# How do physical connection interfaces handle floating voltages

I am sorry if this is a bit of a novice question but I am new to electronics and have come to a problem that I can't quite see an obvious answer for.

A while ago I had a bit of a scary encounter when connecting an Arduino powered machine to my pc. The device worked fine with a PC power supply but once I swopped it for an open frame type one I noticed some serious arcing when I brought the USB cable near my pc's port. After some research I determined that this was caused by the fact that the new power supply was not connected to earth internally and once I made the necessary modifications my problem was fixed.

What I am really stumped about is how consumer electronics product manage this floating voltage problem when interfacing with attached devices. What I am wondering is how for example a laptop running from its battery without any possible earth reference can connect to a USB printer without any problems. Is it that the internal regulator in the laptop immediately assumes the USB ground as its ground reference and if so, how would that prevent an initial arc?

I think the problem you experienced was actually because there was a ground connection at both ends. What you saw was the end results of a ground loop.

When you have two systems each with their own power supply, both connected to ground through different routes, the resistance of that ground path can differ for each device. That means that in fact the electrical potential at ground can be different between the two units.

Connect those two different voltages together, with what basically amounts to a short circuit, and you get a spark.

When you have something that has no ground connection there is no voltage difference, since only one side has a potential to real ground. In the laptop the earth connection does not form part of its circuit in any form whatsoever.

Incidentally, the USB specifications actually state that the shield of a USB cable should only be connected to ground at the host end, not the device end - unless that device is purely bus powered in which case the shield forms part of the ground return for the power.

• Ok I guess that might make sense, the power supply initially only had a live and neutral connection to the mains and I then attached the earth wire to the ground connection to solve the problem. Would this indicate some complicated earth connection through I guess the neutral wire? Also, what I mean is that the laptop will have some internal ground or 0V reference and the printer will have another voltage as its ground or 0V, now say these differs dramatically, why would there not be an arc between them? Commented Dec 24, 2014 at 19:10
• There may be, but you don't see it because of all the plastic around the connector. Also, if the printer is designed properly, the USB socket's shield won't normally be connected to ground. Commented Dec 24, 2014 at 20:34
• Ok but so what your telling me is that the major arc that I saw came from the shield instead of the pins? Commented Dec 24, 2014 at 20:59
• How major is major? It's also possible that there is capacitive coupling from the live inside the power supply to the case; on equipment where this occurs it can sometimes be felt as a sort of velvety texture when touching the case, and rarely, even produce a tiny spark - 'major arc' is a bit of a stretch though. Commented Dec 24, 2014 at 21:23
• Ok, what I saw was that when the usb cable was about a cm or so away from the usb plug there came an arc between the two through the air. It was quite visible and far enough outside the PC so that I could clearly see it Commented Dec 24, 2014 at 22:00