Grounding Output of AC/DC PSU

[ed_reardon]

Hello all,

In the work life I've been handed an issue with an instrument.

The instrument is powered by a (transformer) isolated DIN rail switch-mode PSU and it's been reported many of them are floating well above ground (sometimes around 80VAC) - my initial instinct is to tie 0V PSU output to mains referenced earth, which has been tried on test equipment and certainly stops the floating issue (and the odd nibble from the damn things),  as it is expected to.

However, with isolated PSUs (they're not my area!) is there anything inherently wrong with doing this?  They're clearly no longer entirely separated but as the whole installation is well earthed so that isn't an especially great concern to me.

As an aside, and I've spent the morning trying to work though out,  these instruments are measuring from sensitive electrodes in to instrument amplifier circuits,  the input (and subsequently output) appears much more stable with the PSU 0V to Ground, which is the opposite of what I expected?! Am I going mad or is this plausible?

[Jeroen3]

Every isolated AC/DC converter has leakage current. This leakage current will find a way.
Through your sensors perhaps? Messing up the measurement.

In general, any DC bus in an system should be grounded, unless there is a specific requirement not to. Low leakage power supplies should be used in that case.

[Relayer]

Hello ed_reardon,
If you're talking about the primary of the power supply, then you cannot
connect earth directly to it. It is designed to float.
Does the mains plug have an earth pin?
If not, then to reduce the tickles you may receive when touching any of
the metal parts, is to replace the mains cord with a 3 pin one that has
an earth wire within it, then place the active and neutral wires in the
normal place, then connect the earth wire to a 22nF 275V X2 capacitor
(0.022μF), and the other side of the cap to the earthed metal work.
This should stop or greatly reduce any zaps you may encounter.
Regards,
Relayer