Author Topic: Audio Grounding, XLR  (Read 4012 times)

0 Members and 1 Guest are viewing this topic.

Offline gnifTopic starter

  • Administrator
  • *****
  • Posts: 1672
  • Country: au
Audio Grounding, XLR
« on: July 27, 2017, 10:37:34 pm »
Hi All,

I have been building an balanced mic pre-amp for my PC and I have come across an issue I can't resolve without isolating my circuit from earth ground (obviously not an acceptable option). I would assume the noise I get is a ground loop issue except the frequency is way too high, I am heading squelching and noise, peeking exactly at 1KHz. With some experimentation I have found that this noise is being radiated by the video card and DVI cables, if I place an antenna near any of the cables the noise is horrendous. What I do not understand is how disconnecting the earth ground removes this fault.

I have been informed that using some audio isolation transformers will assist here, but I would like to avoid that if at all possible. I would also like to avoid switching to an external sound card/dac.

The circuit I am building is heavily based on this schematic: http://apuckt.blogspot.com.au/2011/08/balanced-microphone-preamp.html

The main differences are, +-15V instead of 12V and I have skipped the phantom power segment as I do not require it, everything before R6 and R7 is omitted.
 

Offline The Soulman

  • Frequent Contributor
  • **
  • Posts: 949
  • Country: nl
  • The sky is the limit!
Re: Audio Grounding, XLR
« Reply #1 on: July 27, 2017, 10:57:25 pm »
Little bit more info is needed here.
What microphone do you use, what cable (must be shielded) what housing does your pre-amp have, where is its power coming from, how do you connect to your computer?
 

Offline gnifTopic starter

  • Administrator
  • *****
  • Posts: 1672
  • Country: au
Re: Audio Grounding, XLR
« Reply #2 on: July 27, 2017, 11:03:48 pm »
Little bit more info is needed here.
What microphone do you use, what cable (must be shielded) what housing does your pre-amp have, where is its power coming from, how do you connect to your computer?

The mic is some Jaycar cheap thing at the moment, it simply has a transducer in it, completely passive device.
The pre-amp is in a full metal enclosure, grounded.
Mic cable is XLR, shielded, etc.
Cable shield is connected to the enclosure.
Cable ground is connected to the 0V rail.
All grounding is wired using a star configuration.
The connection to the PC is RCA -> Stereo Jack, shielded of course.

Also of note... when disconnected from the PC there is no noise at all, clearly the PC is the source.
« Last Edit: July 27, 2017, 11:22:27 pm by gnif »
 

Offline The Soulman

  • Frequent Contributor
  • **
  • Posts: 949
  • Country: nl
  • The sky is the limit!
Re: Audio Grounding, XLR
« Reply #3 on: July 27, 2017, 11:28:42 pm »
Little bit more info is needed here.
What microphone do you use, what cable (must be shielded) what housing does your pre-amp have, where is its power coming from, how do you connect to your computer?

THe mic is some Jaycar cheap thing at the moment, it simply has a transducer in it, completely passive device.
The pre-amp is in a full metal enclosure, grounded.
Mic cable is XLR, shielded, etc.
Cable shield is connected to the enclosure.
Cable ground is connected to the 0V rail.
All grounding is wired using a star configuration.
The connection to the PC is RCA -> Stereo Jack, shielded ofc.

You have a cable shield and a cable ground?

The cable shield should go thru the xlr pin 1 to star ground.
I've seen some ("onehunglow") cable mfgs connect pin 1 inside the xlr to the xlr housing, this is wrong and defeats the star ground.

Do you have a similar power supply as in the schematic (well 7815 instead of 7812)?
Is the current consumption below 10 mA per rail? (anything over could indicate oscillation).
Don't omit R4 and R5.
Add a ac coupling cap. (47uF or so) to the output, better save than sorry.
 

Offline Someone

  • Super Contributor
  • ***
  • Posts: 4510
  • Country: au
    • send complaints here
Re: Audio Grounding, XLR
« Reply #4 on: July 27, 2017, 11:30:03 pm »
Ignore above as its probably not the XLR or microphone that is at issue here as they sit on their own potential from the preamp. I'm guessing you actually have a problem connecting the signal from the preamp to the computer, confirms this by putting a dummy in place of the microphone input and checking that the noise still appears on the computer input. There are no quick or easy fixes as its a ground loop between the PC and the earthed power supply of the preamp.
 

