XLR Shorting Plugs

Most of what I've found on the web regarding creating XLR shorting plugs (to try and reduce noise picked up through unused inputs - worth a shot) shorts only hot to ground, and not negative as well. Is there a reason for this?

With my RCA shorting plugs, I "shorted" using a 1k resistor - partly because I was worried they might get unintentionally stuck into an output and fry an output stage, and partly for a reason that I thought was quite good but I can't currently remember. Given that the first is impossible for XLRs, is there a good reason to use resistors instead of straight shorts?

Thanks,

Shayne Hodge
 
XLR shorting plugs are usually used when a RCA is wired is parallel with one side of the XLR input. The shorting plug would be inserted into the unused side of the XLR when the RCA is in use. An XLR iput on it's own should inject very little noise into the system as XLR uses a balanced system with an inherently very high CMRR. Most any noise that is picked up should be common to both the +ve and -ve line in the XLR and so cancelled before it can really get too far into the system and cause problems.
 
XLR shorting plugs are usually used when a RCA is wired is parallel with one side of the XLR input. The shorting plug would be inserted into the unused side of the XLR when the RCA is in use.
An XLR iput on it's own should inject very little noise into the system as XLR uses a balanced system with an inherently very high CMRR. Most any noise that is picked up should be common to both the +ve and -ve line in the XLR and so cancelled before it can really get too far into the system and cause problems

Hi and sorry to take up again this old discussion.
But i am doing some noise measurement on usb interfaces and i would like to be sure to measure only noise generated by the interfaces.
So the idea of shorting the mic inputs.
From the thread above i understand that is not necessary ?

Moreover, in the case the shorting is recommendable, i read somewhere that is better to short the 1 and 3 pin to ground using small value resistors (like 100-200 ohm)
I did not quite understand why these resistors are needed.
I am doing now the tests with the mic ins open.
But i would like to have it shorted to be sure that they not capture disturbs around.
This noise measurements are very important to me because i am looking for a very silent interface.
I have noticed already that many interfaces have quite noisy mic preamps when gain goes up.
I am looking for a really silent interface.
If all this is silly tell me without problem.
Any advice would be very welcome and appreciated.
Thanks a lot indeed.
Kind regards, gino
 
Last edited:
The idea of a balanced input is exactly that; it is balanced. If each pin is the same size and length, which they are, the noise (if any) will be present on both +ve and -ve inputs, (pins 2 & 3) and as the inputs are balanced the noise will cancel.
If however you find that by connecting pin 2 to pin 1 stops any noise, the input is not balanced.
 
The idea of a balanced input is exactly that; it is balanced. If each pin is the same size and length, which they are, the noise (if any) will be present on both +ve and -ve inputs, (pins 2 & 3) and as the inputs are balanced the noise will cancel.
If however you find that by connecting pin 2 to pin 1 stops any noise, the input is not balanced

Hi and thanks a lot indeed for the very kind and valuable advice.
Then it could be of some value to use them in ordet to check if the inputs are really balanced of what. Very very interesting. I think i will do it. I am curious. I am in the process of evaluating some usb interfaces and noise is indeed a very important aspect. I like very low noise.
Issue closed.
Thanks a lot again, gino
 
The main noise (and I am agreeing with the others above about balanced) source would be on the balanced line somehow. if you are testing, measure noise with unterminated input, then short pins 2 and 3 together - that shorts the balanced line. If you see a reduction in noise, then you might do something about it. But I am suspecting it won't matter.
 
The main noise (and I am agreeing with the others above about balanced) source would be on the balanced line somehow.
if you are testing, measure noise with unterminated input, then short pins 2 and 3 together - that shorts the balanced line.

Hi and thanks a lot for the valuable reply.
You mean connect 2 with 3 and leave the ground input open ?

If you see a reduction in noise, then you might do something about it. But I am suspecting it won't matter.

Thanks a lot i will do that.
I think that a lot can be learned from these noise measurements.
Actually with THD measurements should provide a good overall measure of the quality of an interface.
For instance i have started measuring a usb powered interface.
Replacing the power from the usb bus with a 2A Samsung charger (i have a special cable to do this) has no impact on the noise of the unit.
So i ordered a new 10A usb charger ... just to play a little :eek:
Thanks a lot again, gino
 
Ground is not part of the input for a balanced system. It is already connected to ground.
That is why it is called ground

Hi and thanks for the very helpful reply.
In my mind, maybe wrongly, shorting an input meant to connect it to the ground. So shorting a XLR meant to connect both + and - to the ground.
But if connecting - and + together is a more telling test i would do that for sure. I am open to good options.
Thanks a lot again, gino
 
Hi and thanks a lot for the valuable reply.
You mean connect 2 with 3 and leave the ground input open ?



Thanks a lot i will do that.
I think that a lot can be learned from these noise measurements.
Actually with THD measurements should provide a good overall measure of the quality of an interface.
For instance i have started measuring a usb powered interface.
Replacing the power from the usb bus with a 2A Samsung charger (i have a special cable to do this) has no impact on the noise of the unit.
So i ordered a new 10A usb charger ... just to play a little :eek:
Thanks a lot again, gino
Pin1 is the connection chassis to shield/screen.
Pin1 forms part of the enclosure. It is not part of the audio.
Forget the word "ground".

Pins 2 & 3 are the signal/audio.
 
