PSU not suitable for use with metal housing [resolved]?

The PSU shipped with Reform - MeanWell GST60A24-P1J has an annoying problem. When moving hand along the metal surface of the Reform while charging it causes repeated electric shocks.
While this is probably not dangerous it is pretty annoying.

I notice that other notebooks that have metal housing do not have this problem, and some PSUs which come only with two pins on the mains plug do. These were however shipped with appliances that have plastic housing, and do not have this problem because of that.


we don’t experience this issue here so it could depend on the electrical setup in your country or house. On the other hand, I know this issue well from having used aluminum MacBooks in the past here in Germany, where it was very pronounced.

I’m not an expert on AC and power electronics, but could it be that the ground pin in your power outlet is not actually grounded? Maybe someone else here has experienced a similar thing and has some insights?

You should not feel any voltage / tingling sensation with a 3 prong power supply. If you can feel anything, then (as @mntmn suggested) your wall socket is most likely not properly grounded)

In the very same outlet it is not a problem with other notebooks with 3pin power supply.

Still there is the thing that this is not binary, and while the pin may be connected the resistance of the connection may vary depending on the wiring.

Did you get yours from MNT in Europe or via Crowd Supply?

Mine is from Crowd Supply and it does this too. It also puts out some pretty nasty RFI when the Reform is charging. If I run the Reform off a different power supply the shocks and RFI go away.

I got it in the EU. If the type is the same it does not really matter.

Perhaps you could try a different 60W+ power supply with a barrel jack? (under 32V should be fine). Or a beefy USB-C supply with an adapter (there is a thread here on which ones work and which don’t).

Yes, I had a box with old notebook PSUs somewhere, will see if there is anything suitable.

Found an old ‘dy’ power supply. The connector is not an exact match - the inner diameter is bigger, and it does not make a connection when fully inserted. With the indicator led it’s easy to tell when it does connect.

It does not have this problem.

On it the ground pin of the wall plug is connected to the output connector shell, on the supplied PSU it is not.

so it’s using the earth wire as the return path? the outer bit of the output barrel connector should be connected to neutral, not earth.

Definitely not. The external metal surfaces should be connected to earth wire if anything.

A barrel connector only has two metal surfaces. Inside is live therefore the outside has to be neutral. If it goes down earth instead it will trip the RCBO.

Nope. The earth wire is there to define the potential at which the user sits, and if you have a metal surface that’s accessible to the user this is what it should be connected to.

That’s why the 2pin PSUs have this problem, and the 3pin do not when constructed correctly.

I don’t know how you would construct the PSU to not trigger RCBO, and I leave that to PSU experts. It’s been made sometime this century, RCBO is not exactly new, and the certifications it claims take up about half of its label so I suppose it would not trigger it (PN 239427-003).

Supposedly connecting the input ground pin to the output ground should not be a problem for RBCO so long as the PSU contains a transformer separating the input and output.

The datasheet of the MW PSU states that there isn’t a connection, though.

:warning: Don’t open the PSU unless you are familiar with local electrical wiring codes. The voltage inside the PSU is dangerous. You may need some certification to work on electrical equipment depending on local regulations. :warning:

The MW PSU is repairable - there are screws under the feet that can be unscrewed to take it apart.

The input ground can be bridged to output quite easily. There is already a jumper wire going from the input connector to the output side, and a part covered in heatshrink tubing connecting the two (probably a capacitor - did not take it out to investigate). Adding the jumper wire resolves the problem.