Raspberry Pi INA219 Power Measurement Tutorial 

Thanks for the positive feedback and the great suggestions!

I haven’t used the ACS712 in a project but you’re right that you would need an ADC like the MCP3008. I have a video on ADC’s: ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-wJgyszOSoQU.html You might also want to ask Allegro Micro and Texas Instruments tech support for recommendations because there are many new current sensors with I2C interfaces: www.allegromicro.com & www.ti.com

Yes, you could use a Python interface such as Tkinter.

Great suggestion!

I made a stepper motor tutorial: ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-LUbhPKBL_IU.html

Oh thank you very much, I have to check this!

You can use Python logging although that could have adverse effects on your SSD card depending on frequency. I usually log to a Syslog server running on my home NAS or use an online service such as Adafruit IO.

A common ground is mandatory to measure voltage because it is a relative value. If you only want to measure current you might be able to use an INA219 without a common ground but there could be some accuracy issues.

Looks like a great chip with a substantial range improvement. The set up for the INA260 would be very similar to the INA219. The main difference would be software. Instead of the ChrisB2 library, you would probably want to use Adafruit CircuitPython which will run on the Raspberry Pi if you use their Blinka library. I have a tutorial that discusses Blinka and CircuitPython: www.rototron.info/circuitpython-nrf52840-pi-tutorial/

rdagger68 hi but this is precisely my problem the Adafruit INA260 installation guide is very vague when it comes to the first part of installing circuit python. learn.adafruit.com/adafruit-ina260-current-voltage-power-sensor-breakout/python-circuitpython it’s talking about installing circuit python and takes u to another guide then it asks you to install some files at some place at the pi etc... it’s extremely confusing. So a video step by step guide would be great.

Sorry, it would probably be a long time before I could make your request. Installing CircuitPython is very easy and I do demonstrate the installation process in the link above at 8:37 in the video: ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-bserx54ZJPM.html I think the Adafruit tutorial is confusing because it is primarily targeting their boards and not the Raspberry Pi. If you follow the instruction in my video to install Blinka, then theoretically the only other install step should be to pip install the INA260 driver: sudo pip3 install adafruit-circuitpython-ina260

I haven’t tried but you could probably build a voltage controlled current source circuit using an opamp and a MOSFET and then control it using PWM. There are probably many adjustable current source chips that could be controlled using PWM or a digital pot. I recommend you search electronics.stackexchange.com for “adjustable current source mA”.

It should be easy connecting 4 INA219 boards. You just need to set the I2C slave address on each board so it's unique. You can have up to 16 boards because there are 16 possible slave addresses. The boards do not require much power so the 3.3V supply pin should be OK.

@@rdagger Thank you so much! I was not expecting such a quick response! Do you have a patreon?

@@shrekogreface7411 I don't have a patreon. Thanks for inquiring.

The INA219 can only measure up to +26VDC.

The logging library is part of MicroPython Lib: github.com/micropython/micropython-lib/tree/master/python-stdlib/logging

All batteries have a discharge curve. This non-linear graph shows voltage as a function of capacity. Generally speaking, rechargeable batteries tend to have a consistent curve over their lifespans. Therefore, by monitoring voltage you can get an estimate of battery capacity. You can also track current consumption and generate a running total to calculate an amp-hours value. By combining the two you can get a better estimate. If you are working with LiPo batteries, you can use a low cost MAX17043 I²C battery fuel gauge module.

You might also want to ask Allegro Micro and Texas Instruments tech support for recommendations because there are many new current sensors with I2C interfaces: www.allegromicro.com & www.ti.com

I don't have any first hand experience but there are probably many options. TI makes the INA169 and LM5056. Linear makes the LTC2945. I'd speak with TI and/or Linear tech support and ask for recommendations. You can also try Allegro MicroSystems: www.allegromicro.com and the Open Energy Monitor forum: community.openenergymonitor.org

Thanks. No plans for TOF. I think Andreas Spiess made a few videos.

You can only read 1 spot at a time with a single sensor.

How to send a message?

