Let’s be logical here, and see if I missed anything.
Before when you powered the device even at 5V it worked? That is to say that 0x44 was an identified address regardless of whether or not that address is valid to A0.
It is not until you provide power to the device on the VPU pin that the I2C bus fails?
If so my guess is failed component, not sure how sensitive it is to esd and without a schematic I cannot identify a TVS diode etc so you could have smooched it.
It is a cheap component, I would get a Arduino and another sensor and see if you can get a
control, Amazon it will cost 20 bucks and as a learning curve it will help.
You will go around in circles otherwise and there is in reality only a limited amount of debugging we can do on here.
With the device powered on with no VPU attached what can you get from the device?
Refer to the data sheet again and see what you can read! Even a zero value is something!
Try to lower the power on the VPU if you have a variable bench supply and maybe add some filtering on the supplies.
Yes you are missing a lot. Your schematic changes were not enough. Look at the number of coffee cups Tom was giving you. The more cups the more he needs to prevent him from having a breakdown.
We have told you time and time again about what is wrong with your schematic. So read Tom's comments and try and comply. Your photograph was very useless because it didn't show how you normally wire it up. We are after seeing what happens when you wire it up normally, not when you start disconnecting random stuff.
We are not interested in the output side of the board, the trouble you are having is with the communication between the controller, which seems to be a Pi (why it still does not say so in your schematic?)
So when you change the voltage driving the board, what do you measure on the I2C lines?
How do we even know what you are saying is what you are really doing? Not deliberate but because of you lack of knowledge.
Thanks @bjorntocreate for your response and the follow up.
I don’t use the VPU to power either my load (LED) or the board. Instead, I only use the VS to power the INA3221 board by 3.3V from my controller. For the LED it is powered separately by different power supply as shown in the schematic. I tried both 3.3V and 5V to power the LED, but I am still having the same problem. Also, in both cases, I was using 3.3V to power the INA3221 via VS pin. Does this make sense to you?
For the VPU, I only used it when you and @Grumpy_Mike suggested to power the board on that way and measure the voltage on SCL and SDA. So, I did that accordingly. Other than that, I don’t use VPU at all. I use VS to power the board and VIN+ to power the LED. Again, when I tried both 3.3V and 5V to power the LED and measured the voltage on SCL and SDA, both read around 3.28V regardless the power supplied to the LED itself. So, I think that the voltage on both SCL and SDA is governed by the voltage provided to VS pin and that it has nothing to do with the voltage provided to VIN+ shown in schematic in the updated post.
Do you think it is clear enough? Or do you think I am still not explaining it correctly?
Sorry your diagram in post #1 doesn't show a Pi, just "controller"?
If the device in the image in the grey case is the Pi, I and many did not recognise it
Please do not go back and edit posts, place any info in new posts to allow anyone following this thread to follow the flow of posts.
What I meant was a schematic of your project, using component symbols, blocks for modules, power supplies, pin labels.
I am going to be straight up honest here and say I have actually no idea what the issue is.
I cannot see how you would see the device if it were not powered so when are you actually seeing the 0x44 appear? Unless it is taking power over the pull-up resistor to VS pin on the i2c bus and logically as soon as those pins get pulled low it will lose power save as to some small amounts of capacitance and inductance.
As a final idea remove everything from the sensor, attach the SDA and SCL pins GND from the PI and 3v3 from the pi and see what happens then, the TS led is not a problem I don’t think as this pin is pulled low at fault and unless that is triggering a transistor etc but I don’t think in any event that is anything to do with the issue as the i2c bus would also report fault and not fail as a fault mode.
It is a very simple circuit and beginner or not you should be able to deal with this unless like I said you have a component issue which is where I am leaning, also if you have another i2c device test it to see if you killed the I/0!
I also assume you are using secure shell (SSH) to access the pi, there are some things diagnostically you can do from there.
There are some excellent books you can invest in if this is something you may want to pursue, also great books on schematic layouts, if you do not want to spend much then get a copy of KICAD which I think is free and I think Eagle is also but I am not that familiar with them, Altium costs a lot but it is cheap for what it is, but you don’t need that, just some experience and knowledge.
Also do not take what we say as being dicks, I know Tom and Mike have many years experience as Engineers as do I, we just need to have things presented in a way that we can assimilate and answer, we appreciate that you are not at that level but as Tom said editing posts is not a good way to keep an audit trail as you have affected the chronology and we have no reference and we would not likely go back unless there was salient data to do so.
