Skip to main content

I have worked on a lot of different systems, by SSL, Neve, Trident and others, and notice that there are MANY opinions on carrying the ground wire in a cable.

I am wondering about advice any here could give me, or even opinions for that matter...

I have both balanced AND unbalanced gear in my patchbays.

Some people say only tie all the Input jacks to ground, others say only the output jacks to ground...( In the cable I mean, as in break the screen on the other jack)
Currently I have both input and output jacks tied to ground and it does OK, but sometimes there are anaomalies.
If I untie the unbalanced stuff, will it still work?

I know for many this is actually a passionate subject with heated tempers ( jeez we are geeks ! :) ) but I'd like to hear all the opinions.

Comments

anonymous Wed, 03/21/2001 - 16:13

...that there are MANY opinions on carrying the ground wire in a cable. I am wondering about advice any here could give me, or even opinions for that matter...

...I have both balanced AND unbalanced gear in my patchbays.

Some people say only tie all the Input jacks to ground, others say only the output jacks to ground... ( In the cable I mean, as in break the screen on the other jack) Currently I have both input and output jacks tied to ground and it does OK, but sometimes there are anomalies.

If I untie the unbalanced stuff, will it still work?

It depends a bit on what kind of wiring you have and how the signals conductors are connected - if you have screened twisted pair "balanced" cable and the signal is running on the pair then dropping the screen at one end should not be a problem.

I generally wire studios as if everything were balanced... so I use wire and cabling with shielded twisted pairs. In a shielded/twisted pair you will have two "signal" wires. One of these is the "hot" or "high" and the other is the "cold" or "low". These are the only wires that always have to be connected at both ends. The shield is often some copper braid or maybe foil that wraps around two signal wires along the length of the cable - the twisted pair. The shield is a conductor, in the sense that it CAN conduct from one end of a cable to the other end, but it is not meant to conduct current anywhere. In most cables there is a wire running with the shield called the "drain" wire. The drain wire is also called a ground wire but that description is really going in the wrong direction... for the shield to work it has to have a low impedance connection to ground via a drain wire. The shield needs to have one such connection. Multiple connections can cause problems and I can describe these but I want to try and keep this post brief so maybe in a follow-up I will explain how differing ground potentials and ground loops can ruin your day. You CAN wire a studio with twisted pair wire and NO shields connected anywhere and it will work - probably not well enough to use but it will work functionally.

So let's talk about the twisted pair first and leave the shield for a moment.

The hot wire is pretty easy assuming you know how the gear itself is done. The hot wire is going to be THE signal wire in all unbalanced connections so that would be the "tip" connection in most cases except for those insert jacks that have the send and return hots on the same Tip/Ring/Sleeve connector. With XLR connections we now have a standard where pin 2 is hot but enough gear already exists with pin 3 hot that it is always good to KNOW what it what. Some people seem to think that the existence of a standard will somehow metamorphise older non-compliant equipment into agreement but this is not the case and not every manufacturer making gear today follows the "pin 2 hot" standard - so you have to figure that out on a piece by piece basis.

The cold wire is a little less obvious because it is carrying the "low side" of a balanced connection or the "signal common" of an unbalanced connection. Unbalanced signal common is pretty much GROUND. So... in a studio with mixed balanced and unbalanced connections you will often see the low side of balanced outputs being connected to ground at the inputs of unbalanced devices. It turns out that this works but depending on the type and how well the balanced outputs are designed there may be some bad side-effects - these have to be dealt with on a case by case basis - I use transformers. You can also see the unbalanced signal common being connected to the low side of balanced inputs and once again this works - but some gear with active balanced inputs (not transformers) really does not like to be unbalanced this way and sometimes they amplify any noise on the ground wire MORE than the audio signal on the hot wire. This is more a function of the circuit design than the balanced to unbalanced connection in itself and once again these problems have to be taken on a case by case basis - I sometime use transformers or active circuitry to fix this.

