Electrical problems (shock touching the metal box)

[ilario]

Dear all,
From time to time, I am experiencing sudden shutdowns at the moment of touching the metal box of one of the two A64-OLinuXino-2Ge8G-IND I am using.
Sometimes, I even feel a tiny electrical shock when touching the metal box (and the Olinuxino goes down).

This happens more often while it is under CPU load (even if

Code Select Expand

cat /sys/class/thermal/thermal_zone0/temp never goes above 60 C under stress with s-tui and using a big heat sink and the metal box lid off).

It does not happen with the other A64-Olinuxino unit I have.
I suspect I damaged something soldering the GPIO1 or the UEXT ports. Or connecting stuff to the GPIO in the wrong way (I am a chemist, not an electrical engineer XD)...

Does anyone else experience this?
Should I connect the metal box to the ground of the building's power line?

Thanks!

[LubOlimex]

You didn't install the box properly. There should be no direct contact between the box and the board. If you touch the box and the board gets influenced (or you get influenced) then the installation is not proper.

Probably some of the spacers were not placed properly.

Edit: Now that I saw your other post - the heatsink should also not touch the box. Investigate where the board touches the box, any metal parts of the board shouldn't make contact with the box at all.

[ilario]

Thanks LubOlimex, I can confirm that the box and the board's ground were not in touch. I also replaced the box with another A64-Olinuxino unit with no improvement. The heatsink I am using is huge, but still it does not touch the lid of the box (also, I am not using the lid, as the jumpers connected to the GPIO1 header are too large to allow the box to close).

I felt the shock also touching other components. So I suspect that the shock was caused by something wrong with the multiplugs power cord (e.g. its integrated SPD surge protector) or with the building's power connection, not due to the A64-Olinuxino unit.

So I followed BakhamMamodar's suggestion (thanks!) and connected the ground pin of the GPIO1 port to the ground of the building, and the problem didn't happen again.

I used the ground connection from the same multiplug power cord, in a very unrefined and unsafe way, as you can see in this picture:



Thanks!