Offline HackedFridgeMagnet

  • Super Contributor
  • ***
  • Posts: 2028
  • Country: au
Re: Audio Grounding, XLR
« Reply #5 on: July 27, 2017, 11:32:03 pm »
The noise is induced in a ground loop. and by breaking that loop you kill the noise.
But you knew that, so my point is when you have a ground loop you don't have a star ground.

Another important point is the "Protective Earth" and the "Gnd" in your circuit aren't necessarily connected and are not normally connected, because the moment you connect two devices you lose your star ground.


To keep your equipment safe and legal (In AU but many other places too) you must make it as one of two Electrical classes.

Single Insulated
Any exposed metal must be attached to protective earth. In which case your RCA sockets need to be isolated from the chassis, and double isolated from mains.

or Double insulated.
Active and neutral must be double isolated from all exposed metal. In which case your metal chassis can directly mount the rca plugs.

The proper definitions are somewhere if you want to look them up.

I guess you should go double insulated into the transformer and don't ground the chassis. Properly rated Heatshrink is allowed as 1 layer of mains insulation.

Hope this helps.


 

Offline HackedFridgeMagnet

  • Super Contributor
  • ***
  • Posts: 2028
  • Country: au
Re: Audio Grounding, XLR
« Reply #6 on: July 27, 2017, 11:37:33 pm »
One other thing, audio transformers aren't necessarily bad, I personally think they sound fine, much better than hum.
But I learnt the hard way don't use a mic level transformer at line level. It will saturate, you would need a line level transformer.
 

Offline The Soulman

  • Frequent Contributor
  • **
  • Posts: 949
  • Country: nl
  • The sky is the limit!
Re: Audio Grounding, XLR
« Reply #7 on: July 27, 2017, 11:42:32 pm »
Also of note... when disconnected from the PC there is no noise at all, clearly the PC is the source.

Measured with a oscilloscope?

Also get some capacitance on the output as a rf filter, some ceramics will do, don't go over a couple of nano Farads.
 

Offline gnifTopic starter

  • Administrator
  • *****
  • Posts: 1672
  • Country: au
Re: Audio Grounding, XLR
« Reply #8 on: July 27, 2017, 11:51:24 pm »
Also of note... when disconnected from the PC there is no noise at all, clearly the PC is the source.

Measured with a oscilloscope?

Also get some capacitance on the output as a rf filter, some ceramics will do, don't go over a couple of nano Farads.

Yes, measured with a scope.
Noise exists even without a mic attached, it is not the source.
I can reproduce the noise simply by connecting audio ground to the PC chassis.

One other thing, audio transformers aren't necessarily bad, I personally think they sound fine, much better than hum.
But I learnt the hard way don't use a mic level transformer at line level. It will saturate, you would need a line level transformer.

Thanks, the reason I am avoiding them is because simply I don't have any... this circuit was built just "because"... out of what I have on hand, I don't really want to invest any money into it.

I think I will go the route of double insulation... Thanks for confirming that this is still a ground loop issue, I wanted to be sure before I went down the wrong rabbit hole :)
 

Offline The Soulman

  • Frequent Contributor
  • **
  • Posts: 949
  • Country: nl
  • The sky is the limit!
Re: Audio Grounding, XLR
« Reply #9 on: July 27, 2017, 11:58:19 pm »
Ignore above as its probably not the XLR or microphone that is at issue here as they sit on their own potential from the preamp. I'm guessing you actually have a problem connecting the signal from the preamp to the computer, confirms this by putting a dummy in place of the microphone input and checking that the noise still appears on the computer input. There are no quick or easy fixes as its a ground loop between the PC and the earthed power supply of the preamp.

Over here it's not common (pun intended) to connect the "center tab" of the output of such a psu directly (inside the psu)
to the mains ground, if it is than yes that is likely to be the cause of the noise.

And you're correct the microphone and its xlr cable can't cause ground-loops on it own, but when the pin-out is wrong it could easily become a antenna and without band-with limiting (as per schematic) in the pre-amp, all sorts of crap can result.
 