The twisted pair or audio pair of wires is how you get a signal from one place to another. Most of the time we have a relatively low impedance source or output being connected through a wire, various connectors, patch bay etc. to a relatively high impedance load or input. One way of looking at this connection is by viewing it as a measurement. The input is essentially measuring an audio voltage as presented at the source terminals (output connector). A voltage is measured as a potential difference between two points. A balanced output would have those two points be the output high and output low whereas the unbalanced output would have those points be the output signal and the output common. Since that unbalanced common is the same as ground and "ground is ground" then you might think that all we would have to connect in an all-unbalanced studio would be hot wires. The problem is that ground is not the same everywhere - the only way that can happen is if all the chassis and conductors carrying ground have ZERO impedance... ZERO. Even a super-conducting wire will have some inductance. So if I want to measure the output of something very accurately and with the least amount of interference I want to measure right at the source - otherwise I will be adding other voltages to my measurement - like ground noise. So two wires for signal - even in an unbalanced situation.

I am zipping though this - there are whole books that cover just this question.

Now - the shield or drain wire, with one very obvious exception, should be connected at one end only. The exception is microphone cables - for two reasons. One is that the shield/drain wire is often the only way the microphone case can be grounded and the other is that phantom power uses the shield/drain wire to carry the return or back to the source - zero volts. With that exception stated it can be said that it is almost universally accepted that the shield/drain wire should be not be used to carry current or be used to ground equipment and that for the purposes of making the shield functional the drain need only be connected at one end. There some "if, ands or buts" regarding this but I'm skipping those as they really don't come in to play often in most studios... maybe in a follow-up.

All the heated talk and controversy seems to revolve around which end gets the shield termination. Connect only at sources or only at destinations? Connect shield/drains at the patchbay and drop at the far ends or drop at the patch and connect at the far ends? There are actually some very good technical papers on this and both sides (source / destination) have been argued. Actually there is some greater advantage in connecting shield/drain wires at inputs to devices as opposed to outputs from devices and here we have to view the patchbay as a device as well. Rather than argue one particular shield religion over another it should be said that the more important thing is to come up with a system that is consistent and does not randomly connect some shields at both ends such that they end up carrying current from one chassis to another or leave some shields randomly unconnected at both ends so they can't do their job. THAT is when problems start happening.

So - a rational shield termination scheme may not solve all your problems. It may be that some chassis are not grounded because all the gear is plugged into "ground lift adaptors"... but this is another topic - power distribution and ground.

For more information, if you really want to know all about wiring AND power distribution, I strongly suggest that you go to http:// and order the reprint of the journal issue on Shields and Ground for $15.00.

Aaron-Carey Wed, 03/21/2001 - 17:14

ok this is exactly the kind of response I was hoping for, and maybe should be required reading!
I am using foil screen twin + drain for everything....In the case of unbalanced equipment I tie the - wire to ground, and use a UNBALANCED 1/4 " plug...
this way, when it is connected, the longer sleeve area ends up connecting what would be the ring and the sleeve in a balanced 1/4 ".

Specifically, I am glad you brought up the mic system ground.
In my studio, the mic goes to a snake, the snake comes to a row on a patchbay...from there it is to be patched into either outboard mic pre's or console mic pre's...I guess these MUST have the ground connected all the way through so I can use phantom power.

Like I said, at the moment, the ground, or shield or drain is connected at both ends.

From here I gotta figure out which connections CAN be broken safely in a consistent manner so that if I patch something in I am not making a ground loop.
SO whats your poison? leave the patchbay side connected, and outboards not? Source? Destination? help! :)

Guest Thu, 03/22/2001 - 05:39

Aaron-John gave you a fairly lengthy response, that barely scratched the surface. My advice to you is to hire someone with experience in this area, have them design a ground scheme for you, then implement that scheme.

Trying to figure it out on your own leads directly to migranes. As John pointed out, there are several different schools of thought as to where to connect the shield/drain wire...there are also differing schools of thought on how many of them to connect at which end [and that's as far as I'm going because this is a can of worms I do not want to open for love nor money].

