Audio Lead Screening

[ThermalRunaway]

Hi everyone. I came across a concept today which I was unfamiliar with so I've been trying to learn more about it. My confusion is based around screening techniques for Audio Leads.

I have been fault-finding some military communications equipment at work, and I found that a number of the units had failed test due to slightly low sidetone levels. I decided to retest the units to observe the failure for myself, and I found that actually the units would pass provided that you allowed them to cycle on the sidetone test for a few times. I also found that the sidetone measurement, taken in Decibels, would fluctuate by quite a high margain so that a pass or fail depended on how the audio analyser caught the measurement. I put passed units back on the test and found that they exhibited the same symptoms.

I decided that there must be some form of problem with the test gear or interconnecting leads to be causing that kind of fluctuation in sidetone measurement, so I set about tracing the problem. I came across a 5-pin comms leads which I felt had faulty cable screening. The cable itself was screened, but having done a continuity test between the shields of the two connectors (one at each end of the lead) I found that the screen was o/c.

I happened to mention this to one of my work colleagues, and he said that the lead probably was not faulty and that it could quite well have been designed that way. He said that quite often, in audio systems, only one end of the lead has the screen connected because this helps to avoid ground loops.

Not being one to take anyone's word on anything, I decided to try and find out more for myself but I've done a bit of internet searching and I haven't turned up much of anything.

Is anyone else familiar with this?

Brian

 

[ThermalRunaway]

Ok I tried my old friend ePanorama and came up with this, if anyone else is interested:

https://www.epanorama.net/documents/groundloop/cable_shielding.html

 

[ThermalRunaway]

I've done some more research on this, and I'm getting the general impression that ground loops are only really a problem in audio systems where the ground for the mains connection forms the ground for the audio signal (because of the earthing on the casing). But the military comms stuff I've been looking at is battery powered.

Hmmmm....

Brian

 

[Nigel Goodwin]

While it's true that ground loops are mostly a problem for mains powered equipment, the same problem can occur with battery powered equipment as well.

Basically you should only have one ground (or chassis) connection between different items - so assume you had two battery powered audio units, with a screened cable between them. Assuming they share a common power supply?, the negative and positive wires are connected together, negative is usually ground/chassis. So if the screened audio lead is connected at both ends, this means you have TWO negative connections between the units - this can cause currents to flow which can be induced into the audio path.

The simple answer is to break ONE of the ground paths, either the negative wire (leaving the screen to carry the power), or the screen itself (at one end), as it no longer needs to be connected at both ends.

 

[ThermalRunaway]

Ok, so he might have been right about that then. But wouldn't an audio lead with the screen only connected at one end be more prone to noise?

Brian

 

[audioguru]

Ok, so he might have been right about that then. But wouldn't an audio lead with the screen only connected at one end be more prone to noise?

I have avoided ground loops and noise pickup on circuits inside a metal box by connecting the cable's shield only at the receiving end.

 

[Blueteeth]

hi,

Ok, so he might have been right about that then. But wouldn't an audio lead with the screen only connected at one end be more prone to noise?

Not an expert, but I would say no, it wouldn't be more prone to noise, unless perhaps, if the cable in quesiton was of significant length (>100m?). Sheilding in cables, in many instances also acts as 'ground return', like most cheap consumer audio cables (core = signal, screen = GND). However, as nigel correctly pointed out, if both 'ends' of the circuit that the cable is connecting already share a common ground, then the screen need not be used a return path, it is there only to shield cable signals from EMI (and to some extent, prevent the signal from *creating* EMI). All that is required is that it is electrically connected to the circuits 'GND', at either end.

Blueteeth.

 

[Paul Obrien]

Another look at ground loops https://www.electro-tech-online.com/custompdfs/2006/07/humloop.pdf

 

[Nigel Goodwin]

Ok, so he might have been right about that then. But wouldn't an audio lead with the screen only connected at one end be more prone to noise?

No, it's still completely screened exactly as before - your scope probe generally works like that (assuming you use a seperate earth lead like most scope users do?).