# USB port and bench power supply ground difference

So I recently coughed up the $250 for a semi decent bench power supply, the thing that got me though is when powering a circuit from this unit and connecting the BusPirate via USB is that the ground levels aren't equal. Measuring the difference between the ground levels with a multimeter gives me a +-0.4V AC signal. I ended up tying the ground levels together to prevent the BusPirate from putting out a voltage greater than 3.3V (relative to the power supply). This works but gives me a fair tingle when plugging out the USB connector and touching the ground sleeve as well as the circuit. I have two questions: 1. Can AC signals this low cause any damage? 2. What precautions should I be taking to prevent damage to myself and components? Here's some additional info: 1. The PSU ground is not connected to mains ground (It does have a mains ground but I'm not using it). 2. When touching the scope ground to the PSU ground (not mains ground) I can see a tiny little spark. 3. The PSU is powering a circuit on a breadboard, I'm using this ground as my one reference. The other reference is from the ground of a USB cable powering the BusPirate (A small piece of test equipment I'm using for USB <-> RS232 3.3V) 4. With the scope ground connected to the PSU ground (not mains ground) and the scope probe on the USB ground I get a 50Hz 3V p-p signal. 5. The USB cable is attached to my laptop which only has two leads. The resistance between the laptop ground at the laptop side and the plug mains ground pin is 800Ohms. • .4V AC could be caused by your multimeter to a great extent if you have a poor multimeter, did you see this with a scope? It is very possible you are picking up 60/50Hz from the power lines in your house, but It is something worth verifying, and the fact that it gives you a tingle is not a great sign either. – Kortuk Jun 5 '12 at 12:46 • Show a diagram of exactly how things are connected. It's not clear now, so it's hard to answer your question. Also show exactly which two points you measured the 400 mV between. – Olin Lathrop Jun 5 '12 at 12:49 • A "semi-decent" bench power supply may or may not be floating. It is important to know this as well. – W5VO Jun 5 '12 at 13:47 ## 2 Answers You have a bench supply and USB supply with AC mains feed. These may have two wire or 3 wire (phase + neutral + ground) leads. The outlets used may or may not have grounds. If both outlets and leads do not provide grounding so that both power supplies have third wire grounded input grounds then the following is likely (based on what you describe and my experiences): - You are very probably getting coupling from mains via a Y capacitor in each lead of mains to input ground. • If you measure the 0.4 VAC with no load and a decent AC meter you will probably see about Vmains/2 or 55 VAC or so in USA. - This capacitor coupled mains voltage can destroy things. • Y caps may be 0.001 or 0.01 uF. "More data needed" including certainly whether PSU ground is connected to mains ground, what a "Bus Pirate" is (Gargoyle will for sure know but I/we ought not need to ask it) etc. BUT some supplies have isolated in and out DC wise and Y capacitors from phase & neutral connected to input ground. If you use a 2 wire power feed and connect in-gnd to out-gnd you get mains/2 capacitively coupled. If you are getting bitten you have >>> 0.4 VAC and probably half mains via Y caps. This is not an answer on why this might be happening, but rather how you might mitigate it. To handle ground-loop problems with a USB peripheral, you can get a USB Isolator like this one for$37.75:

I haven't used this particular device. There are others, like this one but I have no price information (they have a phone number to reauest a quote, which may mean they only sell in bulk).