"Real-world" shielding for audio

[jamsession]

I got some 24 conductor (12 twisted pairs) cable here that has an overall foil shield and braid, but not individually foil or braid shielded. If I use that for a school project I'm working on (low-budget, but I want to do a time and resources-balanced job for a production of "much ado about nothing" next week), and have 6-18" fanouts on them, do I need to wrap each in foil and add braid on those 12-18" fanout ends?

I know it's dependant on the amount of EMI, RFI energy in the area (audience cell phones, wireless mics, interplanetary alien communications, sunspots, etc etc.)

AND it’s a _bugger_ to rework if I get it wrong

AND that's why they make individually shielded snake cable (does your snake cable have separate foil and braid also? Most don't for size considerations, it looks like individual foil and overall foil/braid is common?

But I'm looking for real-world advice - if I'd probably be ok with the fanout ends exposed 1 foot if the rest was overall shielded well

Do you have any references for proper real-world shielding? It’s a fascinating field of study. (Kind of like making CAT5 lan cable, the specs say it's ok to unwind the twisted pair for 3/4" or so when adding the connectors and still stay in spec.) So all shielded cables are unshielded at the XLR pins for 1/2" or so, and curious how the physics work, I know it's a good better best scenario, and hum is a B&^TCH, but individually foil wrapping and braiding 18" and shrinkwrap over them is a lot of work too. I have this otherwise great cable and will use it on this job since I'm donating my time and labor for the most part, just want to get it done as well as reasonable in the time I have, and put the right amount of effort into it, if this cable should work ok.

This is for balanced XLR runs. If I was running unbalanced, shielding would be even more important, right? (that's why we DO balanced to remove at least certain types of interference). I wish I could see pictures of shielding in action, esp. shielding effects of balanced vs unbalanced lines, and diff types of interference. EG: How much intererference would get into 1" unshielded cable, 6", 18", 20'. Plus points to anyone who shares links to that!

fyi - whirlwindusa.com has lots of neat boxes - isolators and balancers, would love to hear experiences with those or similar products.

Sub-Question:
(would welcome stories of sources of hum not including your humming, "I'm on the top of the world" on the way to work. :^) and solutions to fix them.

My recent hum story:
Got the stage all miked proper, and when the wonderful stage hands put the podium out for the keynote speaker, with the AC powered green desk lamp out there (aquired an hour before the show), desk mic stand with an SM58 pointed right at it... you can see where this is going. Yep, picked up a beautiful 60 hz hum. Now I get to help the videographer take that out in post. Since he and the client didnt' notice it, I wonder if I should bother. Uggh. another thing to add to the preshow checklist. Anyone use LED podium lights? the green desk lamp was so bright through the green plastic "lampshade" that it threw off the exposure from the speakers face. another Uggh. that lamp is history as soon as I find a better alternative. I could put an LED bar in there. Gotta love LED lighting.

Cheers,
Jam

 

[BNBSound]

Making a balanced snake out of twisted pair multicable isn't the best route, but it can be done. An audio snake will have the ground conductor braided around the signal conductors for the entire length. In addition to providing the all important ground reference for the mixer, it acts as a barrier to noise and shunts it to earth when properly connected. Your 12 pair snake will actually become an 8 pair and 8 ground snake, and the grounds won't be shielding the conductors. If the whole thing is shielded with foil, you should ground the foil at one end.

The finished piece will still be susceptible to noise to a certain extent, but the true noise canceling properties of a balanced line are in the pair of signal conductors. One carries the signal in its original polarity, and the other carries the same signal in the opposite polarity. When they arrive at the mixer, the first thing the two signals see is a differential amplifier, the output of which is: the difference with respect to ground of the two signals.

This means that while one signal is cruising along right side up, and the other is flying along in tandem up side down, if a noise spike gets impressed upon the signal it gets impressed upon both of them the same way. That is to say that while the positive side has a positive spike in it, so does the negative spike. When they reach the diff amp (or transformer, which acts differently but is functionally the same for our purposes here) The negative noise spike cancels out the positive spike and only the audio signal goes on through the mixer.

Using balanced cable is more than half the battle in a sound system. If there's still hum in it after all the cables are switched to balanced lines, any remaining noise will be from ground loops, which can generally be carefully removed. This does not mean that you should use a ground lifter on any of the electrical cords. Don't do it, it's illegal and just plain stupid. Many audio devices like DI boxes have ground lifts, most split snakes have them as well. You can make one by making up a short mic cable (and labeling it) with the ground disconnected at one connector.

60 Hz hum gets into the system when, for example, the mixer is plugged into a circuit in the booth that's on leg A of a typical three phase system, and the amps are plugged into let's say leg B. Electrical noise can be drastically reduced by making sure all elements of a sound system are on the same leg. With my portable distro, I've had all my sound gear on one side, some seriously old SCR dimmer packs on the other side, and my amps sitting directly next to them. The dimmer noise was so low as to be nonexsistant.

