AtxRaspi powers up randomly

[Jean-zouz]

Hi everyone, 
I'm using ATXraspi system with the volumio music distribution. This little guy is working well except for one thing, it powers up suddenly, without any action from me. So I have added a power supply  switch to my circuit to solve this temporarily but still...  When I switch on the power supply without pressing the power button, the Pi powers up at some point,  maybe 1 or 3 hours later. Does anyone ever experienced this?
Many thanks.
Jean-zouz from Lyon France.

[Felix]

Very strange but could be explained by a faulty power switch, or other hardware glitches.
Can you try it with another switch?

Are you using the momentary button mode?

[Jean-zouz]

Hi Felix,
Yes I'm using a momentary button with integrated LED, one of those with the power sign on it you can find for couple of euro on AliExpress. To describe a bit more: to get rid of this strange behavior, the power switch I have added is before the atxraspi circuit (like if I unplugged the circuit).
Then I'm using very standard configuration: micro USB input,momentary button, and I have soldered terminals on the board to communicate with gpio, like on the YouTube setup video. All the rest is working properly. So I definitely think this is an electrical problem, not software. I'll try with another press button and let you know.
Thanks,
Jean-zouz.

[Felix]

The ATXRaspi should only turn the output ON in response to a button press.
After you try another button, if the behavior still occurs, I would need to have the item returned to me for debugging before I can really diagnose any potential problem with it, but we can talk about that later.

[perky]

That sounds very much like a floating input, maybe a lack of pull-up. It that's true just touching the pin or probing with a meter might induce the same behaviour.

[Felix]

As always, all the boards go through some test procedures, with a real button to ensure they power up and can issue a SHUTDOWN etc.
There's a possibility a floating input (cold solder joint, tombstoned resistor that I did not notice) was able to work here in testing but fail in real use.
A close look at the board could reveal such hardware problems.
If you could post a high resolution closeup photo of the board maybe we can spot something.

[Jean-zouz]

Hi all,
Perky, you are totally right! It powers up when I'm probing using my meter!
I'll post a Hi res picture tomorrow!
Thanks,
Jean-zouz.

[Felix]

Please check this post, just another example where actually flux and soldering residue causes strange behavior:
https://lowpowerlab.com/forum/atxraspi/atxraspi-shutdown-led/

I only discovered after inspection *under* the headers plastic holder, the last place you'd expect something to fail.
And yes probing the circuit can cause that, also light shorts by flux residue can cause other such behaviors.
So a really good clean solder job is required for this board to not have currents drain through unexpected pins.

[Jean-zouz]

Hi Félix,
I'll also check soldering job if I have time tonight. One thing I have also noticed though: the boot OK led seems to constantly have a small current passing through. When I power up, before having the feedback OK from the pi, this boot OK led is On, very weak, but on. Afterwards when I get the boot OK signal, then it is really strongly on. So yes I'll definitively check soldering and send pics asap.
Thanks for support.
Jean-zouz

[Felix]

Jean-zouz,
Very very dim Green LED ON is normal after you power the Pi. It's almost invisible. It's because the Pi makes that pin have a very small voltage (floating pin), until it is set to HIGH by the ATXraspi shutdowncheck.sh script.