May I suggest you talk to the 'tech' at the largest studio in your area...perhaps for a nominal service charge they can affect a scheme for you.

anonymous Thu, 03/22/2001 - 06:23

Fletcher is correct - you really should hire someone... but I have typed another lengthy response anyway.

ok this is exactly the kind of response I was hoping for... ...I am using foil screen twin + drain for everything....In the case of unbalanced equipment I tie the - wire to ground, and use a UNBALANCED 1/4 " plug... this way, when it is connected, the longer sleeve area ends up connecting what would be the ring and the sleeve in a balanced 1/4 ".

right - assuming I am reading this correctly.

The basic process for the wiring I do (before taking into account some specific exceptions that exist because some gear is just designed badly) is to treat the wiring system, connections, bays etc. as if all the i/o is balanced. The audio pair is what caries the signal and makes the connections and the shield is a whole other thing. So with balanced gear you just have to deal with figuring out which pin of the xlr is hot and which pin is low. From there you are on your way. With unbalanced gear the "tip" or signal pin is hot and the audio common (or "sleeve" or "ground") is essentially equivalent to the low or "-" signal found in balanced gear. The connectors in unbalanced gear are often 1/4" though xlr's and phono jacks are used as well... sometimes other things. With 1/4" unbalanced i/o you generally have only one signal connection per connector but there are inserts that use TRS for send and return on the same jack and there are a few pieces out there that have two levels of i/o on one TRS connector - like "+4" on the tip and "-10" on the sleeve. In the normal vanilla unbalanced 1/4" i/o connection you should use a 1/4" "mono" or Tip/Sleeve plug - like a guitar plug. [FYI - I have seen this a lot so keep this in mind... if you use a TRS plug and connect to tip and sleeve you may find that the sleeve is not grounded in some gear. The manufacturer may use a TRS jack and use the ring portion of the jack to ground the longer sleeves found on a guitar type plugs - this is very annoying]. The "low" would be the sleeve connection and the high would be the tip connection. Proceed from there with figuring out how to connect the audio pairs.

The shield/drain connections are really another system altogether. Shield/drain wires don't ground equipment chassis. For shields to work and do the job of reducing noise pickup by the audio pair they do have to drain off to ground - so there has to be a ground connection. When you pass AC current down a wire through a shield you have provided a great path for noise to couple into the audio pair that the shield surrounds. The best way to prevent current from flowing in a wire is to not connecting it to anything... or if it is connected to something at least make sure it does not complete a circuit and become a path for current to flow. Since a shield has to be drained off to ground then ONE connection to ground must be made at an end where the wire is terminated to a piece of gear. To assure that no current flows on the shield the OTHER end must NOT be connected.

Specifically, I am glad you brought up the mic system ground. In my studio, the mic goes to a snake, the snake comes to a row on a patchbay...from there it is to be patched into either outboard mic pre's or console mic pre's...I guess these MUST have the ground connected all the way through so I can use phantom power.

right - so for microphones - first we try and avoid patchbays - and we fail most of the time. So we often use patchbays where the jacks have isolated sleeve connections and we take the microphone tieline in from wherever it comes from and make a connection to a top row jack that has a full normal down to a bottom row jack. The ground wire also connects down. So here the ground loops through the bay and goes to the mic preamp input and the source of phantom power. When you patch the mic tieline to another preamp input you still have that original ground connection but you break the audio pair (full normal breaks to outputs and well as to inputs when patched) and make a new connection for the signal pair and an additional connection for ground/shield/phantom power return. There are jacks that have an additional switch on them that you can wire the ground through and that would act as a breaking normal for ground. Those jacks are made by ADC and maybe Switchcraft. We use Audio Accessories Audioline jacks which are substantially better jacks. Alternatively you can simply not wire any normals for mics and patch everything... or you can do an xlr patch system.

...SO what's your poison? leave the patchbay side connected, and outboards not? Source? Destination? help!