To short an unbalanced input you connect it to ground. To short a balanced input you connect the two connections together. In fact, its the same for both: both have two wires, so connect them together to short the input.
an unbalanced connection has two wires. Signal flow/hot and signal return/cold.

One shorts the two together for a shorted input.

Take an example:
The RCA/Phono socket on the back of an amplifier.
The pin is the signal flow/hot.
The barrel is the signal return/cold.
Both flow and return are isolated from the enclosure.
One inserts a dummy shorting plug into the socket.
That shorts the flow to the return. Neither gets connected to the enclosure.
 
To short an unbalanced input you connect it to ground. To short a balanced input you connect the two connections together. In fact, its the same for both: both have two wires, so connect them together to short the input

Thanks a lot again. I will prepare two plugs.
I am testing some interfaces and most of them have balanced inputs.
However i guess that open inputs should be a more critical testing condition for noise. Am i wrong ?
Regards, gino
 
Pin1 is the connection chassis to shield/screen.
Pin1 forms part of the enclosure. It is not part of the audio.
Forget the word "ground".
Pins 2 & 3 are the signal/audio.

Hi ! thanks for the precious explanation.
I am a little confused about ground honestly.

an unbalanced connection has two wires.
Signal flow/hot and signal return/cold.
One shorts the two together for a shorted input.

and this is fine. I understand this.

Take an example:
The RCA/Phono socket on the back of an amplifier.
The pin is the signal flow/hot.
The barrel is the signal return/cold.
Both flow and return are isolated from the enclosure.
One inserts a dummy shorting plug into the socket.
That shorts the flow to the return. Neither gets connected to the enclosure.

Ok ! i think i have got it ! i was really confusing things
I mean the ground and the barrel.
Ground and grounding techniques are a mistery to me.
For instance ... the mains input.
The center pin (ground) must be connected always to the chassis for safety ? because sometimes i get noise doing this.
Moreover i have a heapdhone amp that when i use rca to xlr adapters at the inputs i get strong noise.
Using balanced connections no noise at all.
Weird.
Regards, gino
 
the "ground" word causes confusion.
I wish this Forum would ban it. And make us use the unique descriptor for whatever wire we are discussing

Yes i can confirm on my side.
But still i have problem to understand if the metallic chassis of an equipment must be connected to the central pin of a mains plug that i think is ground.
I remember i did this and i got a loud noise on an amp.
And actually the center pin was left open.
regards, gino
 
In a class I appliance all exposed metalwork must be connected to the safety ground lead in the mains cable, in a class II device this is not the case, so it depends on the design of the device.

What you should NEVER do is disconnect the chassis safety earth connection in a class I device.

I would question why anyone would care about noise in an unconnected input? Not only will it be different to the noise when the input is terminated by the impedance of the source (Because the noise gain will be different), but you are presumably not using it anyway.

If you want to measure noise, do it with a specified source impedance (150 ohms between pins two and three is typical for a microphone), or short or open but all will give different numbers (and this is to be expected and is perfectly normal), also do it in a specified bandwidth and weighting....

You really, really, need to hit the books for some basic analogue design, I get the feeling you are hacking around in the dark at the moment (And lay off the messing with mains power supplies (including the safety bonds) until you know a hell of a lot more about what you are doing, that stuff is lethal).

Regards, Dan.
 
In a class I appliance all exposed metalwork must be connected to the safety ground lead in the mains cable, in a class II device this is not the case, so it depends on the design of the device.
What you should NEVER do is disconnect the chassis safety earth connection in a class I device.

Hi and thanks a lot for the very helpful advice.
i am pretty sure that it was a class I unit. I saw that the metal chassis was not connected to the center pin and when i tried i got a very nasty noise.
A strange amp by the way ... not very commercial and old one.
I decided not to use for this danger.

I would question why anyone would care about noise in an unconnected input?
Not only will it be different to the noise when the input is terminated by the impedance of the source (Because the noise gain will be different), but you are presumably not using it anyway.
If you want to measure noise, do it with a specified source impedance (150 ohms between pins two and three is typical for a microphone), or short or open but all will give different numbers (and this is to be expected and is perfectly normal), also do it in a specified bandwidth and weighting....
You really, really, need to hit the books for some basic analogue design, I get the feeling you are hacking around in the dark at the moment (And lay off the messing with mains power supplies (including the safety bonds) until you know a hell of a lot more about what you are doing, that stuff is lethal).
Regards, Dan

Thanks again. You are right. But now i have i discovered the Arta software that allows to perfor noise and distortion measurements on usb interfaces and i would like to test some of them and see.
For instance i have already measured one with the inputs open.
I need to prepare these shorting pluges then with the 150 ohm resistor.
Then i could test all with confidence.
I am only interested in measuremets of noise and distortion.
I am sure that the lowest the best.
Thanks again, gino
 
ginetto61 said:
However i guess that open inputs should be a more critical testing condition for noise. Am i wrong ?
That would depend on what you mean by noise (many people use noise and interference interchangeably - which can cause confusion).

If you mean interference, then electric field interference is stopped by adding a short. Magnetic field interference is facilitated by adding a short, as that creates a small loop.

Noise results would depend on whether you have voltage noise or current noise dominating at the input.

As always in electronics (and almost everything else), you can't rely on a recipe; you have to gain understanding. All your threads seem to follow the same path: give me a simple recipe to solve this problem. In most cases there isn't a recipe; in some cases there isn't a real problem!