Don’t connect an LCD. Omit the LCD code. Replace all calls to lcd.message() with print(). A resistor is not necessary.

You can have up to16 INA219 sensors on an I2C bus. Each INA219 must have a unique address which is set using jumpers A0 and A1. See table 1 of the datasheet for more info.

Does the power source being measured share a common ground with the Pi?

Please make sure there is a common ground between the Pi and the source being measured.

Buying from Adafruit is your safest bet.

@@rdagger I looked and its the same board. the simple version with no extra connectors on the side. I bought 5 on amazon.de think it comes from china as its 2 week delivery. Tested a second on bus at other pins. same result , it sees it 1 time then dies. I toyed with adding 3k3 pull ups and changing the pull up to none etc on the pins. ( not the standard ones ). Nothing wrong with the power or gnd ( 3v3 ). Only thing is that I have nothing connected to measure as I just wanted to check the connection and software. maybe it's a bad batch. I will order some more from a different supplier in germany.

after testing with some diff pull up's I decided to do a test again on i2c bus 1. Got it working with 10k pull up added to sensor. Will try again later with 5k on bus 2. but happy to get it working.

If the INA219’s shunt is in series with the panel’s positive lead then you would only be able to measure current if there was a load. Many solar charge controllers have serial ports that log real-time data. I have a tutorial on reading data from charge controllers: ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-1VxP38XlVEQ.html

The video is old. See my website for updates. If you're using Python 3 then you may need to use pip3 install instead of pip install. If that doesn't work, please post the error message.

No. Watts is a measure of power as opposed to watt-hours which is a measure of energy. A 2-watt light bulb will consume electricity at a rate of 2 watts. If you run a 2-watt bulb for one hour, you've used 2 watt-hours.

I have not tried, but you take a look at: github.com/PyMySQL/mysqlclient

I don't understand the question, but if you want to measure household appliances then you would need a different solution because the ina219 can only measure up to 26V DC. Try openenergymonitor.org/

According to the first page of the datasheet it is 0.5% accurate over temperature.

Database integration is a very involved topic. Generally speaking running a database on the Pi is problematic because of the high number of write cycles to the SD card which has a limited lifespan. Nonetheless, many people have had good luck using SQLite: www.sqlite.org/ It has python bindings and is not as write intensive as MySQL or Redis. Another option is to use an online database. In my RocketChat tutorial I use MLab for database storage from the Pi: ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-BevcvRLsa9Y.html MLab offer a free sandbox plan that provides .5GB of data storage: mlab.com/

Sir only particular voltage saving in database apart from every repeated voltage,can this be possible,if possible can you please describe the procedure.

It is a complex topic, but basically you would need to do the following: Create a database, create a table with a numeric field to store the voltage readings, in your program you would need to establish a database connection, create a query or cursor object on the connection and then execute an Insert statement to append the voltage data to the table. Here is an example for SQLite: www.sqlitetutorial.net/sqlite-python/insert/

Its very helpfull,thank you

The resistors used for the LCD display are only to control contrast and brightness. Changing the values will change these levels which should be OK. 4.7K and 5.1K are very close. However, 51 ohm and 10K are not close. 10K stands for 10,000 ohm which is around 200 times larger than 51 ohm. Ideally, you should use potentiometers instead of resistors. This will allow you to fine tune the contrast and brightness. There is an example of using pots with an LCD display in my weather clock tutorial: www.rototron.info/raspberry-pi-international-weather-clock/

I have not tried it but Adafruit has an INA169.

I don't have any first hand experience but there are probably many options. TI makes the INA169 and LM5056. Linear makes the LTC2945. I'd speak with TI and/or Linear tech support and ask for recommendations.

