Load Near Earth Ground

The offline converter usually provide the isolation.

The converter doesn't have an isolation transformer. For example, if the earth ground makes contact with either of the converter outputs ... there's a big poof!

D from BC British Columbia Canada.

Then I'd say that there is most likely a serious code violation. The only caps that are allowed between line connections and PE are the X-capacitors and those must usually be rated 4kV for breakdown:

And in applications such as medical you can't even use those.

A stray capacitance that arcs at 350V sounds scary. Probably there is a clearance and/or creepage path violation.

I think so too. I'm looking into that. I'll probably have to change the converter design to use an isolation transformer.

D from BC British Columbia Canada.

Usually, capacitance from (either side of line) to ground comes from transformer primary winding(s) to its core, which *should be* tied to ground. 2.4nF is a bit large, tho..

Then the converter was not designed to any standard and the menufacturer is exposed to all kinds of legal actions including murder.

Yikes...I'm the one who designed and made the converter. :( I'm exploring if I can use a nonisolated converter (no transformer) and pass hi-pot testing. The load has a high surface area in close proximity to earth ground. That's why the breakdown and the capacitance. A believe a hipot test will make the load arc to ground.

D from BC British Columbia Canada.

If you can't redesign the load I think that's you only option.