One of the easiest systems for shields is to connect them all at a patchbay that has a ground bus and have any wire leaving the patchbay terminated with the shield wrapped back under shrink (or cut off) and disconnected. There are many variations but this is one that is workable. If your patchbays are ground bused and all shields terminate there you can go around a drop shield on the "far ends". In an ideal situation with a low source impedance device connecting to a high input impedance receiving device you get better performance by draining off the shield at the receiving end. This makes takes the end of the shield at the high impedance receiving end directly to ground and reduced the likelihood of RF pickup via capacitive coupling from the hanging shield end to the input. The other end - where the source is a low impedance - will have the hanging shield end but the RF coupling would be lower and more attenuated. You have to remember that a even a short cable like those we use in recording studios will, at very high frequencies (well above audio), have enough inductive impedance to develop a healthy potential from one end to the other. In cases where this becomes a problem our practice is to provide a path for the high frequency potential to drain off - that would be a small capacitor that connects the hanging end of the shield to ground. That would be a .01 or even a .001 microFarad cap. The last time I resorted to this was well over ten years ago with a studio right next to a radio transmitter farm that had a dozen or more AM and FM stations blasting away. Many radio stations just wired shield everywhere and if you have a good tuner and some gain you can hear the ground loop noise in between all the over processed program material...

I am not in a position to give you any more direction. How you proceed is going to be determined by the mechanical aspects of what you actually have and how much of a pain in the ass it will be to work one solution over another. If you were building from scratch that would be another thing altogether.

anonymous Thu, 03/22/2001 - 06:27

Fletcher is correct - you really should hire someone... but I have typed another lengthy response anyway.

ok this is exactly the kind of response I was hoping for... ...I am using foil screen twin + drain for everything....In the case of unbalanced equipment I tie the - wire to ground, and use a UNBALANCED 1/4 " plug... this way, when it is connected, the longer sleeve area ends up connecting what would be the ring and the sleeve in a balanced 1/4 ".

right - assuming I am reading this correctly.

The basic process for the wiring I do (before taking into account some specific exceptions that exist because some gear is just designed badly) is to treat the wiring system, connections, bays etc. as if all the i/o is balanced. The audio pair is what caries the signal and makes the connections and the shield is a whole other thing. So with balanced gear you just have to deal with figuring out which pin of the xlr is hot and which pin is low. From there you are on your way. With unbalanced gear the "tip" or signal pin is hot and the audio common (or "sleeve" or "ground") is essentially equivalent to the low or "-" signal found in balanced gear. The connectors in unbalanced gear are often 1/4" though xlr's and phono jacks are used as well... sometimes other things. With 1/4" unbalanced i/o you generally have only one signal connection per connector but there are inserts that use TRS for send and return on the same jack and there are a few pieces out there that have two levels of i/o on one TRS connector - like "+4" on the tip and "-10" on the sleeve. In the normal vanilla unbalanced 1/4" i/o connection you should use a 1/4" "mono" or Tip/Sleeve plug - like a guitar plug. [FYI - I have seen this a lot so keep this in mind... if you use a TRS plug and connect to tip and sleeve you may find that the sleeve is not grounded in some gear. The manufacturer may use a TRS jack and use the ring portion of the jack to ground the longer sleeves found on a guitar type plugs - this is very annoying]. The "low" would be the sleeve connection and the high would be the tip connection. Proceed from there with figuring out how to connect the audio pairs.

The shield/drain connections are really another system altogether. Shield/drain wires don't ground equipment chassis. For shields to work and do the job of reducing noise pickup by the audio pair they do have to drain off to ground - so there has to be a ground connection. When you pass AC current down a wire through a shield you have provided a great path for noise to couple into the audio pair that the shield surrounds. The best way to prevent current from flowing in a wire is to not connecting it to anything... or if it is connected to something at least make sure it does not complete a circuit and become a path for current to flow. Since a shield has to be drained off to ground then ONE connection to ground must be made at an end where the wire is terminated to a piece of gear. To assure that no current flows on the shield the OTHER end must NOT be connected.

