This Python library supports the INA219 voltage, current and power monitor sensor from Texas Instruments on both Python 2 and 3. The intent of the library is to make it easy to use the quite complex functionality of this sensor.
The library currently only supports continuous reads of voltage and power, but not triggered reads.
The library supports the detection of overflow in the current/power calculations which results in meaningless values for these readings.
The low power mode of the INA219 is supported, so if only occasional reads are being made in a battery based system, current consumption can be minimised.
The library has been tested with the Adafruit INA219 Breakout.
This library and its dependency can be installed from PyPI by executing:
pip3 install pi-ina219
This library provides a generic I2C driver interface which allows the usage
with different I2C driver libraries, such as smbus
,
smbus2 or
Adafruit GPIO library.
See example.py
for implementations using mentioned libraries.
Those three I2C driver libraries are supported by the Raspberry Pi models, but there may be others. Remember to enable the I2C bus under the Advanced Options of raspi-config.
The pi-ina219
library can be installed with support of a specific driver
libraries:
pip3 install pi-ina219[smbus]
pip3 install pi-ina219[smbus2]
pip3 install pi-ina219[Adafruit]
The address of the sensor unless otherwise specified is the default of 0x40.
Note that the bus voltage is that on the load side of the shunt resistor, if you want the voltage on the supply side then you should add the bus voltage and shunt voltage together, or use the supply_voltage() function.
In most cases this will be determined automatically, however if this fails you will see the exception:
Could not determine default I2C bus for platform
In this case just set the bus number in the INA219 constructor, for example:
ina = INA219(SHUNT_OHMS, busnum=1)
This is known to be required with Raspberry Pi 4 and the 'Bullseye' (October 2021) Raspberry Pi OS.
This mode is great for getting started, as it will provide valid readings until the device current capability is exceeded for the value of the shunt resistor connected (3.2A for 0.1Ω shunt resistor). It does this by automatically adjusting the gain as required until the maximum is reached, when a DeviceRangeError exception is thrown to avoid invalid readings being taken.
The downside of this approach is reduced current and power resolution.
#!/usr/bin/env python
from ina219 import INA219
from ina219 import DeviceRangeError
SHUNT_OHMS = 0.1
def read():
ina = INA219(SHUNT_OHMS)
ina.configure()
print("Bus Voltage: %.3f V" % ina.voltage())
try:
print("Bus Current: %.3f mA" % ina.current())
print("Power: %.3f mW" % ina.power())
print("Shunt voltage: %.3f mV" % ina.shunt_voltage())
except DeviceRangeError as e:
# Current out of device range with specified shunt resistor
print(e)
if __name__ == "__main__":
read()
In this mode by understanding the maximum current expected in your system and specifying this in the script you can achieve the best possible current and power resolution. The library will calculate the best gain to achieve the highest resolution based on the maximum expected current.
In this mode if the current exceeds the maximum specified, the gain will be automatically increased, so a valid reading will still result, but at a lower resolution.
As above when the maximum gain is reached, an exception is thrown to avoid invalid readings being taken.
#!/usr/bin/env python
from ina219 import INA219
from ina219 import DeviceRangeError
SHUNT_OHMS = 0.1
MAX_EXPECTED_AMPS = 0.2
def read():
ina = INA219(SHUNT_OHMS, MAX_EXPECTED_AMPS)
ina.configure(ina.RANGE_16V)
print("Bus Voltage: %.3f V" % ina.voltage())
try:
print("Bus Current: %.3f mA" % ina.current())
print("Power: %.3f mW" % ina.power())
print("Shunt voltage: %.3f mV" % ina.shunt_voltage())
except DeviceRangeError as e:
# Current out of device range with specified shunt resistor
print(e)
if __name__ == "__main__":
read()
In this mode by understanding the maximum current expected in your system and specifying this and the gain in the script you can always achieve the best possible current and power resolution, at the price of missing current and power values if a current overflow occurs.
#!/usr/bin/env python
from ina219 import INA219
from ina219 import DeviceRangeError
SHUNT_OHMS = 0.1
MAX_EXPECTED_AMPS = 0.2
def read():
ina = INA219(SHUNT_OHMS, MAX_EXPECTED_AMPS)
ina.configure(ina.RANGE_16V, ina.GAIN_1_40MV)
print("Bus Voltage: %.3f V" % ina.voltage())
try:
print("Bus Current: %.3f mA" % ina.current())
print("Power: %.3f mW" % ina.power())
print("Shunt voltage: %.3f mV" % ina.shunt_voltage())
except DeviceRangeError as e:
print("Current overflow")
if __name__ == "__main__":
read()
The sensor address may be altered as follows:
ina = INA219(SHUNT_OHMS, MAX_EXPECTED_AMPS, address=0x41)
The sensor may be put in low power mode between reads as follows:
ina.configure(ina.RANGE_16V)
while True:
print("Voltage : %.3f V" % ina.voltage())
ina.sleep()
time.sleep(60)
ina.wake()
Note that if you do not wake the device after sleeping, the value returned from a read will be the previous value taken before sleeping.
INA219()
constructs the class. The arguments, are:- shunt_ohms: The value of the shunt resistor in Ohms (mandatory).
- max_expected_amps: The maximum expected current in Amps (optional).
- busnum: The I2C bus number for the device platform, defaults to auto detects 0 or 1 for Raspberry Pi or Beaglebone Black (optional).
