Safety question concerning ground loops

[derMichi]

Hi all,

I have built a subwoofer amplifier based on an lm 3886 with an 24 dB/octave lowpass in front of it. As a source, I use the line-out from my main amp.

Until this morning, I had massive hum problems, caused from a ground loop formed by the two amplifiers and the connection inbetween them.

Now I had to break this ground loop somewhere. Some people suggested to isolate the transformer from the rest of the amplifier, but I had a simpler idea:
I simply cut the connection from the lm3886-amlifier ground to the chassis ground and tadaa! the hum was gone, both amplifiers are dead quiet now.

Now my question to the experts out there: Is this acceptable from a safety point of view? I think it should be, but then, I'm no electrical engineer but a physicist - for us, anything that works is acceptable ;-), we usually prefer the quick-and-dirty-solution.

thank you for any advice!

[eVITAERC]

Try soldering the connection back together through a 10 ohm 2W resistor paralleled with a 0.1uF 16V-ish capacitor. That should null the effects of the ground loop as well as maintain the safety and shielding benefits of the ground.

[derMichi]

could be worth a try.

Although I don't see a safety problem in this setup - the chassis is still grounded via the earth connection, only the low-voltage side of the amplifier is floating. but then, the amplifier does still have an earth connection from the shielded input cables. Correct me if I'm wrong- I'm always eager to learn!

[derMichi]

Sorry for the spamming ;-)

I just found a reason for the earth connection of the internal electronics via a resistor and paralleled capacity and thought, I could as well post it here, as others might have the same problem.

In case of a transformer breakdown, meaning electrical contact between primary and secondary windings, one could have mains voltage on the secondary side, which in turn means, that the secondary side of the circuit MUST be earthed.

And if it is earthed via the above mentioned resistor/capacity combination, the hum induced by the then formed ground loop should be small enough to be inaudible.

There is an excellent article about earthing on the site of Rod Elliot.

[richie00boy]

The lack of 0V to chassis connection will be to the detriment of RFI shielding.

[richie00boy]

Originally posted by derMichi
Sorry for the spamming ;-)

I just found a reason for the earth connection of the internal electronics via a resistor and paralleled capacity and thought, I could as well post it here, as others might have the same problem.

In case of a transformer breakdown, meaning electrical contact between primary and secondary windings, one could have mains voltage on the secondary side, which in turn means, that the secondary side of the circuit MUST be earthed.

And if it is earthed via the above mentioned resistor/capacity combination, the hum induced by the then formed ground loop should be small enough to be inaudible.

There is an excellent article about earthing on the site of Rod Elliot.

That is true in cases of equipment with mains safety earth connection and class-I construction. If you can comply with class-II construction regulations with appropriate insulation and creepage and clearances, you can forgo the mains safety earth (thus no possibility of ground loop) yet still connect case to 0V. This is how the majority of consumer equipment is made. However this is not really suitable for racked equipment as then the cases are galvanically connected so potential for earth loops arises again.