 

[jamsession]

Thanks - it does have a foil and braid around the whole outside - I understand it's not ideal, but since it's got an overall braid and shield, I think it will work ok.

making an 8 channel snake, I unwrap 4 of the pairs and add one wire to the remaining 8 pairs, for the balanced XLR connections.

I know, it's like the guy who didnt buy shoes and complains that his feet hurt. I've made snake out of this cable before, and it worked fine. If I was trucking it to mobile locations I'd want the max shielded route. but for this install, performance next week, have to balance time and expense factors with performance, I think should be ok, but I'm wondering if I need to wrap foil and add braid to the 18" fanout ends. Wouldnt hurt, it's more work and curious from a real-world scenario if its needed.

"probably not until you hear a hum", and then you swap it for a whole 'nother snake. :^)
but until then, touch wood and you saved them a few hundred dollars. I think that's a fair approach. but still want to do the best I can with what I've got.

thanks

 

[Chris15]

Remember that you will need to have aground connection for each channel at the very least at the female end if you want to use phantom power... Also, I'd always prefer more than the length of tail you were talking about... Because when patching, length can be an issue....

Also the comment about shielding in the connectors is somewhat questionable. Most connectors worth speaking of have a metal body which then connects to mains earth via the case of the mixer or whatever.

 

[David Ashton]

I did my apprenticeship as a measurement tech in a high power testing lab and interference was a major problem when your switching 132kv.We did a lot of experimentation to reduce the noise and by far the most efficient way was to tightly twist the pairs, this was much more effective than screening which is counter intuitive but definitely works, so if you twist the ends of your fanout this will certainly help.

 

[museav]

There are many facets to shielding and grounding, but one of the most important here is that for audio cable the shield typically is related to both. There will be some practical problems, such as how do you tie the single shield to multiple grounds? The idea of using conductors for grounds and tying the shield to the chassis of the equipment at one or both ends wouldn't seem practical for a snake.

What is the application? Are these all line level returns or mic inputs? Is it in conduit or run exposed? How long is the run? All of these might affect it.

If you are working with bulk cable and hardware then you might want to consider that other than simply the experience, between materials costs, testing, troubleshooting, reliability and so on, most people find it much more cost effective to simply purchase a snake.

It is a common misunderstanding but the balanced and differential two separate issues. Putting the signal symmetrical on both paths has advantages, but noise rejection is not one of them and this has nothing to do with a balanced circuit. Balanced has to do with the impedance of each path (conductor) with respect to ground and nothing to do with signal symmetry, in a balanced audio line the impedances of the two conductors to ground are the same. With the impedance being the same for each path you will get the same voltage induced in each conductor from any outside sources of interference. When you combine this with differential input circuitry it then cancels the induced noise. But note that this noise cancellation is unrelated to the signal itself or any signal symmetry, you could have the signal on just one conductor and still achieve the same noise rejection.

 

[jamsession]

in this case, 50 and 100' runs, not in conduit, again with overall braid and foil shielding and 12 twisted pairs.

thx

 

[BNBSound]

It is a common misunderstanding but the balanced and differential two separate issues. Putting the signal symmetrical on both paths has advantages, but noise rejection is not one of them and this has nothing to do with a balanced circuit. Balanced has to do with the impedance of each path (conductor) with respect to ground and nothing to do with signal symmetry, in a balanced audio line the impedances of the two conductors to ground are the same. With the impedance being the same for each path you will get the same voltage induced in each conductor from any outside sources of interference. When you combine this with differential input circuitry it then cancels the induced noise. But note that this noise cancellation is unrelated to the signal itself or any signal symmetry, you could have the signal on just one conductor and still achieve the same noise rejection.

I thought this might be yet another long winded useless post, but that last paragraph really cleared up my understanding the concept of a balanced line. The fact that the noise rejection relies on impedance explains why when you have a dynamic mic that's "one legging", as long as the phantom isn't popping you still don't get any noise on the channel.

 

[Chris15]

This reality that twisted pair is more important than shielding is why we can use differential balanced transmission for RS-485 in things like C Bus lighting control and those other sorts of architectural systems as well as the comms in alarm systems etc etc for a kilometre without issues. These systems are helped by the driving voltage they use, 15 volts +- is normal. This compared to the 100mV on a good day out of a mic. 1Vpp Line level is more tolerant as I'm sure people would have noticed.

This property is also the basis of UTP Cat 5 & 6. It's also what the lighting boys use for DMX (which is based on RS 485).

RS485 specifies a minimum difference of just 0.2 volts. But it is also happy with a 15 V differential in most cases. This is why 1km is happy and why you can use Cat5 for DMX without problems any within the standard RS422 also works on balanced output, though it is unidirectional while 485 is half duplex. All benefit from termination. RS422 is used by Yamaha for remote head amp control from memory, not to mention conversion when RS232 needs to go longer distances...