PowerPi

Motivation

While developing low-powered MCU systems (battery-powered), I wanted to figure out how much power my system is using and at what times.
I came across this video and in it, the Nordic Power Profiler Kit II.
Having a RPi and INA219 at hand, I decided to build my own power profiler.

Goals

1. Measure and show current draw, voltage stability and overall power usage
2. Control, visualize and analyze easily from a web browser

Features

• "Ultra Fast" measurements (uS) (e.g. to show power spikes in MCUs)
• "Slow" measurements (seconds, minutes) for long-term power system analysis (e.g. battery charging and solar systems)
• Save measurements to CSV
• Integration with CC/CV source to make it a more powerful lab tool

TO-DOs

• Move to better visualization platform
• Add script to configure as a boot service using systemd
• Clean the code
• Create short overview video
• Compare to professional equipment
• Better exceptions handling and show problems on dashboard

MCU (ESP32) Power Monitoring Example

System Overview

Hardware

1. Raspberry Pi 3B+ (plus has WiFi built-in) connected to INA219 using I2C
2. INA219 in series (high side) between power source and load
   1. Up to 26VDC
   2. Up to 3.2A
3. Common ground between measured circuit and PowerPi

Software

1. Based on pi-ina219 library
2. Uses Panel and Holoviews for web server and visualization
   1. This probably not the best way, but it's the only tool I feel comfortable with for now

Getting Started

Installation

1. Burn a Raspbian image on a MicroSD card, connect to a WiFi and get the IP address

2. Install pigpio - used for get_current_tick() to get micro-seconds precision

   wget https://github.com/joan2937/pigpio/archive/master.zip
   unzip master.zip
   cd pigpio-master
   make
   sudo make install

3. Install pip (to download python libraries) -

   sudo apt install python-pip

4. Download Python libraries

   pip install holoviz panel holoviews pi-ina219

5. Start pigpio daemon (every reboot)

   sudo pigpiod

6. Clone repo

   git clone https://github.com/saharco540/PowerPi.git

7. Start server

   panel serve panel.py --allow-websocket-origin=* --port=8080

8. Locate your RPi IP address and open a browser to http://IP:8080 or if hostname is powerpi - http://powerpi.local:8080

Wiring

Please read further here

Usage

Tabs for different modes

1. Data Logger - Start data logging to CSV for long-term measurements and later analysis
2. Live View - Live visualization of current and voltage
3. Data Explorer - Visualize and analyze data from CSV files

Extras

3D Printed box available (3d-box.stl)