Possible earth loop frustration

[pmiller056]

The purpose of double insulation or reinforced insulation, is that the insulation system is so good, that even under gross fault conditions the external surface of the equipment or externally accessible metal parts never can become live, hence no requirement for a safety earth connection. So, in this case a reference earth connection is never required to pass any fault current. In this case any leakage current passed by the reference earth connection is always going to be less than the 5mA noted. The purpose of supplementary equipotential earth bonding in a fixed installation is quite different.

 

Peter

[KevinE]

Kevin - this is not correct

 

Yes it is, that's why I said that. And the proof of the pudding is in the eating...see the OP's outcome.

 

There is debate in the IET forums that a multiple-connected class-2 system in effect forms an extraneous conductor and should therefore have at least one point tied to earth...certainly in a fixed installation it's recommended if not mandated. Each device will pass a certain AC filter current to chassis via its class Y components, as mentioned by another post. The filter current can be a complex waveform and hum and beat interference can result. Sometimes earthing the screen helps, sometimes makes it worse. It's the same with video distribution systems. It only takes enough AC coupling, signal-wise, to neutral (earth pot.) one one item for it to induce noise on the output. A switching power supply may work at 100khz or so, and very little coupling is needed for a loop to occur.

[kgallen]

a multiple-connected class-2 system in effect forms an extraneous conductor and should therefore have at least one point tied to earth

 

Hmm, interesting, I can see how that might be argued from a wiring regs perspective (and hence your point about fixed installs). This would make me want to create a fully-fault current capable earth connection though, rather than a potentially weak one to just the audio screen via the audio connector. Hearing what you wrote earlier, would you maintain that a relatively weak screen connection would be acceptable given this, rather than a fault-capable connection to the class 2 chassis - to which we might expect (hope?!) the audio screen to be connected either directly or capacitively coupled? (I'm not challenging your comments here, just interested in extending the thought process.)

 

Thanks for the info and clarification (KevinE and Peter), I stand corrected! http://www.blue-room.org.uk/public/style_emoticons/default/smile.gif

 

Kevin