Explanation of how modules "putting noise on the ground" bus?

From circuitbending to homebrew stompboxes & synths, keep the DIY spirit alive!

Moderators: lisa, luketeaford, Kent, Joe.

Post Reply
TL072
Learning to Wiggle
Posts: 19
Joined: Sun Sep 06, 2015 3:42 pm

Explanation of how modules "putting noise on the ground" bus?

Post by TL072 » Fri Jun 26, 2020 3:20 am

I have been reading the threads on here discussing the benefits and issues with different kinds of power-supplies and come across a number of references to how badly designed modules (or at least modules that were not designed to avoid it) "put noise on the ground bus" or something similar to that. Although I imagine there are several ways this can occur, I was hoping someone might be able to fill in this gap in my knowledge a little and explain how this can occur?

To put the question in context, I am probably going to buy a linear power supply as I am working off a meanwell switching power supply at the moment and I am seeing that few of the analogue modules (particularly a tube vca) are not at all happy with the switching power and are much quieter if tested on a low capacity linear supply that I have. Intuitively, while slowly building this system, I have noticed that digital modules can add noise to other modules but at that time I think it was related to inadequate power (e.g. too many modules for the power supply). I have also noised small differences when removing other digital devices from the power-bar on which the eurorack is plugged. In trying to figure out where the noise is coming from, I have also noticed that a digital sequencer with many LEDS will even cause a tone to increase in the noise when more LEDS are on which sounds to me like I am hearing something related to a multiplexer.

Although I am probably going to buy a linear supply anyway as this seems to be the easiest way to maybe avoid some of these problems, I was hoping that someone can point me to articles or shed more light on this issue so that I can might ... (1) be able to understand ways to troubleshoot these issues more generally (2) at least be better at isolating the source of issues like this and (3) maybe even do some kind of DIY filtering to fix noisy modules or something like that (this might be a bit wishful thinking but because I do not understand the problem I am not sure)

User avatar
EATyourGUITAR
has no life
Posts: 4770
Joined: Tue Aug 31, 2010 12:24 am
Location: Providence, RI, USA

Re: Explanation of how modules "putting noise on the ground" bus?

Post by EATyourGUITAR » Fri Jun 26, 2020 4:06 am

without all the details, it would be impossible to determine exactly what the primary cause of noise is for you. poor power distribution is a cause of a lot of problems in synthesizers. you need a quiet power supply. low ripple. you need low impedance from the power supply to every module. you need to have a ground the does not move at all. by move, I mean the voltage at ground is no different from ground anywhere else in the system. differences in electrical potential, especially differences that change rapidly over time, are the source of the noise. there are two kinds of noise, common mode, and differential mode. anything can happen in a synthesizer. the voltage ref is also a power rail. upgrade the power supply, upgrade the bus board. read all of the posts on power supplies from Graham Hinton. He has probably posted more correct information to muffwiggler on the subject than any user here. The only advice I disagree with is that all people need absolutely every improvement possible. the reality is that anything is better than what you have now. you might be happy with half way correct as long as there is no safety issue with fire or injury.
WWW.EATYOURGUITAR.COM <---- MY DIY STUFF

User avatar
Rex Coil 7
Super Deluxe Wiggler
Posts: 6532
Joined: Thu Nov 29, 2012 11:29 am
Location: Captain Of Outer Space

Re: Explanation of how modules "putting noise on the ground" bus?

