Comment by bcaa7f3a8bbc | Hacker News Reader

Comment by bcaa7f3a8bbc | Hacker News Reader

Parent Full thread

The real question is, if the common wisdom is to ground the chassis when possible, and even the induced voltage on the chassis is harmless, they are seen as a widely recognized annoyance. So, my (Question 1) then why does the hardware designers keep using ungrounded design in the first place? Just because they think the output through a isolated transformer is safe enough (also can meet safety regulations), so they want to then omitting the Protective Earthing to simply make the power supply smaller and they think it's okay to give the users some small shocks? Or it's something else?

And here's another question I don't understand. When a piece of electronics don't have a Protective Grounding connection, the common explanation is: the low voltage is created by a SMPS, and an isolation transformer at the output side of the SMPS makes the voltage no longer being Earth-referenced, so in the worst possibly scenario, if a grounded person touches the "hot" output wire, nothing would happen. But if you ground the chassis to the Protective Ground (often the ground of the PCB is connected to the chassis), the output becomes Earth-referenced again and loses the protection of isolation. This is why an isolation transformer has a unconnected Protective Ground at its output side, and many SMPS-based appliances don't use the Protective Ground.

So my Question 2, according to the common wisdom and many regulations, most appliances should be grounded, but from the principle of isolated power supply, it should not be grounded. And in real life, we clearly see lots of laptops use DC power supply without the ground connector, and others with the ground connector. And to further complicated the issue, some DC power supplies only use the ground internally for the EMI filter circuity and/or its own enclosure, but the DC output itself is isolated and is not ground referenced. Then how can an isolated power supply with the ground plug help stopping the induced voltage on a laptop's enclosure, if the output is isolated? Something related to the leakage current of the X/Y capacitors? It seems there's lots of inconsistency here. Just why?

> So, my (Question 1) then why does the hardware designers keep using ungrounded design in the first place?

Ungrounded sockets are still very common (just off the top of my head, in Japan, Philippines, Australia, all through Europe, India...). Even as building codes change it takes many decades for the housing tock to upgrade. Even when the code requires upgrades to commercial buildings it rarely will of residential.

If the leakage is harmless (or "harmless) why risk the user buying a product, getting home, and then being unable to use it. The OOBE is terrible and possibly fatal.

This is also why the "international travel kit" for Apple power supplies contains non-grounded plugs.

The authors final and correct point about Apple is fortunately now obsolete as since it was written Apple has switched to standard Type C connectors. On the other hand they no longer ship grounded connectors by default (in the USA at least).

They do ship grounded connectors still, but it is underneath the molded plastic piece at the bottom of the box which is hard to get out.

That's a false rumor. https://forums.macrumors.com/threads/macbook-pro-16-17-comes...

Hmm. Thanks!

As far as I know Apple does not ship a three-iron plug in the USA except when attached to a cord. They do in some places like the UK.

Apple plugs in the UK have three pins because it's necessary to operate the safety slider in UK sockets.

But (for laptops) the earth pin is not connected internally.

(rummages through Apple travel adaptors): holy cow you're right!

Per the common product safety standards (national standards based on IEC 60950-1, 60335-x, 62368-x), Class I and Class II construction are defined per the level and type of protection from shock for 'grounded' and floating enclosures, respectively.

Per your SMPS reference, the output reference depends on the design intent and intended end-use environment. Depending on the scoped standard, outputs greater than 50Vdc will require a minimum separation from the both exposed dead metal and stuff referenced to an earthing terminal. Where the output is considered safe to to touch, there are only functional spacings required. Also, Class II construction is, in fact, not intended to be 'grounded', and must not be a hazard for a fault condition where the enclosure touches either mains voltage or metal referenced to an earthing terminal.

Class II construction requirements are supposed to mitigate a single fault condition associated with seperation from and contact with a hazardous voltage or power source.

Most potentials between the user and ground are due to current flow through the y caps (noise filter caps rated to between a hazardous voltage level and the user). Touch current, depending on the scoped standard and category and class of equipment is typically limited to less to 5mA (some equipment is allowed 25mA). Class II equipment is typically limited to 250uA. Given the assumed human-body models in IEC60990, the voltage at max allowed current leakage would range from imperceptible to noticable but not effectual.

Shock hazards can exist from improper construction or improper installation and of equipment.

Reply on news.ycombinator.com