Specifically, I am glad you brought up the mic system ground. In my studio, the mic goes to a snake, the snake comes to a row on a patchbay...from there it is to be patched into either outboard mic pre's or console mic pre's...I guess these MUST have the ground connected all the way through so I can use phantom power.

right - so for microphones - first we try and avoid patchbays - and we fail most of the time. So we often use patchbays where the jacks have isolated sleeve connections and we take the microphone tieline in from wherever it comes from and make a connection to a top row jack that has a full normal down to a bottom row jack. The ground wire also connects down. So here the ground loops through the bay and goes to the mic preamp input and the source of phantom power. When you patch the mic tieline to another preamp input you still have that original ground connection but you break the audio pair (full normal breaks to outputs and well as to inputs when patched) and make a new connection for the signal pair and an additional connection for ground/shield/phantom power return. There are jacks that have an additional switch on them that you can wire the ground through and that would act as a breaking normal for ground. Those jacks are made by ADC and maybe Switchcraft. We use Audio Accessories Audioline jacks which are substantially better jacks. Alternatively you can simply not wire any normals for mics and patch everything... or you can do an xlr patch system.

...SO what's your poison? leave the patchbay side connected, and outboards not? Source? Destination? help!

One of the easiest systems for shields is to connect them all at a patchbay that has a ground bus and have any wire leaving the patchbay terminated with the shield wrapped back under shrink (or cut off) and disconnected. There are many variations but this is one that is workable. If your patchbays are ground bused and all shields terminate there you can go around a drop shield on the "far ends". In an ideal situation with a low source impedance device connecting to a high input impedance receiving device you get better performance by draining off the shield at the receiving end. This makes takes the end of the shield at the high impedance receiving end directly to ground and reduced the likelihood of RF pickup via capacitive coupling from the hanging shield end to the input. The other end - where the source is a low impedance - will have the hanging shield end but the RF coupling would be lower and more attenuated. You have to remember that a even a short cable like those we use in recording studios will, at very high frequencies (well above audio), have enough inductive impedance to develop a healthy potential from one end to the other. In cases where this becomes a problem our practice is to provide a path for the high frequency potential to drain off - that would be a small capacitor that connects the hanging end of the shield to ground. That would be a .01 or even a .001 microFarad cap. The last time I resorted to this was well over ten years ago with a studio right next to a radio transmitter farm that had a dozen or more AM and FM stations blasting away. Many radio stations just wired shield everywhere and if you have a good tuner and some gain you can hear the ground loop noise in between all the over processed program material...

I am not in a position to give you any more direction. How you proceed is going to be determined by the mechanical aspects of what you actually have and how much of a pain in the ass it will be to work one solution over another. If you were building from scratch that would be another thing altogether.

Aaron-Carey Thu, 03/22/2001 - 08:29

Originally posted by Fletcher:
May I suggest you talk to the 'tech' at the largest studio in your area...perhaps for a nominal service charge they can affect a scheme for you.

The sorry sad thing is: I was the tech for the largest studio in the area! There is one other, very well respected guy, but he feels as I " six of one half dozen of the other ".

I think I even bring this issue up because there are more and more and more high speed digital devices, like computers, computer monitors, network cables, and other nasty things present in studios nowadays, especially the smaller ones...the older, conventional grounding techniques may not cover it all anymore. John how is your success when there are computers present? is it a different animal, or do the same laws apply ?

fred01 Thu, 03/22/2001 - 08:58

I agree with everything that John has written here (At least all of it that I read) so I won't restate what he has said.

I hope I'm not repeating anything he said but...

If you use the "float the shields at the patchbay" method, be sure to attach the shield of either the input or the output jack so that the patch cord itself has shielding. I use the input side as I will explain in a moment.

If you use a mike line to mic preamp input patchbay, be sure to use jacks that have normaling switches for all three conductors and that are mounted so that the barrels are isolated. The shields attach to both the mike line and mike preamp input jacks for this bay only.

That will take care of 98% of your mike level patching. But what if you want to use an external mike preamp mounted outside the console? You need to have the shield carry through from the mike line to the preamp input. IF you tie the shield to the input jacks on non-mike level patchbays, then you can have external mike preamps come up on non-mike level patchbays and still be able to patch from mike line to external mike pre. The shield will carry through.

Fred Forssell