Post by Rex Coil 7 » Fri Jun 26, 2020 4:19 am

EATyourGUITAR wrote:
Fri Jun 26, 2020 4:06 am
... read all of the posts on power supplies from Graham Hinton. He has probably posted more correct information to muffwiggler on the subject than any user here. The only advice I disagree with is that all people need absolutely every improvement possible. the reality is that anything is better than what you have now. you might be happy with half way correct as long as there is no safety issue with fire or injury.
I see Graham's insistence as more about how we should form a frame of mind around. Said another way, I think the school of thinking which states "cover as many bases as possible" is more about a frame of mind rather than an actual literal practice. In other words, always be thinking about doing things as top drawer as one can. Not that's it's "wrong" or "incorrect" to not catch every niggle, it's that it is best to think about every spot you can reduce impedance (so to speak). I see no problem with encouraging that frame of mind. In fact, I used to teach it (in a totally different career field). More about not settling for good enough ... always shoot for top drawer, be thinking about how to optimize the power flow at each turn (so to speak). There isn't a darned thing wrong with that way of thinking, not from my point of view. Obviously one must take into consideration costs vs returns in their own situation (which goes without saying) ... "costs" can be material, or labor.
Never Quit, Die Falling Forward
5U PROJECT - (skip pages 4 through 6, boring junk) ... https://www.muffwiggler.com/forum/viewt ... highlight=

User avatar
Graham Hinton
Super Deluxe Wiggler
Posts: 3221
Joined: Fri Jul 02, 2010 6:28 pm
Location: England
Contact:

Re: Explanation of how modules "putting noise on the ground" bus?

Post by Graham Hinton » Fri Jun 26, 2020 7:19 am

TL072 wrote:
Fri Jun 26, 2020 3:20 am
Although I am probably going to buy a linear supply anyway as this seems to be the easiest way to maybe avoid some of these problems, I was hoping that someone can point me to articles or shed more light on this issue so that I can might ... (1) be able to understand ways to troubleshoot these issues more generally (2) at least be better at isolating the source of issues like this and (3) maybe even do some kind of DIY filtering to fix noisy modules or something like that (this might be a bit wishful thinking but because I do not understand the problem I am not sure)
You are actually talking about two separate problems and putting noise into the ground is solved by improving the power distribution not the PSU itself.

Digital modules, and even some analogue modules like envelopes, use more current on the +ve supply than on the negative. The difference goes into the 0V. If you look up the spec for your modules individually or the whole system you can get an idea of the size of the problem. Don't forget that the currents are dynamically changing so there are dc and ac components. The AC component may be due to a slowly flashing LED or the workings of a microcontroller or DSP. Whatever the source the 0V current multiplied by the resistance of the 0V distribution produces a voltage drop in the wiring that is also carrying the signal return currents so the two get summed.

1) Understanding: it's just Kirchhoff's Law and Ohm's Law. If you are not familiar with them look them up. Familiarity is not just knowing the formulae, it's having a feel for them.

2) The voltage drops are given by V= I.R where I is what the module does (only changeable by better design) and R is the only factor you can do something about. If you can hear a noise due to this it is going to need to be reduced by at least 20dB which is a factor of ten of the voltage, it may need more depending how bad it is. That means that the 0V distribution resistance needs to be reduced by at least a factor of ten which means ten times thicker wires and pcb traces. That isn't really practical and it is expensive so that's why I advocate busbars as a solution where you can easily achieve a factor of 100 lower figures. We're talking milliohms and microohms here, much smaller quantitites than you are used to dealing with.

3) You can filter the voltage rails, but not the 0V. Putting filters on the module power inputs can cause problems if the module is dual function, they really need a filter per function.
EATyour GUITAR wrote:The only advice I disagree with is that all people need absolutely every improvement possible.
I've never given that advice, that's just your interpretation and exaggeration.

I will advise people to spend about 10% of the cost of their modules on the power system. Some people spend more on a custom case and then put an inadequate power system in it.

TL072
Learning to Wiggle
Posts: 19
Joined: Sun Sep 06, 2015 3:42 pm

Re: Explanation of how modules "putting noise on the ground" bus?

Post by TL072 » Fri Jun 26, 2020 8:36 am

I realise I might have confused my question in the last paragraph of my request by talking about "this issue" as if I meant the power-supply issue but both your answers have given me some things to think about (many thanks!) and look into things which I had not even thought of re: the power-supply and internal wiring.

For the benefit of anyone who comes across this thread, your answer led me to this which has already helped a bit to understand what is going on:
Differential (Normal) Mode Noise and Common Mode Noise-Causes and Measures

The point I take away from that article as well is that the layout of a PCB could arguably accentuate other noise issues in the system (there is part about loop and area impacting upon radiated noise for example).

