13
\$\begingroup\$

A basic nugget of info is missing in my end-to-end picture of how grounding works and why it's important. When a voltage is applied in a circuit the electric current starts flowing (or the field establishes itself). Now in an AC home circuit the current flows through the circuit like in DC, but happens to reverse directions 50 or 60 times a second too (Hz).

So why do some appliances have electric current leak onto their metallic surfaces at all in the first place. Should'nt all appliances's internals be designed so that there can never (or rarely) be any leakage of current ?

The whole point of my question is that why do we blame the lack of grounding when an appliance gives an electric shock - isn't the appliance equally to blame for being designed in a way that allows for charge leakage ?

Hence in the event of an electric shock isn't it equally important to investigate the appliance (in this case it's actually a custom assembled desktop computer) to find out why its circuit is leaking charge to its metallic body parts, instead of always expecting the grounding to remove that excess charge to the earth.

Another way to paraphrase this question is - are some appliances (especially assembled computers) likely/expected to leak charge. Hence in the event of rare shocks sometimes shouldn't it be more important to investigate the appliance itself for having the proclivity to receive leaked charge instead of blindly checking the grounding

\$\endgroup\$
  • 1
    \$\begingroup\$ Appliances fall into several classes. Class I appliances have their chassis grounded (earthed). Class II appliances are double insulated. \$\endgroup\$ – Nick Alexeev Mar 12 '13 at 3:07
9
\$\begingroup\$

It is not unusual for the mains supply to be deliberately connected to ground via small high voltage capacitors, to reduce emitted radio interference. These capacitors are rated to withstand high voltages safely, and to "fail safe" (i.e. not form short circuits in the event of an accident or excessive temperature.) They are usually identified as "Class Y" or "Class X2" on their case markings, typically 0.1 uF 275V or 400V.

These will conduct a small AC current to a metal case, and if the metal case is NOT correctly earthed, it is possible to get a mild shock from this current, but it should not be dangerous.

I have also measured about 110V AC on exposed metalwork simply from the capacitance within a (230V) mains transformer (the short circuit current was only 30 microamps but the "tingle" could be felt)

I would however agree that any other source of leakage from AC mains to metalwork should be investigated - the dangerous ones would usually show up with DC resistance measurements, unlike the above.

\$\endgroup\$
  • 1
    \$\begingroup\$ Thanks - this was new info to me. So basically if there is more than very mild current on rare occasions detected on a desktop computer - its cause for investigation. What's bothering me is how could new components from decent OEMs (Cooler Master, Asus) leak current to the cabinet. In a computer all cables inside are well insulated and all pins are deep inside plastic covers. The 600W PSU is enclosed and though it powers the motherboard, the motherboard itself cannot be touching the inside of the cabinet though any mount-point which isn't well insulated.. I just dont get why.. \$\endgroup\$ – PKM Mar 12 '13 at 4:19
  • 1
    \$\begingroup\$ Components like PSUs leak current (AC current) to their earth (ground) connections through the EMC filter capacitors. And their ground connections are connected directly to the chassis metalwork. There is nothing rare about it : this connection is a legal requirement unless the alternative "double-insulated" design practice is followed. \$\endgroup\$ – Brian Drummond Mar 12 '13 at 9:39
  • \$\begingroup\$ BTW -- you should drop the ref to 100nF caps and X2 caps -- it's only the Y caps that go line to earth, and those are about 5-10x smaller than typical X2 caps \$\endgroup\$ – ThreePhaseEel Nov 4 '16 at 22:36
4
\$\begingroup\$

The appliance's case may become hot == connected to live wire due to a design mistake or improper use (dropping it on the floor). Those things happen, much as any and all software has bugs. It would be nice if such errors do NOT cost human lives. For that reason, we ground the case, and if a short occurs, the excessive current travels to ground, circuit breaker trips (or better, the residual-current device trips), and no one is hurt.

To clarify: Charge is NOT supposed to leak to ground. Any such event means that the appliance is defective and has to be repaired or exchanged. Interestingly, the median current, required to kill a human 30 ma, is also the standard value for residual current devices to trip.

Now, why does current flow through a person, connecting the appliance's case to ground? Why not insulate all power supplies from ground, and then it would be not possible to close a circuit through the person, touching the live case?

FIXMEUP:
Unfortunately, I am not sure. I think this is because Earth has significant capacitance and before it is charged enough for current flow to stop, the person will be long dead.

\$\endgroup\$
1
\$\begingroup\$

There are two major sources of mains leakage in a properly functioning Class I (earthed chassis) appliance: deliberate capacitive connections from mains to earth and stray capacitance.

First, and foremost in the case of most consumer/light-commercial grade Class I equipment (desktop PCs, mains-powered test equipment), are the class Y capacitors, often around 4.7-10nF or so, from mains to earth at the power input. These provide a path that allows internal high frequency noise to return to its origin instead of entering or exiting the box -- along with the rest of the mains entry filter parts, they provide a "noise firewall" that keeps your box from making a hash of your favorite radio station.

For medical equipment, where low leakage is paramount, and older appliances, where noise filtering of this nature isn't needed, however, these capacitors are not present. As a result, the primary leakage sources are now parasitic, or "stray" capacitances, from mains wiring to earthed metal and also between the two sides of the mains transformer if one is present and the secondary is earthed. These capacitances are smaller than Y capacitors in most cases, but can still provide a bit of leakage current, especially for appliances that have large AC motors or the likes in them.

Newer appliances, with their sophisticated controls, and some other devices (such as microwaves) are a balance between the two sources of leakage current.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.