My educated guess is that it would be biologically hazardous for the potential difference (voltage) between signal ground and the chassis to rise too high.
Imagine plugging in a RCA plug and getting a nice shock because the audio ground is 150 volts higher/lower than the chassis ground potential.
How does the audio ground come to be 150 volts higher or lower than the chassis ground if neither the power supply or signal grounds are tied to the chassis?
se
If you want your family to stay safe you have to design equipment on the basis that a mains live connection will break loose (or mains transformer insulation breaks down) and will then be long enough to touch something inside the amp which is connected to something outside the amp which you or one of your children is touching at the same time. This may seem unlikely, but it happens. Grounding all external metalwork means that fuses blow instead of people getting killed.
Grounding all external metalwork can introduce hum loops, so ground breakers will break the hum loop while maintaining safety.
Grounding all external metalwork can introduce hum loops, so ground breakers will break the hum loop while maintaining safety.
Yes, but there is plenty of consumer audio gear out there which meets Class II specs and doesn't even use the mains safety ground.
And if there's no way for the AC mains hot to come into contact with the power supply/signal ground or signal output, i.e. if no AC mains hot lead was long enough to reach any of those points, then tying the power supply/signal grounds to the chassis would provide no additional safety benefit.
There is however at least one reason to tie the power supply/signal ground to the chassis other than safety, and that's so that the chassis can be more effective at RFI shielding. And if you tie it to the chassis through a pair of diodes, you're not going to get any advantage there as you'll never likely have 0.7 volts of RFI between the chassis and power supply/signal ground.
So if there's no chance of the AC mains hot contacting power supply/signal ground or signal out, but you still have the chassis tied to the AC safety ground, then the better way of tying the power supply/signal ground to the chassis is via a series RC network (say 50 ohms and 0.01uF).
That provides a high impedance at audio frequencies and helps reduce ground loop noise, but a low impedance at radio frequencies so the chassis can better serve as a shield.
se
This is getting pretty far OT and covered by a million other threads already, but...
Such devices typically tie post transformer GND (0V) to the chassis, then fuse the transformer secondaries. Device failure and somehow shorting to chassis would cause the fuse to blow. Tying the audio GND (common) to the chassis then also protects the device inputs/outputs from conducting higher voltages.
Such devices typically tie post transformer GND (0V) to the chassis, then fuse the transformer secondaries. Device failure and somehow shorting to chassis would cause the fuse to blow. Tying the audio GND (common) to the chassis then also protects the device inputs/outputs from conducting higher voltages.
But meeting all the Class II specs is not an easy task.
Then you have to have the unit tested by an accredited agency.
I place a 4.7 - 10 ohm 1/2W - 1W resistor in parallel with the diodes which has always proven sufficient as a ground loop breaker and doesn't leave the audio gnd floating. In some cases I add a small ceramic cap to assure the chassis and system ground are close to the same RF potential. Further in such instances I will also use small ceramic caps (typically 0.01uF) from the RCA ground buses to the chassis.
Last edited:
And shielding. With the metal chassis grounded the field lines that terminate on the chassis set up a minuscule ground current. With the chassis grounded, these currents will flow to ground and not into sensitive circuits.
~Tom
That wasn't my point.
Class II gear doesn't use the AC mains safety ground. Therefore, if the AC mains hot were to come into contact with the chassis, there's no return path back to neutral, the line fuse wouldn't blow and the chassis remains hot whether or not you have the power supply/signal ground tied to the chassis.
se
Why would they flow to ground?
Shielding works in two ways, reflection loss and absorption loss. With reflection loss, the impedance mismatch between the interfering wave and the chassis causes the interfering wave to reflect off the surface, just like a signal in an improperly terminated transmission line gets reflected back to the source.
Absorption loss is due to eddy current losses in the chassis' metal.
se
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.