I have a video on I2C displays: ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-i5A5AsDRRwQ.html

Generally speaking, you cannot measure current unless there is a load because current is a flow of electrical charge carriers (usually electrons). Connecting an INA219 across the solar panel output as you described would create a short circuit. This is usually undesirable and can damage equipment. However, there is an exception with respect to some solar panels. I recommend you check with your solar vendor before attempting to measure short-circuit current (Isc). So to answer your question, it might be OK to measure Isc which could give you relevant data assuming the Isc and the light-generated current have similar values. Again you would need to verify with the vendor and I would also inquire regarding the impact of long duration tests. I don’t think you would be able to simultaneously measure current and voltage because when you short the solar panel leads you should have zero potential (0V). You also have to be careful when measuring Isc that you don’t exceed the 3.2A maximum of the INA219 breakout board.

Sorry I never created a schematic. The drill had a proprietary motor driver IC with no info available. I just removed the trigger slide and replaced it with a potentiometer.

Take a look at my vegetable garden automation video. I connect an ina219 to an esp32: ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-x219R5rzRtU.htmlsi=tXsOLa8lNoLr3zWx

Generally speaking, household appliances around the world run on mains electricity which is usually AC between 100V and 240V. The INA219 is DC and only rated for +26V. Therefore, it would not work. Please be careful because mains electricity can be very dangerous.

@@rdagger then how can I measure ac devices sir

Please consult with an electrical engineer before experimenting with lethal AC voltages. Sorry I’m not aware of any safe hobbyist solutions. You could try the Open Energy Monitor forum: community.openenergymonitor.org/

You could time the charging of a capacitor with a couple of resistors but it would not be nearly as accurate or reliable.

Do you have a common ground between the 18650 and the Pi?

Yes a Pi ground, the LED negative, and the 18650 negative are all on the same bread board rail. The positive of the 18650 is in Vin+ on the INA and the positive of the LED is in Vin-. The VCC, GND, SDA and SCL on the INA are all straight to GPIO pins. Is your power supply current regulated in any way?

The wiring sounds good. I was not regulating the current. The voltage drop on the resistor isn't right. That would be a 17A load assuming a 0.1Ω shunt. Did you mean 1.7mV? Approximately, how many mA is your LED? I'd inspect all the solder connections on the board carefully with a magnifying glass.

You could place an INA219 breakout board inline with a battery to measure the voltage and current draw as long as you are not pulling more the 3.2A. I don't understand what you mean by auto detect. You can have up to 16 INA219 boards connected to a Pi. Each would need a unique I2C address.

At the moment you have to manually tell the software the address of the device after using i2c to detect the address. How about adding a function that auto-scans address and configures each device. 3.2A? Getting close to the limits of some circuits (one draws 2.8a @3.6v!)

It's possible but you would have to ask ChrisB2 (the creator of the INA219 library): github.com/chrisb2/pi_ina219

The author of the pi_ina219 library used in the video also has a MicroPython version: github.com/chrisb2/pyb_ina219 Chris's library is well documented and has sample code.

I don't think it will work if you have the Vin+ and Vin- terminals floating. I'm guessing the voltage readings could be arbitrary. Try connecting a load such as a bulb or an LED and see if the readings are accurate.

There is a post on the Adafruit forum that addresses floating inputs on the INA219: forums.adafruit.com/viewtopic.php?f=19&t=38745&start=15#p192276

Ok so if the load is not connected I could get strange readings. As suggested I have used an LED with 120 ohm resistor supplied from a 3.7 v cell. I get current out of range error when the code runs. I have the max current set to 2A the load is only drawing a few mA.

I just performed a similar test with an LED and it worked. Here's the code I used: from ina219 import INA219, DeviceRangeError from time import sleep SHUNT_OHMS = 0.1 MAX_EXPECTED_AMPS = 2.0 ina = INA219(SHUNT_OHMS, MAX_EXPECTED_AMPS) ina.configure(ina.RANGE_16V) def read_ina219(): try: print('Bus Voltage: {0:0.2f}V'.format(ina.voltage())) print('Bus Current: {0:0.2f}mA'.format(ina.current())) print('Power: {0:0.2f}mW'.format(ina.power())) print('Shunt Voltage: {0:0.2f}mV '.format(ina.shunt_voltage())) except DeviceRangeError as e: # Current out of device range with specified shunt resister print(e) while 1: read_ina219() sleep(1)

Thanks for all your info. Unfortunately I am getting the same problem using your code, current out of range (overflow) for gain 0.32V device limit reached. Also the bus voltage is incorrect measured 3.7v printed value 2.96v.