Can you elaborate upon this:
EATyourGUITAR wrote:
Fri Jun 26, 2020 4:06 am
you need to have a ground the does not move at all. by move, I mean the voltage at ground is no different from ground anywhere else in the system.
I get the point in the abstract, but, beyond a virtual ground, I do not quite understand how ground could change between modules all wired to the same ground bus?

Relatedly, I also realise that my original question could have been written more clear and the point about a moving ground is probably related. I understood from some of the posts on power that one module (module A) which is not patched to the other (module B) could actually mess with the power supply enough that module B would be more noisey only because module A was powered by the same bus. It seemed to me that if this was the situation then maybe an adaptor with de-coupling capacitors or something could make module A not screw with the other modules as much. So my question might be: what are some ways that a module could add noise to a power bus but I think your point might be that the module does not actually change ground but its design means that ground is altered by the module from the perspective of another (patched) module and that can appear to other modules as if ground is changing (and therefore create noise)?
EATyourGUITAR wrote:
Fri Jun 26, 2020 4:06 am
you need low impedance from the power supply to every module.
I know I am now getting a bit away from my question but I remember seeing things on this in the power supply discussions but it was not fully explained as far as I recall. Given that the copper traces on bus boards are thin they have a higher resistance (impedance) than say a thick chunk of copper. This means greater power loss at the ends of the boards (e.g. furthest from the power supply) and therefore noise? Also if I recall even if there is not significant power loss resistance can add noise in some cases I think (related to heat in some way I think)?

I have 2 bus boards powered from one side in parallel but there is the possiblity of powering from the center as well (for people finding this thread). Could powering from the center reduce noise a lot (in theory) because the power distribution would be more even? The power supply is only powering 2 x 80hp rows (from the one side) so I had not really thought too hard about this originally when looking at complex wiring diagrams for really large 12U cases etc. but now I am wondering if it could be a factor.
EATyourGUITAR wrote:
Fri Jun 26, 2020 4:06 am
the reality is that anything is better than what you have now. you might be happy with half way correct as long as there is no safety issue with fire or injury.
I appreciate the frankness, clarity, and pragmatism. :mrgreen:

To kind of also reply to Rex Coil 7 as well on this point, I guess it is always a difficult balance between planning/building to avoid problems from the start (e.g. do it well from the beginning) and pre-mature over-optimisation/engineering. For a diy person like myself, I have to recognise that I do not/cannot, for example, solder to the standard of satelite assembly yet things can work well anyway (or have more or less worked until now or enough to enjoy music making) but then suddenly you realise that maybe the corners you cut were not as trivial as you originally thought!

I will definitely need to read more about the things you have raised. I know questions like mine are sort of "I want to know in detail why something is an issue without actually knowing / being able to immediately understand all the details since I am not an electrical engineer" and you not being able to have the synth in front of you so I really appreciate your help and any further comments that others might add.

I just created a page on the sdiy wiki page on troubleshooting noise problems and added a few of your points. As I find things I may put them there to hopefully collect some of the information in one place.
Last edited by TL072 on Fri Jun 26, 2020 9:13 am, edited 1 time in total.

TL072
Learning to Wiggle
Posts: 19
Joined: Sun Sep 06, 2015 3:42 pm

Re: Explanation of how modules "putting noise on the ground" bus?

Post by TL072 » Fri Jun 26, 2020 9:03 am

Many thanks Graham! I think our replies were made around the same time and i think you filled in some gaps in my last post (which I will leave not edit to create more confusion). You have definitely provided further detail to chew on.
Graham Hinton wrote:
Fri Jun 26, 2020 7:19 am
You are actually talking about two separate problems and putting noise into the ground is solved by improving the power distribution not the PSU itself.

