1

Why the power supply plugs of some modern electronic equipment have ground and some other ones -- with the same power, same manufacturer, same purpose -- don't?

For example: I bought 90 Toshiba Portege laptops, all of the same model, about 50 of them were delivered with grounded power supplies and some 40 with no ground.

I am aware of Why would we need to ground an AC source? but it does not address the question above.

Paulo Ney
  • 133
  • 5
  • 2
    what ground are you talking about? – jsotola Feb 01 '20 at 03:06
  • @jsotola the one on the plugs – Paulo Ney Feb 01 '20 at 03:08
  • 3
    Does this answer your question? [Why would we need to ground an AC source?](https://electronics.stackexchange.com/questions/116147/why-would-we-need-to-ground-an-ac-source) – Blair Fonville Feb 01 '20 at 06:26
  • @BlairFonville, I am afraid not. It is a lot of theory around the issue, but does not address the -- pratical -- point of the question. I have bought about 100 laptops from Toshiba, all of the same model - about half of them came with grounded power supplies and half not! – Paulo Ney Feb 01 '20 at 18:57
  • @BlairFonville The linked post doesn't address the question. It deals with AC mains networks. The question asks about SMPSes. – Ariser Feb 15 '20 at 11:46

2 Answers2

2

Here comes the long story.

A SMPS is a rather complex circuit considering the rather simple and common task it is performing. Generating a low noise and constant DC voltage from mains AC voltage.

It consists of several blocks

  • mains filtering
  • power factor correction
  • converter
  • control circuit
  • output filtering

In design of these blocks there are a lot of constraints between them but also a some of degrees of freedom. In the center of a SMPS is the converter which can be designed adhering to different principles (buck, Cuk, SEPIC, just to name some very common ones). Also within a converter of a single type there can be variations in parameters affecting

  • ripple

  • stability

  • overload handling

  • brownout behaviour

To sum it up, building a SMPS delivering 19V DC with 1.82 A, 0.5 V output ripple and at least 0.85 efficiency there are a lot of different designs around with a lot of varying other properties.

A lot of difference lies in EMC behaviour. So for the one design it may be beneficial to ground something inside the SMPS while for another it might be detremential.

Why are there now SMPS with and w/o grounding for a single appliance?

Companies like Toshiba and LG don't design the SMPS themselves, it is not their core business. They source it from more than one supplier to keep their supply chain redundant. Many SMPS manufacturers have specialised in certain converter concepts. They supply "their" concept for the cheapest price and with optimisations at their best.

LG and Toshiba then don't care about uniformity of the SMPS they put into the boxes. They aren't Apple after all.

Ariser
  • 3,846
  • 3
  • 23
  • 43
0

Only PS units that need Y caps in a Pi Filter to bypass EMI to PE gnd will use a 3 pronged plug. This may be necessary to pass CE conducted emissions tests on higher power units.

The DC side is still isolated at line frequency, but will have CM crosstalk at the switching rate + harmonics, that requires attention.

Tony Stewart EE75
  • 1
  • 3
  • 54
  • 182
  • Interesting! And how do you explain laptops by the same manufactures and same series some have and some don't. LG TV's in the same series -- some do and some don't. – Paulo Ney Feb 01 '20 at 03:39
  • ... and not always the highest power unit having the ground pin ... – Paulo Ney Feb 01 '20 at 03:50
  • I explained it in the need for shunt caps to PE. (in addition to the common mode and differential mode chokes used in a Pi filter. – Tony Stewart EE75 Feb 01 '20 at 04:07
  • and how come the same electronics is delivered by a manufacturer like Toshiba - with and without ground in their power supply plugs? How malleable is this "need for shunt caps to PE" that manufactures feel free to use and not! – Paulo Ney Feb 01 '20 at 18:59
  • There are hundreds of ways to convert AC into DC and a wide range of frequencies to choose from 20kHz to several MHz switching rates each which affects choices in magnetics and cost. So it is not an indiscriminate choice, rather a design to absorb or shunt noise. The test results may dictate a need for shunt attenuation. – Tony Stewart EE75 Feb 01 '20 at 19:09