Plus I am on the spectrum as I suspect a lot of guys on here are so we need some order or the world will burn!
What was the difference between the way you powered things when you got 4V5 (or so) and when you didn't get it. So that is two diagrams one with the high voltage I2C and the other with the low one. And try and make them proper ones.
@TomGeorge, here is the schematic of my project. I power the INA3221 board by connecting VS pin to the 3.3V pin on RPI. I also have a common ground between the board, RPI, LED and the 3.3V power supply. I connect the SCL to SCL pin (#5) on the RPI and similarly, I connect SDA pin on the board to SDA pin (#3) on RPI. I connect the +ve side of the LED to VIN- on the board. I connect +ve side of the 3.3V power supply (that is supposed to power the LED) to the VIN+ pin on the board.
Could you see anything wrong or nonsense in this schematic?
@bjorntocreate
As you can see I don't use the VPU pin on the INA3221 board, I power the board by 3.3V pin on RPI through VS pin on the board. Having this setup, I can measure 3.28v on SCL and SDA.
For the LED (load), it gets its power from the independent 3.3V power supply. However all of them have a common background (Power supply, LED, RPI and INA3221 Board).
Do you think I have to use the VPU and for what?
@Grumpy_Mike
I measured that 4.8V on SCL and SDA when I connected 5V power to VPU pin as you suggested before. I thought you wanted to investigate how the circuit works. But, in my setup, I haven't used the VPU pin at all.
When I have the 3.3V pin of the RPI connected to VS pin on the board, I can measure 3.28v on both SCL and SDA. Then, when I turn on power supply and the LED illuminates, I still measure 3.28v on SCL and SDA.
Do you think I need to have the VPU included? and if so, why do I need it and shall I connect it to what source, the 3.3V from RPI or the 3.3V from the power supply?
Thanks for the image but that is not a schematic.
Can we please have a circuit diagram?
An image of a hand drawn schematic will be fine, include ALL power supplies, component names and pin labels.
The LED seems to be directly connected between 3.3volt and ground (shorted 330 ohm resistor).
With the 10 ohm resistor also across the LED.
That will shortcircuit/overload the 3.3volt breadboard supply.
How is that breadboard supply powered (voltage/current). I see via the DC socket.
Leo..
@Wawa
Yes, the breadboard is powered by 3.3V supplier which is plugged to 9v DC adapter.
I know what you mean, but when you look into the GIF I posted, it is working normally without a shortcircuit. Actually, this small 10ohm resistor was suggested by @Grumpy_Mike and it works. Without it I wasn’t able to read i2c address at all while and while not connecting the power supply.
I see the orange wire, the 10 ohm resistor, TWO wires of the 330 ohm resistor and the LED,
all plugged into the same 5-pin rail of the breadboard. Shorting the 330ohm resistor.
Maybe I see it wrong, or you don't know yet how a breadboard works.
Leo..
Oops, you are right. I would change that. Maybe I didn’t have issues as I run it for just few seconds to quickly check the i2c address.
Thanks for highlighting this .
But, do you have an explanation why do I loose the i2c address when I turn on the power supply?
As I mentioned earlier I only use VS to power the INA3221 board from the 3.3v pin on the RPi. So, I don’t use the VPU pin on the board. The LED gets the power from the 3.3v power supply. So, there is no need for the VPU pin, right?
@TomGeorge
Is this schematic/digram what you are asking for? Or it doesn't help?
As you can see, I power the board by 3.3v through VS pin. Do I need to utilize the VPU pin as well in some way, or we don’t need it here?
I didn’t connect any external resistors other than the ones connected to the LED.
For the INA3221 board, I can read the following values:
Between SCL and VS: 9.90 k.ohm
Between SDA and VS: 9.98 k.ohm
So, I think these resistors are already imbedded in the board, or do I need to hook external resistors as well?
The INA's built-in 10k pull up should be ok for these short interconnections.
Shorting the MCU and INA supply with a ~2volt Vf LED with 0.1ohm CL resistor can't have been good.
Powering the whole setup that could draw 500mA through a linear breadboard supply with 9volt in is another red flag. How hot do the regulators get.
Don't forget that his is an Arduino forum, with only a few members having experience with a Pi.
IMHO, you shouldn't even have asked here for help with a Pi.
Leo..
.