Digital modules, and even some analogue modules like envelopes, use more current on the +ve supply than on the negative. The difference goes into the 0V. If you look up the spec for your modules individually or the whole system you can get an idea of the size of the problem. Don't forget that the currents are dynamically changing so there are dc and ac components. The AC component may be due to a slowly flashing LED or the workings of a microcontroller or DSP. Whatever the source the 0V current multiplied by the resistance of the 0V distribution produces a voltage drop in the wiring that is also carrying the signal return currents so the two get summed.
I think I understand and I am probably just repeating what you are saying but: because the +ve is (at a given moment) moving more current to 0v than the -12v is pulling out of the 0v, there will be a current through the 0v back to the mains through the power-supply which changes the electrical potential against which the part of the circuit using the -12v "sees" (or vis-versa). So parts of the audio signal that make reference to 0v (most I assume) then have that changing value super-imposed upon the signal.

Phew, I think I understand and thank you for explaining.
Graham Hinton wrote:
Fri Jun 26, 2020 7:19 am
2) The voltage drops are given by V= I.R where I is what the module does (only changeable by better design) and R is the only factor you can do something about. If you can hear a noise due to this it is going to need to be reduced by at least 20dB which is a factor of ten of the voltage, it may need more depending how bad it is. That means that the 0V distribution resistance needs to be reduced by at least a factor of ten which means ten times thicker wires and pcb traces. That isn't really practical and it is expensive so that's why I advocate busbars as a solution where you can easily achieve a factor of 100 lower figures. We're talking milliohms and microohms here, much smaller quantitites than you are used to dealing with.
At some point I was trying to find a short on trace (no luck but solved another way) using a mili or micro ohm meter so I think I almost understand this point but for someone with limited electrical knowledge it is counter-intuitive (but very helpful to understand) to think about small lengths of wires having enough resistance that they impact upon a signal that is often +/- 12v!

I think this discussion has definitely not changed my opinion that I should get a linear supply (and now maybe improve the bus) but I think your point means that in the very short term, I might re-arrange some modules and re-locate the point at which the power enters the bus to see if there are any improvements out of curiosity.

User avatar
EATyourGUITAR
has no life
Posts: 4770
Joined: Tue Aug 31, 2010 12:24 am
Location: Providence, RI, USA

Re: Explanation of how modules "putting noise on the ground" bus?

Post by EATyourGUITAR » Fri Jun 26, 2020 9:56 am

Although I believe I understand that my opinion of Mr Hinton being a pedantic know it all to be opinion shared by all, I'm sure Mr Hinton will be pedantic and attack my assessment or my opinion as pedantic does. That makes it more pedantic. Shit. How can I have an opening about Mr Hinton without invoking debate or being accused of spreading lies. Do you really think that everyone who has an opinion of you or a first impression is objectively wrong? So many times I try to give Graham the respect he deserves while also being fair. Believe me I do not want to perpetuate this petty back and forth between me and Graham.
WWW.EATYOURGUITAR.COM <---- MY DIY STUFF

Elahrairah
Learning to Wiggle
Posts: 24
Joined: Sat May 02, 2020 12:25 pm

Re: Explanation of how modules "putting noise on the ground" bus?

Post by Elahrairah » Fri Jun 26, 2020 10:11 am

EATyourGUITAR wrote:
Fri Jun 26, 2020 9:56 am
Believe me I do not want to perpetuate this petty back and forth between me and Graham.
You do that by walking away. We all have different skill sets, and it's rare to be both a good communicator and a subject matter expert (and spend your time on messageboards).


On topic of ground noise; I think a key misconception is the difference between theory and reality. In many people's understanding, we assume the 0V line is perfectly ground, because when I put a ground point in my schematic/simulator it just works like expected. In real life, everything has imperfections and your wiring or PCB traces have resistance. It may be micro-ohms, I dont have the tools to measure that.

You can see the effects by replacing the ground points in your simulated schametic with a wire that goes to a shared point way off to the side, then a 1-Ohm resistor, then the theoretical ground. This puts some resistance, importantly SHARED. Now the voltage of that "0V" will fluctuate with usage, and you'll come to understand the value of the relatively huge bus bars. Try the same with an inductor.

Elahrairah
Learning to Wiggle
Posts: 24
Joined: Sat May 02, 2020 12:25 pm

Re: Explanation of how modules "putting noise on the ground" bus?

