1
\$\begingroup\$

1) can someone clarify the differences between safety ground and technical ground? Why or why wouldn't we always have both? I'm thinking mostly in a North American context.

2) Are there best practices for rack grounding/bonding? I sometimes see AV equipment racks with large ground lugs bonding each one together and then terminating to a ground bus. I've seen plenty of systems that didn't incorporate this and worked just fine, too. So what's the safest and most technically sound approach moving forward?

\$\endgroup\$
  • \$\begingroup\$ Neither is there one that fits all, nor free lunch, its all complex and depends on how the systems are designed. "technical" GND is just an arbitrary point where we reference all voltages of a system to. \$\endgroup\$ – PlasmaHH Nov 6 '17 at 10:05
  • \$\begingroup\$ Pat Brown makes it into everything :) \$\endgroup\$ – inbinder Nov 6 '17 at 10:09
3
\$\begingroup\$

Religious issues ahead!

The basic idea of a technical ground is to keep the power frequency leakage current away from the connections serving as a reference for the gear in the racks. IT DOES NOT WORK, more precisely it does not work at any frequency much above a few hundred Hz, which was mostly ok (If delicate and prone to go wrong), sort of, in the 1950's and 60's era studios where radio transmitters were usually not located in the same building right next to the racks and microphones, but today everyone carries cell phones, switch mode power in the low ultrasonic region is a thing, and there is a LOT more random RF around.

There are tradeoffs, but as close as it gets to the right answer is probably to use a 'mesh grounding' scheme, along with careful attention to what you do with cable screens (Hint, telescoping one end is usually NOT the answer), and sometimes those big parallel earth conductors between the racks (But it is important that these cables follow the same route the signal lines take).

The key realisation is that modest amounts of circulating current do NOT matter, it is the voltage developed by those currents across the network impedance that can cause issues, so the answer is to minimise this impedance, tie everything to everything else, at multiple points with fat cables, screens, racks, building metalwork, conduit, bond the lot!

The remaining problems are usually gear that has the 'Pin 1 problem', either fix it or get a bit sneaky, and fit a common bonding panel to the back of the rack, with the screens for any circuits involving such gear bonded to the mesh there instead of via the equipment pin 1. All cables entering the rack (as well as any other bonding) should have screens bonded to this input panel.

Computer networking is often less critical as it is essentially all transformer coupled.

Tony Walderon when he was at Cadac had much to write on the subject, and AES48 is useful as wellDesigning an interference free audio system Pt 1 Pt 2.

AES X-13 is worth a look as well.

\$\endgroup\$
  • \$\begingroup\$ --Barring Pin 1 problems, (i haven't run into one in a while) you think the answer in the end is --"It depends?" \$\endgroup\$ – inbinder Nov 6 '17 at 13:14
  • 2
    \$\begingroup\$ All engineering is a trade off, but a mesh bonding network works at least as well as anything else, and has the virtue of actually making sense from a basic physics and electromagnetism perspective. \$\endgroup\$ – Dan Mills Nov 6 '17 at 13:20

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.