenge of remote monitoring: minimizing total cost of ownership.

Key Takeaways:

Duty cycling is the dominant factor in battery life
STM32L0's shutdown mode (1µA) enables multi-year operation
Cellular connectivity is viable for battery-powered applications
Proper power management requires holistic system design
Real-world deployments validate theoretical calculations

Whether you're monitoring fuel tanks, water levels, environmental conditions, or industrial processes, the principles demonstrated here apply across countless applications. The combination of efficient hardware, optimized firmware, and intelligent duty cycling creates a deploy-and-forget solution that delivers ROI through eliminated maintenance costs and reliable 24/7 monitoring.

Ready to deploy your own remote monitoring system? The complete source code, documentation, and technical support are available through Norvi's official channels.

License & Credits

Hardware: NORVI EC-M12-BC-C6-C-A Industrial IoT Controller
Firmware: Open-source under MIT License
Cloud Platform: ThingsBoard IoT Platform

Kavinda Mahesh

2 projects • 0 followers

IoT enthusiast at Norvi.io | ESP32, Arduino & industrial IoT | Building smarter automation & connected solutions

Embed the widget on your own site

Ultra-Low-Power Industrial Data Logger with 11-Year Battery

Ultra-Low-Power Industrial Data Logger with 11-Year Battery

Things used in this project

Hardware components

Software apps and online services

Hand tools and fabrication machines

Story

Introduction

Application Overview

Why This Matters

How It Works

System Architecture

Power Consumption Profile

Battery Life Projections

Schematics

Pin Configuration

EC-M12-BC-C6-C-A Key Pins:

Critical Design Elements:

Power Sequencing:

Jumper Configuration

Code

Complete Project Repository

Core Firmware Architecture

Build Instructions

Testing & Calibration

Troubleshooting Calibration Issues:

Network Performance Testing

Optimization Strategies

Environmental Considerations

Weatherproofing Best Practices

Troubleshooting Guide

Real-World Deployment Case Studies

Advanced Features & Extensions

License & Credits

Credits

Kavinda Mahesh

Comments

Related channels and tags