Post by Elahrairah » Fri Jun 26, 2020 10:23 am

TL072 wrote:
Fri Jun 26, 2020 8:36 am
Could powering from the center reduce noise a lot (in theory) because the power distribution would be more even?
It helps by reducing the distance from any module to the power supply. For a fixed quality of wire, distance scales the resistance and inductance problems we're talking about. If you research wire qualities, the gauge scales in ohms per meter (or milliohms per mile, whatever). So halving the distance will halve your resistance (except for common factors like the eurorack headers being little).

User avatar
Rex Coil 7
Super Deluxe Wiggler
Posts: 6532
Joined: Thu Nov 29, 2012 11:29 am
Location: Captain Of Outer Space

Re: Explanation of how modules "putting noise on the ground" bus?

Post by Rex Coil 7 » Fri Jun 26, 2020 10:30 am

... one other thing to keep in mind ... while we need our resident experts' assistance around here, it serves to remember that often times it's difficult to get two experts in any given field to agree that water is wet, let alone more complex issues.

8-)
Never Quit, Die Falling Forward
5U PROJECT - (skip pages 4 through 6, boring junk) ... https://www.muffwiggler.com/forum/viewt ... highlight=

User avatar
Graham Hinton
Super Deluxe Wiggler
Posts: 3221
Joined: Fri Jul 02, 2010 6:28 pm
Location: England
Contact:

Re: Explanation of how modules "putting noise on the ground" bus?

Post by Graham Hinton » Fri Jun 26, 2020 10:37 am

TL072 wrote:
Fri Jun 26, 2020 9:03 am
I think I understand and I am probably just repeating what you are saying but: because the +ve is (at a given moment) moving more current to 0v than the -12v is pulling out of the 0v, there will be a current through the 0v back to the mains through the power-supply which changes the electrical potential against which the part of the circuit using the -12v "sees" (or vis-versa). So parts of the audio signal that make reference to 0v (most I assume) then have that changing value super-imposed upon the signal.
Almost. The energy is supplied by the mains, but the DC power currents do not return there. Their circuit is completed at the PSU.

The important point is that what you think of as 0V is changing all over the system due the the resistance in the conductors. It can never be zero unless you are going to run superconductors at absolute zero temperature, but you can significantly reduce it and improve noises by 20dB to 40dB which is normally enough.
At some point I was trying to find a short on trace (no luck but solved another way) using a mili or micro ohm meter so I think I almost understand this point but for someone with limited electrical knowledge it is counter-intuitive (but very helpful to understand) to think about small lengths of wires having enough resistance that they impact upon a signal that is often +/- 12v!
There is a drawing on my website that shows all the resistances that have to be considered.
Analysing Power Distribution
I think this discussion has definitely not changed my opinion that I should get a linear supply (and now maybe improve the bus) but I think your point means that in the very short term, I might re-arrange some modules and re-locate the point at which the power enters the bus to see if there are any improvements out of curiosity.
Only small improvements can be made by re-arranging the deck chairs on The Titanic. PCBs and ribbon cable are a very poor way of distributing power, you can get away with it on very small systems, but when you expand the more money you spend the worse it gets.
For the benefit of anyone who comes across this thread, your answer led me to this which has already helped a bit to understand what is going on:
Differential (Normal) Mode Noise and Common Mode Noise-Causes and Measures

The point I take away from that article as well is that the layout of a PCB could arguably accentuate other noise issues in the system (there is part about loop and area impacting upon radiated noise for example).
There are different types of noise and interference mechanisms. In most modular synthesizers most of the interference comes from the modules themselves rather than from external radiation. You can consider each one separately and then apply the principle of superposition to see the net result, but it helps to quantify each and not waste time addressing ones that do not contribute much.

User avatar
Rob Kam
Common Wiggler
Posts: 66
Joined: Mon Jul 09, 2012 3:16 am
Location: London, UK
Contact:

Re: Explanation of how modules "putting noise on the ground" bus?

