Gophert NPS-1601 SMPS earth ground reference

[VSV_electron]

Hello!

All who read this of course know about the popular Asian SMPS, also known as a "poor man's lab bench power supply" (which is not, of course).

The problem with the Gophert NPS-1601 as I see it - it is not Earth Ground referenced. There are only two output terminals ("+" and "-") on the front panel and they are completely isolated so that the secondary of PS is floating.

The aluminum case of the power supply is earth ground connected but that's about the operator safety.
However since there is NO earth ground terminal, using the PS "as is" in the floating mode is prone to all kinds of leakage current effects in the circuits connected to this PS causing damage to the circuits in certain conditions.

Is there a reason the manufacturer of the Gophert NPS-1601 has not provided the earth ground terminal and is there anything wrong with adding the earth ground terminal to the front panel of the PS and connecting the "-" to the earth ground?
Otherwise, going the easy way - is there anything wrong with connecting the "-" to the ground earth internally through a simple mechanical switch (to have it earth grounded on demand) ?

Just for reference:

[VSV_electron]

Another relevant question as a quick follow up: As a general rule is that a bad/good practice earth ground referencing SMPS power supplies when using them in bench set-up with the semiconductor circuits?

For instance I have a bunch of the well engineered and respected Mean Well compact 12 V and 5V SMPS units such as this one, LRS-50-12:
https://www.meanwell-web.com/en-gb/ac-dc-single-output-enclosed-power-supply-lrs--50--12

It has the Earth Ground terminal explicitly provided on the terminal block. Can I safely connect the common earth ground to the secondary "-" terminal of the Mean Well LRS-50-12 ? This in effect will make the secondary "-" connected to the chassis of LRS-50-12 and to the mains ground of course when I provide it.

I tried to find any comments here on the forum or elsewhere regarding ground referencing the secondary of the SMPS as a general rule and my search failed.

[electr_peter]

Hi,

The problem with the Gophert NPS-1601 as I see it - it is not Earth Ground referenced. There are only two output terminals ("+" and "-") on the front panel and they are completely isolated so that the secondary of PS is floating.

Isolated DC output is desirable feature in a lab PSU as it allows easy series/parallel connections with possibility to add earth reference when needed.

Gophert NPS-1601 is a low cost and compact size adjustable DC power supply - there simply is not enough space to cleanly integrate third earth terminal. Quality overall is OK for the price (but not top notch), but terminals on Gophert PSU's are of low quality already. Also, in lower cost segment two output terminals are common. Don't expect too much from this price point in noise/leakage/connectors.

[VSV_electron]

To sum it up - is that a normal practice to earth ground reference the secondary "-" of SMPS power supplies?

I ask because I could be missing some fundamental knowledge again as it happened when I connected the floating SMPS to a GM328 transistor tester and consequently destroyed a few Power MOSFETs because of the SMPS leakage current through the transistor and into the mains earth via my ESD wrist strap. Who would imagine that!?

Since then I'm very aware of the SMPS leakage current and while still aiming at using SMPS for powering up the semiconductor circuits I now want to prevent the leakage current - hence the idea of ground referencing the SMPS to avoid all kings of possible damage from the SMPS leakage current to sensitive circuits and devices under test.

[thm_w]

You probably posted already, but, how much AC voltage and current do you measure from output to earth?

I guess Gophert is only good as a high power supply and not a low power/precise lab supply. I have some meanwell supplies that have little leakage to earth (2VAC at <1mA). So it must be possible to design them.

100% of the time I am working on something the PSU will not be earthed. Though of course if you add in a oscilloscope or function generator, then that will earth your circuit.
If its something sensitive/low current I'd use a linear supply. Almost any linear high quality supply will have little leakage (https://www.eevblog.com/forum/beginners/ac-leakage-does-it-matter/).

[VSV_electron]

You probably posted already, but, how much AC voltage and current do you measure from output to earth?
...

Thanks for the link on the leakage current. I found a few more.

Is there a proper technique/set-up for measuring the SMPS (or any PS for that matter) leakage current? When you refer to VAC measurement related to the leakage current, how to measure it properly and between what points?

[Phil1977]

I never heard about any written test standard or similar.

But to get reasonable results, just measure between the floating output and PE. It makes sense to measure as well the AC voltage as the AC current. Typical values are around half of mains voltage and below 1mA of current. DC leakage current component would be at least strange. It should not matter much if you measure between (+) or (-) to PE.

As it was already written, having a floating output is an important quality of a bench power supply. Anyhow if you do not need the galvanic isolation then it often is good to typically connect the (-)-pole to PE. It often reduces interferences and it enhances your safety - if e.g. the insulation in the PS fails and mains voltage makes its way to the output then an earthed DC output protects you and everything connected.