GND when building multi-board audio equipment

[pipthepilot]

Hello Everyone,

I'm designing a 4 channel audio mixer and to make it easier to build and test, I decided to build different sections on separate PCBs. However, by separating the boards, I realise I'm not sure how to manage the grounding as my design currently has ground loops.

The design uses the following PCBs that i'm designing...

• Power board: Distributes ±15V and 48V phantom power
• Balanced input/output board: XLR sockets, phantom power switch and coupling caps
• 4x Channel boards: Stereo Preamp, Level and EQ
• Bus & Mixer board: L&R master BUS and mixer

The diagram below is an approximate layout of my design.


Power is supplied to the input & output board which sends power and balanced input signal to each of the channel PCBs. The bus/mixer board is also supplied power from the power board. Each channel send Left & Right signal to the bus/mixer board and the master left/right signal is sent back to the output.

I guess my question is should I include a GND signal for the audio out from each channel board to the mixer, and mixer to output or are those unnecessary?

Thanks in advance for the help,
Philip

 

[ccaudle]

should I include a GND signal for the audio out from each channel board to the mixer

A good first step is to remove the term "GND" from your vocabulary, and instead think in terms of function and current flow.
You have two needs: a current path for power supply return current, and a stable reference for the audio signal.
Mixing those two functions together thoughtlessly is a recipe for superimposing power supply noise onto your audio signal.

This paper is a good explanation for a more useful way to think of the signal and power flow.
Putzeys: The G Word, or How to Get Your Audio off the Ground

Read that paper a couple of times to let it sink in, then rephrase your question as "should I have a reference signal for my audio signal which has low loop area and is less likely to pick up noise, or should I let the reference for the audio signal take a poorly defined circuitous path around the entire chassis, forming a huge loop antenna to pick up magnetic fields" and the optimal approach will probably become more obvious.

But note that you are providing the audio reference signal, not "including a GND," which requires a change in how you think about the circuitry on the bus and mixer PCB.

 

[pipthepilot]

A good first step is to remove the term "GND" from your vocabulary, and instead think in terms of function and current flow.
You have two needs: a current path for power supply return current, and a stable reference for the audio signal.
Mixing those two functions together thoughtlessly is a recipe for superimposing power supply noise onto your audio signal.

This paper is a good explanation for a more useful way to think of the signal and power flow.
Putzeys: The G Word, or How to Get Your Audio off the Ground

Read that paper a couple of times to let it sink in, then rephrase your question as "should I have a reference signal for my audio signal which has low loop area and is less likely to pick up noise, or should I let the reference for the audio signal take a poorly defined circuitous path around the entire chassis, forming a huge loop antenna to pick up magnetic fields" and the optimal approach will probably become more obvious.

But note that you are providing the audio reference signal, not "including a GND," which requires a change in how you think about the circuitry on the bus and mixer PCB.

Thank you, haven't had chance to read yet but definitely will. You're absolutely spot on that I don't have the correct vocabulary to think of the circuitry correctly.

 

[Matador]

This paper is a good explanation for a more useful way to think of the signal and power flow.
Putzeys: The G Word, or How to Get Your Audio off the Ground

The advice in there needs to be considered VERY carefully: he implements a fully differential amplifier circuit at the end, and proclaims that no ground reference is needed (because of course it isn't), and then proceeds to tie pin 1 of the XLR's into the PCB 'GND' pour which is exactly what he spent the previous 5 pages of text saying was the evil-ness of most implementations (and directly contradicts what he said was the pin1 error in Figure 23, and the remedy in Figure 24).

 

[ccaudle]

then proceeds to tie pin 1 of the XLR's into the PCB 'GND' pour

That is discussed in the text however:

Circuits designed according to the differential
method explained earlier are insensitive to pin 1
problems. The PCB copper fill functions more like a
chassis than a reference. Later in our demo project
we’ll be cheerfully tying pin 1 and the shell to the
ground with no ill effect. But this is not how most
equipment is designed. Most are single-ended internally
and they do use ground as a global reference.
“Ground current” means current through your
internal reference.

 

[sahib]

A good first step is to remove the term "GND" from your vocabulary,

I wouldn't say that he should remove the term GND from his vocabulary as it has its own purpose.

and instead think in terms of function and current flow.

But that is good advice.

Unfortunately the electrical and electronics engineering is full of misplaced and in a lot of cases wrong descriptions. The key is to dig deep in theory and question everything you read.

 

[Matador]

That is discussed in the text however:

I get that, but it's essentially a footnote in a single paragraph in a dozen pages. It's like writing a paper on the evils of salt in the diet, then at the end saying "Now let me present a recipe that uses a little bit of salt just to show that the taste isn't impacted."

The board layout (and schematic for that matter) he presents directly contradicts his advice in Figure 24, and for no reason. I understand why he went against his own advice, as that advice doesn't apply in this specific circuit, however he also could have tied pin 1 to chassis per Figure 24, and left that pour isolated/hierarchal, which is advice that works in EVERY circuit, and directly follows Figure 24, and has no downside even in a fully differential amplifier.

However the rest of the article contains a lot of good advice IMHO.