Post by Rob Kam » Fri Jun 26, 2020 11:25 am

My $0.02: Ensure that the analog ground, digital ground and chassis are completely separate, (p. 63 of Troubleshooting Analog Circuits by Robert A. Pease, Newnes, 1991) and avoid ground loops, (p. 103). However non-banana patch cables are gratuitous ground loops.
Last edited by Rob Kam on Fri Jun 26, 2020 4:13 pm, edited 1 time in total.

TL072
Learning to Wiggle
Posts: 19
Joined: Sun Sep 06, 2015 3:42 pm

Re: Explanation of how modules "putting noise on the ground" bus?

Post by TL072 » Fri Jun 26, 2020 3:30 pm

Troubleshooting Analog Circuits by Robert A. Pease seems to be on Archive.org

User avatar
Rob Kam
Common Wiggler
Posts: 66
Joined: Mon Jul 09, 2012 3:16 am
Location: London, UK
Contact:

Re: Explanation of how modules "putting noise on the ground" bus?

Post by Rob Kam » Fri Jun 26, 2020 4:14 pm

That's excellent, I've stuck the link in my post above.

User avatar
Graham Hinton
Super Deluxe Wiggler
Posts: 3221
Joined: Fri Jul 02, 2010 6:28 pm
Location: England
Contact:

Re: Explanation of how modules "putting noise on the ground" bus?

Post by Graham Hinton » Fri Jun 26, 2020 5:04 pm

Rob Kam wrote:
Fri Jun 26, 2020 11:25 am
My $0.02: Ensure that the analog ground, digital ground and chassis are completely separate, (p. 63 of Troubleshooting Analog Circuits by Robert A. Pease, Newnes, 1991) and avoid ground loops, (p. 103). However non-banana patch cables are gratuitous ground loops.
While I have much admiration for Bob Pease's pioneering work and achievements at National Semiconductor you should not take his remarks out of context. He had as much experience of modern complex audio systems as I have of semiconductor fabrication.

Not only is this advice wrong it is impossible to implement in a large modular system with distributed power. It is only relevant considering a single pcb. For years the standard advice (from non-audio people) has been that analogue grounds and digital grounds should not meet, but they have to somewhere unless you can keep all the circuitry opto-isolated. ADC and DAC data sheets have shown analogue and digital grounds with the instruction that they should only be joined at one place. So what do you do if you have several on one board and multiple boards? There is no answer.

Analog Devices have recently updated the recommendation and explained that the reason the ICs have separate grounds is because of the voltage drops in the tiny bonding wires inside the ICs and actually they are supposed to be joined outside the IC. That's what happens when you have fresh graduates writing application notes on subjects that they have no experience of and quote text book rules.

Most of the text books on noise have been written from the perspective of industrial measurement systems amplifying tiny signals from thermocouples in a harsh electrical environment. That isn't the problem we have.

You can't completely separate analogue, digital and chassis grounds in a modular synthesizer. Most don't have a chassis ground anyway even though they should. You can't avoid ground loops, you have to learn how to deal with them.

User avatar
Rob Kam
Common Wiggler
Posts: 66
Joined: Mon Jul 09, 2012 3:16 am
Location: London, UK
Contact:

Re: Explanation of how modules "putting noise on the ground" bus?

Post by Rob Kam » Sat Jun 27, 2020 6:32 am

Rob Kam wrote:
Fri Jun 26, 2020 11:25 am
My $0.02: Ensure that the analog ground, digital ground and chassis are completely separate, (p. 63 of Troubleshooting Analog Circuits by Robert A. Pease, Newnes, 1991) and avoid ground loops, (p. 103). However non-banana patch cables are gratuitous ground loops.
That should have included "except for connecting AGND and DGND at the PSU".

User avatar
Graham Hinton
Super Deluxe Wiggler
Posts: 3221
Joined: Fri Jul 02, 2010 6:28 pm
Location: England
Contact:

Re: Explanation of how modules "putting noise on the ground" bus?