I don't understand what you mean by foot stepper. Are you trying to convert mechanical energy to electrical energy and measure the current? Or are you referring to stepper motors?

I don't have any first hand experience but there are probably many options. TI makes the INA169 and LM5056. Linear makes the LTC2945. I'd speak with TI and/or Linear tech support and ask for recommendations. You can also try Allegro MicroSystems: www.allegromicro.com and the Open Energy Monitor forum: community.openenergymonitor.org

kindly help

Make sure you have the correct voltage and gain settings for your circuit. You can use multiple sampling and 12 bit resolution to improve accuracy. Please check out the Github site for more info: github.com/chrisb2/pi_ina219

When you pip install the ina219 library, the dependencies should also be installed. Did you following the instructions on my website? Are you using the latest version of Raspbian Jessie?

I now what the Problem is. In the LXTerminal it works perfeclty but It doesn't work with Phython3. The error message cames only when I would open it with Python3. I wanted to creat a window to show the current and voltage.

You need to use pip3 to target Python3.

Yes, you can use the photo.

Thanks alot, your content was very helpful btw. @@rdagger

1. Yes, you can measure only voltage. 2. Yes. Each board needs a unique I2C address. This breakout board has jumpers for 4 different addresses (0x40, 0x41, 0x44 and 0x45) 3. Yes. The LCD is only for demonstration purposes.

Thx!!!!

You could use the ina226, it can measure up to 36V and is very accurate.

​@@rdagger Hi and thx for the video! on your response on the point 3, how have the voltage shown in raspbian instead of the LCD screen? please and thank you

You can use the python print command.

Not sure what you mean by directly, but you can measure current and voltage of a solar panel assuming you have a load.

@@rdagger that was exactly what I mean, thank you. So I need to find some light bulb and place it between INA219 and solar panel

@@kuban6141 Place the INA219 between the solar panel positive lead and the bulb positive lead. A better way to monitor solar power is to interface with a charge controller. I have a tutorial on the subject: www.rototron.info/raspberry-pi-solar-serial-rest-api-tutorial/

@@rdagger Thank you very much!

In theory it would work but you would have to wire the INA219 in series.

Thanks for the prompt reply! I guess that would limit me to 26v?

Yes. Furthremore, I don't think this would give you the desired results since the voltage would be cumulative. I'd ask around the RC forums.

Yes, see my site for details.

@@rdagger thanks

No, the resolution is only 12 bit over 0-26V. I don't know of any hobbyist sensors that will meet your accuracy requirements. I recommend you ask Texas Instruments tech support. They make a lot of other voltage/current sensors.

@@rdagger thank you, a friend of mine mentioned that I might be able to use a mosfet do you have any experience working with these?

DIY high accuracy, stability and linearity can be challenging due to the influences of noise, EMF, temperature, etc. It may also require expensive equipment to calibrate. There are many new high resolution ADC chips available. I recommend you post your question to a metrology forum: www.eevblog.com/forum/metrology/ Be sure to provide a detailed outline of your requirements such as accuracy, sensitivity, resolution, range, are you measuring voltage only or voltage & current, are you measuring DC only or DC & AC and a summary of your project.

Connect the positive of your battery to Vin+ on the INA219. Connect the negative of your battery to the INA219 ground.

rdagger68 great, thank you! Also, you said you can daisy chain multiple ina’s. If you wire them all to 3.3 and ground, and all to the same 2 i2c pins, will they all be given different hexadecimal values? If so, how many different ID’s can be used?

You need to set a unique address for each INA219 using pins A0 and/or A1. Table 1 in the datasheet lists the 16 possible combinations. Many of the breakout boards have AO and A1 pads that can easily be jumped with solder to VS+ which gives you 4 easy to set addresses 0x40, 0x41, 0x044 and 0x45.

