[Al_The_Man]

Actually the theory behind the old established practice of only grounding one end of a shielded cable is to prevent any ground loop that may be present in different parts of a machine and another to impose a field into the shielded conductors. IOW one end of the cable GND may differ from the GND potential of the other, thereby inducing a voltage in the shield if grounded at both ends.
The now common practice of earth grounding a machine with sensitive electronic equipment is to ground bond all metalic parts of the machine, known as Equi-Potential bonding, this essentially eliminates any difference of potential between grounded points, allowing the new practice of grounding any and all shielded cables at both ends.
As per the Siemens paper on equi-potential bonding - section 3.

[mactec54]

Actually the theory behind the old established practice of only grounding one end of a shielded cable is to prevent any ground loop that may be present in different parts of a machine and another to impose a field into the shielded conductors. IOW one end of the cable GND may differ from the GND potential of the other, thereby inducing a voltage in the shield if grounded at both ends.
The now common practice of earth grounding a machine with sensitive electronic equipment is to ground bond all metalic parts of the machine, known as Equi-Potential bonding, this essentially eliminates any difference of potential between grounded points, allowing the new practice of grounding any and all shielded cables at both ends.
As per the Siemens paper on equi-potential bonding - section 3.

We are talking motor cables if a Shielded Cable is used, this has nothing to do with whether there whole system is bonded or not, a Ground loop can not be formed this is standard industrial practice and a requirement in most countries

Analog circuits using a shielded cable are connected at one end only even if everything is bonded