Power supply GND at different potential to scope gnd

[Sparky49]

Hi

I've recently purchased a Rhom    BM92A12MWV-EVK-001 Power delivery dev board for examination. I also bought an Inatech 45W USB power supply which has PD3 capabilities. The Inatech has a plastic earth pin (in the UK) and is double insulated. When I connect the Rhom board to the Inatech, I measure open between the live and the output +V and gnd, and also with the neutral to +V and gnd.

However, when I attempt to examine the output with my scope, I notice that a tiny spark appears when I connect the gnd lead of the scope probe to gnd on the dev board. This seems quite bad, what am I missing here?

I've done some tests and the scope seems to have survived, but I'd rather double check what is causing this before continuing.

[Ian.M]

That's pretty much usual for non-grounded SMPSUs.   See https://www.eevblog.com/forum/chat/does-this-sound-acceptable-for-an-apple-product/ for an explanation

[Sparky49]

Thanks Ian, I have to confess I'm still a little unsure about whether I should be able to probe.

On one hand the scope gnd is connected to earth which is isolated from the potential on the circuit gnd, but on the other hand there is enough potential difference to create spark between the two. :/

[Ian.M]

First check the leakage current to ground of the USB PSU is within the permitted limits for a double-insulated item of IT equipment (IIRC that's 0.75mA max.) - I described this test in the other topic. 

Then hook it up to the board under test and ground the circuit's 0V rail to the same mains ground as the scope using a separate wire securely connected at both ends *before* you plug in the mains supply to the USB PSU.  Do *NOT* disconnect the extra  ground wire while bench testing if the PSU still has mains connected.

That will safely ground out the leakage current and eliminate the risk to the circuit and scope.

[soldar]

Floating PSU's can be a PITA sometimes with this kind of problem. I have modified many so as to ground the negative and that avoids this kind of problem. I have seen many PSU's and other appliances which have a small capacitive leak and I just add a ground to remove the problem.  I have grounded many devices that came with no ground.

These parasitic leaks can cause other problems. If the residence has GFCI's fitted in the individual outlets this is usually not a problem as it will not trip them but if the residence has a single differential breaker for the entire residence it can happen that a lot of capacitive leaks add up and the breaker trips very easily and it is a nuisance.