Post by Graham Hinton » Sat Jun 27, 2020 6:50 am

Rob Kam wrote:
Sat Jun 27, 2020 6:32 am
That should have included "except for connecting AGND and DGND at the PSU".
No. That only works for a small self powered device, like a 19" FX unit or outboard, where the distances involved are a few inches. It cannot be implemented over a large modular system without creating more problems and it cannot be implemented on existing modules without changing the power connector.

You are repeating text book information without thinking it through. What do you do if there are several PSUs?

User avatar
Rob Kam
Common Wiggler
Posts: 66
Joined: Mon Jul 09, 2012 3:16 am
Location: London, UK
Contact:

Re: Explanation of how modules "putting noise on the ground" bus?

Post by Rob Kam » Sat Jun 27, 2020 7:56 am

If it's required on the module, Eurorack 16 pin and Dotcom 6 pin MTA connectors provide for +5V. Extended star wiring seems to be the sensible option but then every cable with a signal ground included is another loop.

Most of the references I can find, are really talking about PCB (if not IC) layout, not large/multiple modular systems. Please, if you could recommend some source of better information on powering modular systems?
Last edited by Rob Kam on Sat Jun 27, 2020 11:54 am, edited 1 time in total.

User avatar
Graham Hinton
Super Deluxe Wiggler
Posts: 3221
Joined: Fri Jul 02, 2010 6:28 pm
Location: England
Contact:

Re: Explanation of how modules "putting noise on the ground" bus?

Post by Graham Hinton » Sat Jun 27, 2020 9:32 am

Rob Kam wrote:
Sat Jun 27, 2020 7:56 am
If it's required on the module, Eurorack 16 pin and Dotcom 6 pin MTA connectors provide for +5V.
They are not widely used and don't solve the problem, they just inject current from the digital circuitry into the existing 0V.
Whether you have a digital power rail or locally regulate the positive analogue rail the same current still goes into the 0V.
Extended star wiring seems to be the sensible option but then every cable including signal ground is another loop.
Again, you are repeating text book information. Please stop and think what such statements mean. Star and loop are vague terms that are often misunderstood and used incorrectly.

Star could mean anything because it does not define the resistances to the centre. What is going to be better: sixteen 0.5mm2 cables a metre long fanning out from one point or a 2 metre long busbar with sixteen short 1mm2 cables along it's length? The answer is the one with the lowest resistance between any two modules and it's not the one that physically looks like a star.

Ground loop is often used to describe common impedance coupling problems. A true loop that picks up magnetic fields is a minor problem in systems with no gain. Consider how much gain you have to apply to a ribbon microphone or guitar pickup to hear a hum. It also depends on how large the loop is, if the mains Earth wiring becomes part of a signal return path and you use outlets in different locations in a building then you will get a hum.

Consider a stereo hi-fi cable: you have two screened cables that are connected together in the the equipment each end. Does that form a loop? Yes, but the two cables are moulded together so the area between them is minimum and it can be used with a magnetic pickup with a lot of gain.
Most of the references I can find, are really talking about PCB (if not IC) layout, not large/multiple modular systems. Please, if you could recommend some source of better information on powering modular systems?
You aren't going to find a text book on this subject because by definition textbooks are following rather than leading practise and the people actively doing it are too busy to write them. I haven't written one and I'm not going to any time soon.

AES48 is the recommended practise for audio equipment grounding and shielding. It is applicable to unbalanced, patchable systems with distributed power, but that requires interpretation. Search for my posts here on AES48 and you'll get the gist.
Shields & Grounds AES Journal special excerpt ($15)
AES48-2005: AES standard on interconnections. ($30)
The Journal reprint is worth getting because it contains Neil Muncy's original paper on "the pin 1 problem" (referring to incorrect XLR wiring), beware that some of the articles are US-centric and do not consider mains power differences in other countries. The actual standard is terse and probably disappointing for $30.

The bottom line of it is that cable screens should be connected to chassis ground and not 0V. Nearly all modules have no chassis ground and yes they create "ground loops" all over the place, but they usually causes shifts in the 0V potential differences rather than pick up hum, unless of course there is some amplification.

Post Reply

Return to “Music Tech DIY”