I have set up this no load configuration with an 18650 battery around 4.09V. I am using the code without an LCD display from www.rototron.info/raspberry-pi-ina219-tutorial/ and am getting the following results: Bus Voltage: 2.57V Current out of range (overflow), for gain 0.32V, device limit reached Do you have any suggestions about what to do next? I am simply trying to get a Pi output with the correct battery voltage.

The INA219 can only measure DC voltage up to 26V.

whta can i use for that?? for a.c

Please consult with an electrical engineer before experimenting with lethal AC voltages. Sorry I’m not aware of any safe hobbyist solutions. You could try the Open Energy Monitor forum: community.openenergymonitor.org/

You can measure the voltage and current coming from a solar panel as long as you do not exceed 26V and 3.2A. However, most solar charge controllers already provide this data. Please check out my solar tutorial: ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-1VxP38XlVEQ.html

@@rdagger I saw your video but my panel output is around 21 volt and 2.5 A max. Actually I am doing college project so I just want to calculate power generation by panel and power consumption by pi. So can I use this method ?

You should be able to use an INA219 to track the voltage and current of a panel. A 2nd INA219 should be able to track the Pi's voltage and current consumption.

@@rdaggermeans this set will work ?

It depends on your components and the desired goals. A good charge controller would probably provide more meaningful results on its data bus. An INA219 inline with the solar panel should give you accurate voltage levels during the day and show how much current the charge controller is drawing. An INA219 inline with the Pi will show you the Pi's voltage and how much current it is drawing.

Any board that supports I2C should be compatible assuming you can get a driver for it.

@@rdagger ya so I got the driver as well as my pcb supports ina219 but the thing is I am not able get the data from it. Would be great if you could guide me 🙏🏻

@@SohanChhangani What board are you using? What driver are you using? What error message are you getting?

@@rdagger custom pcb with cm3566 som

@@SohanChhangani Sorry I don’t have any experience with the CM3566. Generally speaking, I would do the following: 1. Verify your wiring is correct and try to keep the lengths as short as possible. 2. Check with the manufacturer and see if external pull-up resistors are required. 3. Find an Android I2C test utility and see if the address of the INA219 shows up. 4. Find a forum that specializes in Embedded Android Development or Android IOT Development.

You could measure the voltage at the LED anode. You would connect Vin+ to the anode and leave Vin- disconnected. If you had a second INA219 you could also measure the voltage at the cathode and then you would be able to determine forward voltage drop. There would need to be a common ground between the INA219 and your circuit.

Is there absolutely no way to do it just through the LED's legs without connecting its circuit's ground to the INA219/raspberry pi 3 ?

The LED cathode might be at ground. If that's the case you have your ground. However, the cathode could instead be in series with a resistor. The case of the device in question might also be at ground. Generally speaking, you can't measure voltage of a circuit without a ground reference because the definition of voltage is a potential difference. Why do you want to measure the voltage?

I have a device which measures power across home appliances and based on the measured values it would turn on some LEDs as to indicate how much I am consuming. I am trying to detect when some of those LEDs are ON/OFF and sending that to a raspberry pi which would in turn send me an email/sms warning. The measuring device is very tightly locked and I prefer not to touch its internal circuits which is why I am left with only the LED legs. I hope there is some way I can achieve this ? Thank you for being helpful !

Any idea Please ? I can't find anymore help on this

What error message are you getting? Did you use pip3 to target python 3?

What is the error message?

======================== RESTART: /home/pi/ina219.py ======================== Traceback (most recent call last): File "/home/pi/ina219.py", line 1, in from ina219 import INA219, DeviceRangeError File "/home/pi/ina219.py", line 1, in from ina219 import INA219, DeviceRangeError ImportError: cannot import name INA219 >>> ======================== RESTART: /home/pi/ina219.py ======================== Traceback (most recent call last): File "/home/pi/ina219.py", line 2, in from ina219 import INA219 File "/home/pi/ina219.py", line 2, in from ina219 import INA219 ImportError: cannot import name INA219 >>>

this error . can u help me

Are you using auto gain? What voltage and current are you trying to measure?

the led . I want measure voltage and current this led. . auto gain ? can u give email to me. I sent pic. help me pls

Yes as long as the load doesn’t exceed 3.2 amps.

