I noticed that I was getting mild shocks while touching my plugged in laptop and touching a metal shielded USB device plugged into the USB port of my Raspberry Pi 3 at the same time. I used a multi meter to measure the AC voltage between the USB grounds on my laptop and Rpi3, and found there to be 130V.

My country uses 240V mains.

The laptop charger is a DELL USB C charger with ground AC pin.

The Raspberry Pi 3 charger is the official 5V supply and this does not have a ground AC pin.

Both are rated correctly for 240V.

I checked the voltage between the USB C ground of the DELL USB C charger to mains ground, and it's zero. It's also zero with the Pi 3 5V supply microUSB ground to AC ground. But only between the USB C charger and Pi 3 supply grounds there is 130V.

I also tested with another Pi 3 supply of the same brand, and that one gave me 25V between the two.

Anyone have any idea what could be going on here?

  • 3
    \$\begingroup\$ reverse the pins on the 5V supply. You are measuring leakage Voltage betwen the Dell line filter to ground and the 5V leakage to line by coupling capacitance.. approx > 1 MOhm on the unknown meter. That results in TBD xxx uA * ~ 1Meg=130V \$\endgroup\$ Mar 12, 2019 at 4:26

1 Answer 1


The problem is most probably caused by "Y" filter capacitors between the P & N mains lines and the ground line in one of the supplies.

If the ground line is grounded no voltage occurs compared to mains ground. If the ground line is ungrounded it floats at half supply due to the capacitor voltage divider formed by the two Y caps.

As the impedance is high the voltage you see when you measure it depends on the meter design and the value of the Y caps. eg a supply with 0.001 uF Y caps is more liable to be 'loaded' by a meter than one using 0.01 uF Y caps.

The impedance of this floating voltage is such that an unpleasant sharp feeling may be felt if the ungrounded ground is touched but the currents are well below those usually deemed dangerous. However, equipment can be damaged by exposure to this AC capacitor-coupled voltage.

A solution is usually to ensure that all "grounds" are grounded to a common ground. If this grounding occurs when an interdevice signal cable is plugged in it may cause damage as connection is made. Long ago I had a brand new printer destroyed in this way the first time it was plugged into a PC.

  • \$\begingroup\$ The power supply would not be UL listed if there was a likelihood of it failing in a way that would injure someone . Of course it may not be UL listed in any case. \$\endgroup\$ Mar 12, 2019 at 19:05
  • \$\begingroup\$ @Harper I agree, at least in theory. However, what I describe is not a failure, notionally cannot "injure anyone", and in the case described is due to the interaction of two or more "objects". I say objects as that can be another system and/or a person. The voltage to "mains ground" exists because of a supply with the lack of an earth in the mains lead AND two Y filter capacitors. This combination makes little sense in practice if the ground is not returned to true ground in some manner but may be required to meet regulatory requirements. ... \$\endgroup\$
    – Russell McMahon
    Mar 12, 2019 at 21:18
  • \$\begingroup\$ @Harper ... Over the years I have met MANY cord connected domestic & commercial computer equipment power supplies that have this "output ground floats at about half mains voltage at high impedance" feature. Even power supplies from companies such as Toshiba , who are Jedi-Masters [tm] of the laptop art, can have this "feature". \$\endgroup\$
    – Russell McMahon
    Mar 12, 2019 at 21:19

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.