- address: The I2C address of the INA219, defaults to 0x40 (optional).
- log_level: Set to logging.INFO to see the detailed calibration calculations and logging.DEBUG to see register operations (optional).
configure()
configures and calibrates how the INA219 will take measurements. The arguments, which are all optional, are:- voltage_range: The full scale voltage range, this is either 16V or 32V,
represented by one of the following constants (optional).
- RANGE_16V: Range zero to 16 volts
- RANGE_32V: Range zero to 32 volts (default). Device only supports up to 26V.
- gain: The gain, which controls the maximum range of the shunt voltage,
represented by one of the following constants (optional).
- GAIN_1_40MV: Maximum shunt voltage 40mV
- GAIN_2_80MV: Maximum shunt voltage 80mV
- GAIN_4_160MV: Maximum shunt voltage 160mV
- GAIN_8_320MV: Maximum shunt voltage 320mV
- GAIN_AUTO: Automatically calculate the gain (default)
- bus_adc: The bus ADC resolution (9, 10, 11, or 12-bit), or
set the number of samples used when averaging results, represented by
one of the following constants (optional).
- ADC_9BIT: 9 bit, conversion time 84us.
- ADC_10BIT: 10 bit, conversion time 148us.
- ADC_11BIT: 11 bit, conversion time 276us.
- ADC_12BIT: 12 bit, conversion time 532us (default).
- ADC_2SAMP: 2 samples at 12 bit, conversion time 1.06ms.
- ADC_4SAMP: 4 samples at 12 bit, conversion time 2.13ms.
- ADC_8SAMP: 8 samples at 12 bit, conversion time 4.26ms.
- ADC_16SAMP: 16 samples at 12 bit, conversion time 8.51ms
- ADC_32SAMP: 32 samples at 12 bit, conversion time 17.02ms.
- ADC_64SAMP: 64 samples at 12 bit, conversion time 34.05ms.
- ADC_128SAMP: 128 samples at 12 bit, conversion time 68.10ms.
- shunt_adc: The shunt ADC resolution (9, 10, 11, or 12-bit), or
set the number of samples used when averaging results, represented by
one of the following constants (optional).
- ADC_9BIT: 9 bit, conversion time 84us.
- ADC_10BIT: 10 bit, conversion time 148us.
- ADC_11BIT: 11 bit, conversion time 276us.
- ADC_12BIT: 12 bit, conversion time 532us (default).
- ADC_2SAMP: 2 samples at 12 bit, conversion time 1.06ms.
- ADC_4SAMP: 4 samples at 12 bit, conversion time 2.13ms.
- ADC_8SAMP: 8 samples at 12 bit, conversion time 4.26ms.
- ADC_16SAMP: 16 samples at 12 bit, conversion time 8.51ms
- ADC_32SAMP: 32 samples at 12 bit, conversion time 17.02ms.
- ADC_64SAMP: 64 samples at 12 bit, conversion time 34.05ms.
- ADC_128SAMP: 128 samples at 12 bit, conversion time 68.10ms.
- voltage_range: The full scale voltage range, this is either 16V or 32V,
represented by one of the following constants (optional).
voltage()
Returns the bus voltage in volts (V).supply_voltage()
Returns the bus supply voltage in volts (V). This is the sum of the bus voltage and shunt voltage. A DeviceRangeError exception is thrown if current overflow occurs.current()
Returns the bus current in milliamps (mA). A DeviceRangeError exception is thrown if current overflow occurs.power()
Returns the bus power consumption in milliwatts (mW). A DeviceRangeError exception is thrown if current overflow occurs.shunt_voltage()
Returns the shunt voltage in millivolts (mV). A DeviceRangeError exception is thrown if current overflow occurs.current_overflow()
Returns 'True' if an overflow has occured. Alternatively handle the DeviceRangeError exception as shown in the examples above.sleep()
Put the INA219 into power down mode.wake()
Wake the INA219 from power down mode.reset()
Reset the INA219 to its default configuration.is_conversion_ready()
check if conversion was done before reading the next measurement results.
On a Raspberry Pi 2 Model B running Raspbian Jesse and reading a 12-bit voltage in a loop, a read occurred approximately every 10 milliseconds.
On a Raspberry Pi 4 running Raspbian Buster a read occurred approximately every 570 microseconds.
To understand the calibration calculation results and automatic gain increases, informational output can be enabled with:
ina = INA219(SHUNT_OHMS, log_level=logging.INFO)
Detailed logging of device register operations can be enabled with:
ina = INA219(SHUNT_OHMS, log_level=logging.DEBUG)
Install development dependencies first (recommended to use virtual environments). This includes package dependencies, testing dependencies and static analysis dependencies.
pip3 install -r requirements-dev.txt
python3 -m unittest discover -s tests -p 'test_*.py'
A single unit test class may be run as follows:
python3 -m unittest tests.test_configuration.TestConfiguration
Code coverage metrics may be generated and viewed with:
coverage run --branch --source=ina219 -m unittest discover -s tests -p 'test_*.py'
coverage report -m
coverage xml
This library adheres to the PEP8 standard and follows the idiomatic style described in the book Writing Idiomatic Python by Jeff Knupp.
To perform local linting and type checks, run:
flake8 .
mypy -p ina219 --strict