Offline The Soulman

  • Frequent Contributor
  • **
  • Posts: 949
  • Country: nl
  • The sky is the limit!
Re: Audio Grounding, XLR
« Reply #10 on: July 28, 2017, 12:03:47 am »
I think I will go the route of double insulation... Thanks for confirming that this is still a ground loop issue, I wanted to be sure before I went down the wrong rabbit hole :)

Still don't now what you are powering it now from.  :)
 

Offline HackedFridgeMagnet

  • Super Contributor
  • ***
  • Posts: 2028
  • Country: au
Re: Audio Grounding, XLR
« Reply #11 on: July 28, 2017, 12:12:21 am »
One other thing if you are double insulated and your chassis is attached to circuit ground, then you will need rubber feet as any metal object or equipment that comes into contact with your chassis can cause a ground loop.
 

Offline gnifTopic starter

  • Administrator
  • *****
  • Posts: 1672
  • Country: au
Re: Audio Grounding, XLR
« Reply #12 on: July 28, 2017, 12:58:59 am »
I think I will go the route of double insulation... Thanks for confirming that this is still a ground loop issue, I wanted to be sure before I went down the wrong rabbit hole :)

Still don't now what you are powering it now from.  :)

The circuit is in the schematic... lower 1/2
 

Offline gnifTopic starter

  • Administrator
  • *****
  • Posts: 1672
  • Country: au
Re: Audio Grounding, XLR
« Reply #13 on: July 28, 2017, 10:21:36 am »
I have made all the suggested changes, the noise from the PC is gone from the signal!  :-+

Now I have another issue... as I raise the gain a 50Hz signal appears on the output, even with the input disconnected and shorted.



I would assume power supply but it seems clean, I can't for the life of me figure out where this is coming from. Again I call on the communities wisdom :)
I am not sure either if that is an acceptable noise level for this kind of a device.

If I probe before R10 it doesn't show on the scope, either the signal is too low to see it, or it's not there. After R10 it is barely perceivable.
« Last Edit: July 28, 2017, 10:25:28 am by gnif »
 

Offline capt bullshot

  • Super Contributor
  • ***
  • Posts: 3033
  • Country: de
    • Mostly useless stuff, but nice to have: wunderkis.de
Re: Audio Grounding, XLR
« Reply #14 on: July 28, 2017, 11:25:38 am »
OK, you'll have to avoid safety ground in the preamp to kill the noise.
Do not connect the case and / or GND of the amplifier to protective ground.
Otherwise you get a ground loop through the audio connecting cable between the amp and the PC. Part the noise from the PC diverts through the cable shield and protective ground. Since there's no ideal conductor in the world, this causes a voltage drop (noise) over the shield which in turn results in the noise that you experience.
Since you don't want to isolate this path, your only option is to isolate the amplifier circuit from protective ground.
This is safety-wise not a big deal, a metal enclosure doesn't have to be protective grounded if you take care of the line voltage isolation.
Safety devices hinder evolution
 

Offline gnifTopic starter

  • Administrator
  • *****
  • Posts: 1672
  • Country: au
Re: Audio Grounding, XLR
« Reply #15 on: July 28, 2017, 11:38:44 am »
That was with nothing connected, no PC, input shorted, no output... just on the bench on the scope. I have already isolated earth ground from the chassis, and signal ground.
 

Offline capt bullshot

  • Super Contributor
  • ***
  • Posts: 3033
  • Country: de
    • Mostly useless stuff, but nice to have: wunderkis.de
Re: Audio Grounding, XLR
« Reply #16 on: July 28, 2017, 11:52:25 am »
Missed your last post stating you've done that before what I said

Your signal looks like 100Hz rectified line hum.
Most common: ripple on the power supply, you ruled that out. BTW did you check the +17V for ripple?
Other cause: bad GND wiring, coupling C4 / C5 charge current from xformer into the audio path. Sounds plausible, since you don't see it at the output of the first amp stage but on the second stage and this OPAMPs + input is tied to GND.
So check your layout / wiring. It is quite important that all amplifier related GND comes from the GND junction of C4 / C5 and the connection from the xformer goes there, but no common piece of conductor for both paths exists.
Safety devices hinder evolution
 

Offline gnifTopic starter

  • Administrator
  • *****
  • Posts: 1672
  • Country: au
Re: Audio Grounding, XLR
« Reply #17 on: July 28, 2017, 03:02:33 pm »
I figured it out! It was a power issue in the end  :palm: I am not sure why the scope wouldn't show it up though, I couldn't see any drops in voltages on either of the rails, but after swapping out the two main filter caps for fresh ones the noise was gone. It was the +15V rail losing regulation, so the opamp was become biased to -15V
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf