WS2812b led strip and arduino, should I give up?

Grumpy_Mike:
So is the monitor connected to a mains input that includes a ground connection? Have you checked that the ground in your home wiring is actually bonded to the real ground?

You could have some faulty wiring causing a problem with your earth leakage currents.

Yes monitor is connected to mains with good ground.
Real ground.

Then you are wrong about something, how did you check this? Have you had the monitor PAT tested recently?

As your computer does not reset there would seem to be nothing wrong with that and no danger to it. But connecting power supplies when things are powered up is a big no no in electronics as I have told you before.

Grumpy_Mike:
Then you are wrong about something, how did you check this? Have you had the monitor PAT tested recently?

As your computer does not reset there would seem to be nothing wrong with that and no danger to it. But connecting power supplies when things are powered up is a big no no in electronics as I have told you before.

Home is new and the electrical wirings are new so the ground should be ok.
In Mike I trust so if you say that is ok it's ok but I still don't understand why the monitor goes black for one second.

In any case I will not connect things if pc is powered on anymore.

Thanks

but I still don't understand why the monitor goes black for one second.

The monitor goes black because something is disturbing the monitor circuit. It is not affecting the PC otherwise the PC would reset, as they tend to do at the slightest opportunity.
The problem occurs when a second supply is connected. This second supply is floating with respect to the monitor and so when it is connected the ground of the power supply and the ground on the monitor are brought to the same potential. This causes a very short inrush of current which disturbs the monitor’s operation. The result is that some protection circuit in the monitor shuts it down. Then the inrush stops as the two supplies are now equalised and the monitor comes on.

Just as a test remove the LED strip and just connect the power supplies’ negitave terminal to the rest of the circuit. It should still happen.

I didn’t ask before but I am assuming the monitor is a CRT one.

Grumpy_Mike:
The monitor goes black because something is disturbing the monitor circuit. It is not affecting the PC otherwise the PC would reset, as they tend to do at the slightest opportunity.
The problem occurs when a second supply is connected. This second supply is floating with respect to the monitor and so when it is connected the ground of the power supply and the ground on the monitor are brought to the same potential. This causes a very short inrush of current which disturbs the monitor’s operation. The result is that some protection circuit in the monitor shuts it down. Then the inrush stops as the two supplies are now equalised and the monitor comes on.

Just as a test remove the LED strip and just connect the power supplies’ negitave terminal to the rest of the circuit. It should still happen.

I didn’t ask before but I am assuming the monitor is a CRT one.

if I remove the LED strip and just connect the power supply, nothing change, monitor goes black for a second.
the monitor is a new LCD connected via display port to a dedicated gpu

do you think that something like this:

should help preventing this problem and making my computer "more safe"?

is there someone who is using USB Isolator here with an ESP? Does it work well?

The thing is this.

We don't know what you are doing wrong, despite all the back-and-forth questions.

If we can ever fathom what you are doing wrong then we could tell you not to do it, but so far we can't, so we cannot recommend some random device.

So far I don't think we know exactly what your 5 V power supply is, and a detailed, precisely focused photograph of your whole arrangement in bright, uniform light would be most illuminating.

do you think that something like this:
Adafruit USB Isolator - 100mA Isolated Low/Full Speed USB : ID 2107 : $34.95 : Adafruit Industries, Unique & fun DIY electronics and kits

should help preventing this problem and making my computer "more safe"?

No I imagine the problem will still be there because their is no indication that their is anything to do with the USB connection.

Just power the Arduino separately and try the test again.

I quite agree with Paul__B in what he says.

this is is the power supply for the led strip.

and this is the build

if you need more info, I'm here.

a more clear graph of the connections:

Thanks for that, it is gives a better understanding of what you have, your construction looks quite good.

I am not sure how much reliance to put on all those declarations of conformity on that power supply but the big one that is missing is the UL mark. It is not surprising as it is perhaps the hardest one to obtain with respect to a power supply. UL is largely driven over concerned with the device starting fires, hence the name Underwriters Laboratory, it arose from insurers wanting to minimise their risk. I do know that UL has a much stricter limit on the earth leakage current than some of the others like CE. I am not saying go for a UL rated power supply and it will solve the problem, they are a lot more expensive and the UL rating mark might be false on Far East designs.

What it doesn't change is my guess that the problem is probably due to earth leakage and the best bet is to have things connected up before applying the power. I know this means powering down then up before making changes to things, but that is what you should do anyway. Even plugging a monitor into a computer that is running can damage things. Their are things you can plug into a computer when it is running, like USB but they have been specifically designed to be "hot swapable", these designs connect the power rails before connecting signals.

I have a similar problem with the Raspberry Pi, you can't plug many things into the GPIO pins without resetting the computer. In the case of the Pi that is a bad thing, because its main storage is an SD card, and unexpected interrupts runs the danger of destroying the card. And being Linux the boot up process is not swift.

Grumpy_Mike:
Thanks for that, it is gives a better understanding of what you have, your construction looks quite good.

I am not sure how much reliance to put on all those declarations of conformity on that power supply but the big one that is missing is the UL mark. It is not surprising as it is perhaps the hardest one to obtain with respect to a power supply. UL is largely driven over concerned with the device starting fires, hence the name Underwriters Laboratory, it arose from insurers wanting to minimise their risk. I do know that UL has a much stricter limit on the earth leakage current than some of the others like CE. I am not saying go for a UL rated power supply and it will solve the problem, they are a lot more expensive and the UL rating mark might be false on Far East designs.

What it doesn't change is my guess that the problem is probably due to earth leakage and the best bet is to have things connected up before applying the power. I know this means powering down then up before making changes to things, but that is what you should do anyway. Even plugging a monitor into a computer that is running can damage things. Their are things you can plug into a computer when it is running, like USB but they have been specifically designed to be "hot swapable", these designs connect the power rails before connecting signals.

I have a similar problem with the Raspberry Pi, you can't plug many things into the GPIO pins without resetting the computer. In the case of the Pi that is a bad thing, because its main storage is an SD card, and unexpected interrupts runs the danger of destroying the card. And being Linux the boot up process is not swift.

Thanks for the answer and the patience.
This will be my last question to not bother you.

Suppose that there is an earth leakage, how connecting things before powering up them all can be safier?
Doesn't the problem remains?

Thanks for the explanations and the patience :slight_smile:

Suppose that there is an earth leakage,

There is earth leakage in all power supplies, it just simply a matter of how much.

how connecting things before powering up them all can be safier?

Because when they are connected together before powering all the elements of the circuit are at the same potential. Then when they are powered the earth leakage currents start all at the same time so there is not a sudden change in the current. It is the sudden change in the current that causes a change in voltage which is what I think is upsetting your monitor.

You could say it is hidden in the start up process, the monitor is not yet on so you don’t see it go off. That is a bit of an over simplification but not too far from what happens.

This will be my last question to not bother you.

Don’t worry, keep asking about what you don’t understand.

Grumpy_Mike:
Don’t worry, keep asking about what you don’t understand.

Thanks for making arduino's world a better place Mike,
I really appreciate your help. In a world where internet is full of shitstorms and trolls I'm really glad to find people like you and the others here :slight_smile:

for what it worth here the results of those circuit.

ambilight video 1, youtube video

ambilight video 2, AC Odissey gameplay

it's pretty cool :slight_smile:

Not exactly mentioned so far, but it is generally the case that a power supply which has a ground lead on the AC cord, will also have that ground lead connected to the negative of the output. Which is why Mike is suggesting that it has something to do with the odd behaviour.

You can verify that connection with a multimeter.

Paul__B:
Not exactly mentioned so far, but it is generally the case that a power supply which has a ground lead on the AC cord, will also have that ground lead connected to the negative of the output. Which is why Mike is suggesting that it has something to do with the odd behaviour.

You can verify that connection with a multimeter.

I'm sorry, probably I'm too noob to understand it, can you explain in different words where should I connect the multimeter?

You will almost certainly see continuity when you probe between the ground pin on the mains power input and the sleeve (negative) of the output connector.

for what it worth here the results of those circuit.

You might find this hard to believe but I have never seen an Ambilight in action before thanks.

Yes with a power supply with that much current output you are likely to have a 3 terminal mains connector. However in my experance the outputs still float rather than being connected to ground.

Grumpy_Mike:
Yes with a power supply with that much current output you are likely to have a 3 terminal mains connector. However in my experience the outputs still float rather than being connected to ground.

Which is why we want to know whether it is connected or not as part of the forensic trail.

To have the secondary not grounded, the transformers need to comply with "double insulation" standards even though it is not so marked as it has a ground pin.

Regarding the "that much current output", almost all 19 V laptop power bricks with a three pin plug do have the output grounded.

my ignorance in this matter doesn't let me understand well what you are saying.
should I change my power supply for a better one?