Why is it safer for ground to be connected to the chassis for an electronic system?

For example, the ground on the mains supply of an AC motor in a box fan is connected to the outer chassis of the fan?

My only explanation would be that in the event of something accidentally shorting to the chassis, its shorted to ground as opposed to the chassis now being a live voltage?

  • 3
    \$\begingroup\$ Yes, that's the reason. \$\endgroup\$ – endolith Aug 14 '11 at 2:17
  • \$\begingroup\$ But in the event of a dangerous voltage being shorted to ground via the chassis, wouldn't there be a large current flowing to ground via the chassis? \$\endgroup\$ – NickHalden Aug 14 '11 at 3:34
  • 4
    \$\begingroup\$ Which should blow a fuse/breaker somewhere (hopefully) as opposed to being an electrocution hazard. \$\endgroup\$ – W5VO Aug 14 '11 at 4:10

Short Answer

All exposed metal should be "grounded" by connecting it to a properly designed low impedance ground circuit in order to ...

  • Stop fault energy flowing into YOU or to other unintended places. The grounded metal will provide a much lower impedance than you will. Earth potential rise should be lower to much than phase voltage and you should at worst feel a reduced shock for a very short period.

  • Assist fault detection & interuption equipment to work properly (or at all). [eg fuses, circuit breakers, ground fault interrupters/earth leak circuit breakers etc.] A large amount of current, greatly in excess of the usual design maximum, will glow from "phase" to ground and cause the fuse or circuit breaker to interrupt the voltage.

  • If an eath leak circuit breaker / ground fault interrupter is used - allow nearly complete protection against electric shock.

Long(er) Answer:

The main objects in power distribution to "appliances" (let's use that to refer to everything from large fixed wired things to small portable devices) are

  • (1) To make the energy go to where we want it

  • (2) To stop the energy going where we don't want it.

  • (3) To detect when energy goes where it shouldn't and stop it doing so

  • (4) To maximise efficiency of energy transfer.

  • (5) To keep users safe and surroundings free from damage.

5 is arguably a mixture of 2 & 3.

The interconnection and "earthing" of all exposed metals is related to 2 & 3 & 5

Point 2 - Stop it going elsewhere: We don't want energy to leave the appliance - eg into users, into other equipment, into the ground (except a temporarily as part of ensuring 3 & 5. If there is a path to ground via a ground lead from a broken mains wire, or via failed insulation or an incorrectly operated switch etc then the energy cannot get out into the world.

If fault energy flows into a case and then to ground via a low-impedance-by-design via ground lead oath, then it is very unlikely to flow through you. This is usually regarded as 'a good thing' [tm].

Fault energy flows into the case or chassis and then via the ground wire to ground, thereby completing the circuit. What happens then is covered by point 3.

Point 3 - detect and stop fault energy flow. Energy flowing via a ground lead is not meant to be there. An AC mains or "grid" feed to an appliance is designed to be able to provide far more current than it is intended or designed to handle. This is so that the voltage drop will be low under normal use conditions. Ground circuits are also designed to be able to handle in the sort term far more current than the mains circuit is designed to provide. So during a phase (= live wire) to ground or to chassis or to metal case fault, the current flow via phase and back via ground will be far larger than it is designed to be in bnormal use. This means that the fault condition is easily distinguised from normal operation - so that objects with no more brains or capability than a piece of wire - commonly termed a "fuse" - can detect and terminate the flow.

Visitors have arrived .... more soon ...

| improve this answer | |
  • \$\begingroup\$ Ah, but if really crapping timing happened and I was barefoot in soil touching the chassis right as a phase wire shorted to the chassis... then current would have 2 paths to ground (1 via my body and 1 via the chassis back to gnd wire on mains) and I could still be injured before the fuse/breaker triggers? \$\endgroup\$ – NickHalden Aug 14 '11 at 6:18
  • 1
    \$\begingroup\$ You asked why it was done - not whether it was foolproof :-). The object is that ground rise is as small as is reasonably possible, and it will be using the wired ground which is designed to be low impedance, compared to the ground you are standing on and your body contact which may be lowish impedance ... SO you may or may not get a shock and it may or may not be a significant one. I think that if everything was done as intended (pigs may fly) then you would have a good chance of having shock potentials greatly reduced. As ever, your mileage may vary :-). \$\endgroup\$ – Russell McMahon Aug 14 '11 at 8:11
  • 1
    \$\begingroup\$ If an earth leak triggered breaker* was used you would experience little or no shock and would feel little or nothing. * = ELCB / Ground fault interrupter etc. In most cases an ELCB will allow you to touch phase while you are grounded and feel only a single short kick, with the circuit then being safe and dead. I imagine this MAY kill the occasional person but it would be extremely unlucky. \$\endgroup\$ – Russell McMahon Aug 14 '11 at 8:16
  • \$\begingroup\$ Ok cool, got it. \$\endgroup\$ – NickHalden Aug 14 '11 at 16:53

If exposed surfaces weren't connected by low impedance to some common node, it would be entirely possible for the various devices around us to float to different, potentially hazardous potentials. So it makes sense for everything to be tied to something. The most readily available and arguably safest thing for devices to use as a common reference is earth ground.

| improve this answer | |

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.