Hi, I'm really sorry to bother you but I'm actually tryna measure the amps, volts, and power of a charging 24 volt battery with a RPI that's being charged by a solar panel (I'm building a portable solar charger for my school project). So, if I need to measure amps that's more than 3.2 amps I need to get a lower value resistor for the INA board? Also, you mentioned that another alegro board is suitable for higher amps, If I get that, will the steps in this tutorial be the same? I'm struggling with this and I really need to pull my grades up :( I hope you can help me. Thank you and have a great day

@@antisocialboy818 The Allegro is just and AC or DC current sensor. It doesn't measure voltage and it produces an analog output as opposed to digital. There are many other sensors that would probably be a better fit such as the INA260: www.adafruit.com/product/4226 You should ask Texas Instruments tech support for a recommendation.

@@rdagger will do, thank you :)

I don’t have any tutorials on measuring battery life. You could use an INA219 to determine average current consumption which could be used to derive run-time: Battery capacity (mAh) / Average current consumption (mA) = Hours of expected run-time

What do you want to measure?

rdagger68 I want to measure the current, and voltage produced by the solar panels.

First when dealing with lethal voltages I strongly recommend that you confer with a qualified electrician. Since you are working with solar, the most efficient approach would be to use a 96V MPPT charge controller with an RS232 or RS485 data output which could easily be polled by a microcontroller or computer such as the Raspberry Pi to provide detailed energy usage. I have a tutorial on the subject: ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-1VxP38XlVEQ.html Another option is the Open Energy Monitor project which provides comprehensive open source solutions for monitoring solar power and other energy use: openenergymonitor.org/

rdagger68 is there a way to measure the voltage produced by the solar panel with the method showed in this video? Instead of measuring the power consumed by the lamp, maybe I can do the same for the panel..

rdagger68 thank You! Very good tutorial

No. You can use a hall effect-based sensor such as the Allegro ACS712. Please be careful if working with potentially lethal mains electricity.

Not a problem working with electricity.....im also a lic electrician. :-) Thanks for your help!!

Please consult with an electrical engineer or electrician before experimenting with lethal voltages. You could try the Open Energy Monitor forum: community.openenergymonitor.org/

@@rdagger know that, be careful, and also know how dangerous 150V / 100 Amp is... now that the disclaimer is out of the way, anyone help, and no OEM can't at this stage as all their stuff is AC based.

You can try Allegro MicroSystems: www.allegromicro.com/

I installed all the libraries but nothing is showing

Did you enable I2C using raspi-config? Did you verify the address using i2cdetect?

@@rdagger Thanks for the quick response time! I enabled I2C via the Raspberry Pi Configuration menu and i wouldn't get any detection using i2cdetect. I made sure i connect the pins to the proper gpio pins and everything.

@@rdagger Could there be an issue if the wirelength of the wires I used to connect the pins is a bit long? The wires about a foot long.

@@stonkyboy7606 1 foot is within spec but I would test the wires with a meter and try swapping them out. What Pi are you using? You could have a bad INA219.

Yes. Each one needs a unique address. See table 1 in the datasheet for the address pin settings.

@rdagger68 Ah ok thanks! Will the library also work for INA226?

I haven’t tried, but I would guess no. You could submit a feature request to chrisb2. It looks like he has made updates to the library since the video came out.

I don't think it would be a good choice because the input bias is probably too high. I'd check with Texas Instruments support. They can recommend an optimal device for your needs.

@@rdagger can you name

@@rdagger actually i m working on wood moisture meter and i have to measure serial current though wood. which i consider will be in micro

Dave Jones makes a precision current adapter: www.eevblog.com/projects/ucurrent/

Maybe I'm too tired but I do not see an error message. I see a prompt to continue.

rdagger68 Why when I update sudo can not just any posts need additional memory y / n

Can you speak Indonesian?

Sorry no.

I don't understand. Maybe you have a bad sd card. Did you expand the SD card?