XLR Connector Assembly Question

ekstasis

New member
Hi, I hope it's Ok to put this question in this forum, I couldn't really find the right place for it . . .

Anyway, I'm putting together an XLR cable. I know that the shield of the cable goes to pin 1, but there's a G-Terminal in addition to pins 1,2,3 and I'm wondering if I am supposed to tie that to anything or just leave it alone.

This terminal will apparently be in contact with the housing of the XLR connector once it's put together.

Thanks in advance.
 
If your cable is 3 conductor cable (shield/ground), signal high and signal low, the shield should be jumpered over from pin 1 to the "G" connection. This connection provides ground to the housing when it's assembled and connects the chassis of the mic pre and the mic housing together. This helps to eliminate ground potentials which can cuase pops, clicks and other nasty things like static zapps.
 
mshilarious said:
Here's the long-winded explanation:

http://www.rane.com/note110.html

Short answer: you can if you want. In theory it provides an extra inch of shielding.

About the only thing I agree with in that explanation is Figure 1b, and even then, I disagree with the details.

Here's my opinion on the matter, for whatever it's worth:


What is a Ground Loop?

The term "ground loop" is a bit of a silly term. The real cause is not the loop (which is harmless), but rather the improper grounding in one piece of equipment and/or one or more electrical outlets. This improper grounding causes the device's signal ground to be partially floating (at a high resistance to the actual earth ground).

The result of this is that the path of least resistance from the poorly-designed gear to ground is through the audio cables to the other piece of gear. The piece of gear with a proper ground ends up being a sink for all of the extra crap coming from the bad piece of gear, and all of that noise goes straight through the cable ground and into the secondary device.

Depending on how the devices are designed, most of this noise may be rejected, but if you have enough noise---say an initial signal of several volts---even a fraction of a percent of that signal will result in audible noise in the audio signal.


What Causes Ground Loops?

All grounding problems fundamentally stem from two sources: equipment designers and home/office/studio builders who don't know what the hell they're doing. :D

In the case of grounding problems due to equipment designers, generally speaking, this is caused by either a poorly designed power supply or by long, narrow ground traces on circuit boards. Either of these issues can cause a relatively high resistance between the signal ground and the actual ground wire coming from the wall. (Don't get me started on gear with two prong plugs.) This is bad.

A proper design, as hinted at by figure 1b in the link above, is to have as low a resistance as possible between the signal ground and the actual earth ground (which may not necessarily require that the chassis be grounded, unlike what is shown in that figure).

This can be done in several ways:

  • 1. Two solid ground planes---one on each side of every circuit board. This is the best way, but it requires a multi-layer circuit board design that most designers aren't willing to spend money to implement, particularly if that would be adding two more layers to a board that already has... say 6 layers.... :D (Note that the need for such extremes is greatly reduced by proper case grounding, but only if you have either an external power supply or proper shielding between the power supply and the audio section.)
  • 2. Multiple ground points---for the ground plane (such that it is), make sure that none of the ground paths are narrow, and make sure that it is tied to ground in multiple places along any traces of significant length across the board. Just remember this little rhyme and you'll get the basic idea: under an ohm sends noise running home; if two ohms are earned, the product's returned. :D
  • 3. Jack-to-case grounding---as shown in figure 1b in the link above. It is strongly recommended that any external jacks be tied directly to earth ground, and the chassis is usually solidly grounded. If this is not the case, substitute a direct, heavy gauge wire to the earth ground. (Note, however, that in an ideal world, the chassis should ALWAYS be grounded for safety reasons.)

The second screw up, which causes true ground loops (which are relatively rare) is caused by poor building design. A true ground loop is what happens when the actual earth grounds of two outlets have different electrical potential. This means that one of the grounds is not adequately grounded and is instead acting more like a neutral wire.

Use of substandard wire gauges, substandard grounding rods, having separate grounding rods for separate parts of a building with no electrical path between the two rods, etc. can cause true ground loops. You will probably never see a real ground loop in your entire life unless you are operating in a relatively old building, but I suppose it's possible....

The third and final cause of ground loops is consumer equipment. The reason that most unbalanced gear doesn't exhibit ground loops (as the person who wrote the above article failed to mention) is that most consumer gear doesn't have a ground AT ALL! By not hooking up the ground, you end up without ground loops, but all of your gear ends up with some voltage potential, and if one piece of gear ends up with current leakage in its power supply, you'll get shocked whenever you touch any of your equipment.

This problem is exacerbated (as the person noted, but failed to explain) when you mix professional and consumer equipment. It has nothing to do with mixing balanced and unbalanced equipment. Most gear with balanced connections have three prong power cords, and most gear with unbalanced connections don't. When you introduce a piece of gear that actually has an earth ground... well, you're back to having a single current sink and all that noise comes flooding through your audio cables because it's the path of least resistance to ground again.


