Hi, Indeed. The way I get around this in all my gear, both commercial and diy is to isolate the signal ground from the chassis. I always use the third wire (AC mains ground) and attach it securely to the chassis. The internal circuitry ground is connected to the chassis via a type X2 capacitor and parallel resistor. (some folks prefer a diode bridge) There can be no other connections between the two grounds. This is important. Beware of chassis mounted jacks as they can defeat the protection. What this does is prevent ground loops through the AC mains but allows the chassis to still protect the users. A side benefit is the chassis still can act as an EMI shield for the internal components. This arrangement complies with electrical standards and makes for a quiet piece of gear no matter what it is connected to. Any faults in the circuitry are either handled by the fuse (an absolute necessity) or at least kept from harming the user. The typical component values are between 0.1 and 0.2uf and 100-150 ohms 1-2 watt size is usually fine. The capacitor needs to be AC mains rated thus the X2 designation. Usually the voltage ratings are 275 or 350 VAC. There are a few less common AC mains rated caps but the X2s are easy to source and not costly. Even though many companies used other common types in the past they are not recommended as they are not self healing and can short through. Good discussion.

Hi Pos, That could happen....but to have that occur you have to violate another portion of the codes. The one about all exposed metal parts need to be either double insulated from contact or earth grounded. The failed transformer should be one or the other and then there is no hazard.

The resistor is actually best thought of as a low frequency path for crud between the circuitry and earth. Many designs leave it out. I find that it helps with the S/N a bit. Also the use of a large rectifier (often a bridge) between the two grounds could fail if there is the second grounding violation. I don't care for the rectifiers as they leave the chassis about 0.7 volts different from the earth and 0.7 Volts is a lot of potential noise that is not eliminated.

Just my two cents on how the codes work.

Hi, There are several ways to make the chassis to signal ground connections. Everything from the old school way of using the chassis as the signal ground (usually too noisy for me) to huge bridge rectifiers. Any of them can comply with the various electrical codes as long as the user is protected from accidental contact with a live chassis. How we all lived through the 2 wire AC mains days with tube gear is a miracle. What I find is there is a sort of sweet spot when using an X2 capacitor and a resistor in the 120 to 150 ohm range. It is not designed to protect from faults like lightning etc or really any external faults. That is why the chassis is AC mains earth grounded. It will provide a path to the earth ground if there is an internal fault but that is not the main purpose. It is not really all that good at that as the impedance is fairly high. Its purpose is generally accepted as two fold. One it acts as a ground loop prevention measure when other gear is connected that passes signals to the subject device. The signal ground on the one will not find an alternate path through the AC mains and cause hum. Second it allows the chassis to act as an EMI shield without being in direct connection. IMO your one ohm resistor will comply with the codes....but may not provide as much ground loop hum rejection as is possible. And as nearly everyone knows...I hate hum and noise with a passion and the higher value resistor helps. BTW wattage is not really critical (I use 1 watt ones) as fault protection is not the primary function...that is what the chassis and three wire mains connection is supposed to do. Even if the resistor failed, the user is still protected. Now to be difficult...I could make a case for if the resistor and X2 fail open and an internal circuit fault energized the signal ground and it was connected to either an input or output cable that had a circuit ground conductor and the user was touching the ground conductor and something else that really was grounded there could be a hazard. Such a failure would almost always manifest itself as an anomaly in the gear and require attention. But then that sequence can happen with anything attached to the AC mains like lights, appliances etc. Opinions anyone?