Bill Whitlock
Mr. Whitlock has presented this topic at recent AES Conventions. As many members are not able to make it to a convention, we were fortunate to have Mr. Whitlock as our guest.
He started the discussion by asking: is all noise picked up by cables, presumably from the air like a radio? To which he responded: this is rarely an issue.
He discussed grounding, which ultimately is a connection to soil. For electronics it has also come to mean the common return path for various circuits. Hence, the meaning of "ground" has become vague, ambiguous and often quite fanciful.
He then discussed ground loops. When ground is at different potentials in different parts of the circuit, there is the likelihood that this voltage difference will cause a current to flow — which is noise for an audio circuit. He discussed various bad practices of breaking the connection of the ground pin in order to circumvent this noise.
He discussed the differences between unbalanced and balanced interconnects. Unbalanced is a poor interface one conductor has zero impedance to ground. In a balanced interface, both conductors have the same impedance to ground. Unbalanced circuits are extremely susceptible to noise. Consumer audio and virtually all video interfaces are unbalanced. Simply put, the terms professional and unbalanced do not belong in the same interface.
Mr. Whitlock spent some time discussion power conditioners. He concluded by saying that if all power supplies were properly grounded we wouldn't need power conditioners. Need to ground power supplies at the supply, not via the circuit of the device.
In a discussion on cables, he said that twisting a cable makes it, essentially, immune to magnetic fields. Star-quad cables uses two pairs of wire to ensure a significant reduction in EMI. However, it is very important to maintain the twist, so at the end of the cable keep the leads to the connector as short as possible so as not to break the twist. With a star-quad cable, even at a few inches from a power cable will result in almost no EMI pickup (assuming equal power on each conductor). Essentially as the distance increases from the power cable, the opposite fields cancel.
He also discussed shield current induced noise. Any shielded cable with a drain wire is roughly 100 times worse than one with a braided shield. The drain wire takes all the current and it is not randomized, hence is noisy. This can be avoided by only grounding one end (driver end). This is also solved by the "Hybrid ground".
He illustrated various troubleshooting techniques using dummy devices.