What Can I Do to Fix These Problems In My Studio?

One commonly held myth is that you can reduce ground hum by disconnecting the earth ground of one device. The problem is that this only does anything if you disconnect the ground of the device that is properly grounded. The end result is a giant floating ground bus that will shock you when you touch the gear. (They don't call it 60 hurts for nothing. :D) Never do this.

For most so-called ground loops (the ones caused by bad gear), the best cure I know of is to ground bus everything to a building earth ground (third prong).

If you have a -true- ground loop (caused by two grounds that are both at least moderately decent, but on separate circuits without a shared ground), then and only then should you solve this by lifting a ground. Do not lift the earth ground of the power cable, though. Lift the ground pin and shield connection in the signal cables.

As a general rule, if your grounds are solid (with ground bussing if necessary to compensate for badly designed power supplies in certain pieces of gear), you should never need to lift a ground unless you are running between buildings or something insane.

As a general rule, microphone cables should ALWAYS have the metal case connected to the ground pin. In the worst case, it does no harm. In the best case, when you hook two cables together, it prevents those metal cases from turning into a giant antenna in electrically noisy environments....
 
I disagree with everyone. :D

Don't hook the shell up. It can cause more problems than it solves. It really only solves one problem, and a rare one- enough RF to cause a problem in about 1/2" of unshielded cable.

It can actually cause hum by creating those uneven paths to ground. It can also affect phantom power returns, and so I don't do it on mic cables or snakes. It can also make stopping hum in line-level gear with a shield lift switch ineffective in the case of a metal chassis-mount XLR, for instance. The shield cannot simply be lifted with a switch. Some gear still has separate signal and chassis grounds, at different potentials. Hooking them together through the XLR shell can cause hum and also phantom power problems.

The people who made your gear decided whether or not to ground the XLR shell when they made it, making the connection in the chassis-mount XLR or not. If they think it should be grounded and could be with no problems, they did it for you, inside the gear. Don't do it in your cables unless you come across a situation where you have to. Unless you live right under a radio tower, I bet you won't. It's easier than spending your time undoing a bunch of pin1/shell connections.

Dgatwood's advice is actually good, assuming perfect gear and electrical design. You will find those pretty much 0% of the time. :) It's also good to understand why they would do it in the first place. In lots of real pro places, the whole place is designed around reducing RF and noise. Connecting pin1 and the XLR shell is one of the first things that has to go when you go into the real world.

Neutrik makes a connector that has goodies ( a capacitor and stuff) built in to get around some of the pin 1 /shell problems if you want. Markertek has them for $6 each.
 
Last edited:
boingoman said:
Don't hook the shell up. It can cause more problems than it solves. It really only solves one problem, and a rare one- enough RF to cause a problem in about 1/2" of unshielded cable.

Rare? Hardly. Every single school auditorium I ever recorded in caused problems for me. Light dimmers can produce MASSIVE amounts of EMI. :D


boingoman said:
It can actually cause hum by creating those uneven paths to ground. It can also affect phantom power returns, and so I don't do it on mic cables or snakes.

I guess it really depends on what's on the other end of the cable. It is true that this can make a shield lift much harder if you're going to plug things other than mics into the other end of a snake, so to that end, yes, it can create uneven paths to ground. It's probably best to have a handful of ground lift adapters, which, if made correctly, can lift both the signal and case ground.

Of course, again, you should only need this if you are running cables between gear in different rooms or on a different circuit, and only then if the building grounds suck.


boingoman said:
It can also make stopping hum in line-level gear with a shield lift switch ineffective in the case of a metal chassis-mount XLR, for instance. The shield cannot simply be lifted with a switch.

Depends on the XLR chassis connector that the equipment manufacturer used. Good connectors have the chassis contact separate from the shield contact for this very reason.


boingoman said:
Some gear still has separate signal and chassis grounds, at different potentials. Hooking them together through the XLR shell can cause hum and also phantom power problems.

Yes and no. It isn't actually causing the hum problem. The lack of proper connection between the two ground planes is simply masking the power supply leakage problem that already exists in that piece of equipment. (Where I come from, those pieces of gear are referred to as "shock hazards".... :D) When you hook up a cable with a proper ground, the path of least resistance from the piece of gear's crappy chassis ground to earth is through your audio cable. The correct way to fix this is not to leave the chassis ground floating. The correct way to fix this is to properly ground the chassis.

There are very few grounding problems that can't be solved by grounding the chassis and signal grounds of problematic gear with a 10 AWG wire directly to the third prong on an outlet....
 
Back
Top