|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
95.96.177.246
In Reply to: RE: Common Mode Noise posted by Mercman on January 24, 2017 at 10:56:04
Hi Mercman,
Thankx for posting this. It highlights a 'hidden' problem that I also had in my system and probably also many inmates may have in their sound systems without noticing. As for me, I only discovered this hidden problem when I made unplanned changes in power connections because I had too dismantle and re-install my sound system when my living-room needed some fresh painting.
In the end solving noise problems and solving related ground loop problems really paid off in my system!! So I definitely want too encourage other inmates too have a close look at their systems. May be similar sound quality improvements can be made in their systems.
On the other hand, finding and eliminating ground loops and suppressing associated noise problems was very time consuming and a lot of work. I spend much time on reading tutorials and papers, much on experimenting, buying parts, installing them, etc, etc.
So I decided Mercman's post is a good reason for writing down this very, very loooong post about what I tried in my system. Hopefully it will give other inmates an impression if 'hunting' for ground loops and associated noise problems in their system, might be worth the trouble.
* But first some considerations when I red your post.
While your friends solution focuses on suppressing these noise currents through filtering, you also focus on the pathway these noise currents flow. You point at a video https://www.youtube.com/watch?v=sMmdDLFP4So where Juergen Reis (MBL) discusses the 'back door leakage currents' that are found in complete systems which can have a negative impact on the sound of our systems.
Don't know why Juergen Reis (MBL) calles them 'back door leakage currents' but these 'back door leakage currents' are also better known as: ground loops. Don't know why Juergen chooses too call them: 'back door leakage currents'. (Probably just too make him look smarter and make the problem look more mysterious)
While adding filters too suppress noise is one solution, I was very happy that you also considered pointing at the pathway's these noise currents may use. Changing or eliminating these pathway's may be much more effective than putting a filter on it.
And what to think of: not creating those pathway's in the first place? (through prober system design and/or proper balanced interconnection)
However avoiding or eliminating those pathway's is easier said than done. Especially since many inmates nowadays will have sound systems that are (inter-)connected too:
1. a computer, TV-set, cable-box, streaming device, wired internet, etc, etc.
These consumer grade devices mostly use noisy PSU's and use (potentially problematic!) unbalanced connections.
2. coax-cables from cable-companies or antenna-systems that may introduce an extra (potentially problematic) pathway too the electrical safety ground.
3. consumer equipment that has a build-inn design defect also know as the 'Pin 1 problem'.
Combining all 3 points together make an ideal situation for creating (unintentionally) all kinds of ground loops causing minor (or bigger) sound quality problems.
Like me, most inmates probably will assume that a sound system has no noise problems caused by ground loops, as long there is no 50/60 Hz hum coming from the speakers. But unfortunately ground loops ('back door leakage currents') come in all sorts and can have all kind of (U)HF noise frequencies having negative impact on the digital parts of the sound system.
In his tutorial 'A Ham's Guide to RFI, Ferrites, Baluns, and Audio Interfacing', Jim Brown (vice chair of the AES Standards Committee working Group on Grounding and EMC Practices) makes a bold claim. He suggests that the most common way hum, buzz, and noise enters the circuit board of our music equipment are caused by design defects. Somehow, somewhere the cable shield has a (unwanted) connection too the circuit board ground. On page 5 Jim Brown writes: "Nearly all consumer equipment, including even the most expensive "high futility" gear, is built with this problem. Virtually all computer sound cards have these problems. So do most RS-232 interfaces and nearly all HAM equipment".
Although focused on HAM equipment Jim Brown's tutorial was a must read for me. It helped me a lot. It can be downloaded from: http://audiosystemsgroup.com/RFI-Ham.pdf
Just like you discovered, I also discovered that Jim Brown was right. The sound coming from my system could be improved by a significant amount when attention is paid not to create unwanted ground loops in the first place and/or put filtering on those loops if they can't be avoided.
As already suggested by Soundcheck, cleaning up the cable mess, is a good starting point. But that did not do anything in my system. My system setup is fairly simple and there are not a whole lot of cables.
System setup:
Coax cable -> cable box -> HDMI -> HD TV -> HDMI -> HTPC -> Cat 6 cable -> Merging Hapi -> XLR AES/EBU -> PMC TwoTwo.8
- the HTPC has two network interfaces. One NIC is connected to the Merging Hapi. One NIC connects to a wired LAN cat 5e cable for internet connection.
- With help of a Neutrik impedance transformer the cable box RCA S/Pdif output is connected to the AES/EBU XLR input on the Merging Hapi.
(I use this connection when watching movies on TV. I like too hear the movie sound through my sound system)
- At the cat5e cable which connects too the internet I use a Baaske internet isolators.
Although my sound system is quite simple, I only discovered by accident that my system could sound significantly better. Here's how. For many years I powered my HTPC from a separate power socket across the room. I feared the noise that the ATX computer PSU might throw back into the power line thus spoiling the power input of the amplifiers in my Klein & Hummel O300 active speakers. Only by coincident, I found out that my system definitely sounded much better when the noisy ATX computer PSU was plugged into the same power strip that also powers the active speakers and the Merging Hapi.
Looking back, I know now that I made a huge beginner mistake. Like most computers, also my HTPC has a 'Pin 1 problem' by design. With help of this design defect I had created a massive ground loop that added more noise than the ATX PSU would do by itself.
So now the ATX computer PSU, the active speakers and the Merging Hapi unit are all powered from the same power strip. However, I've added an Epcos power line filter into the ATX PSU power cable. The Epcos power line filter suppresses the noise that the ATX PSU throw's back into the power strip, which again improved the sound by a significant amount !
Because of this success I got heavily infected with 'ground-loop-hunting fever'.
I began reading many internet papers/tutorials on ground loops. Beside Jim Brown's tutorial, I would also recommend reading:
'Understanding, finding & eliminating ground loops in audio & video systems' written by Bill Whitlock, (http://web.mit.edu/jhawk/tmp/p/EST016_Ground_Loops_handout.pdf)
Although it's more of an advertorial (and not exactly a white paper) it still provides in-depth high quality information
After reading Bill Whitlock's paper, I now looked suspiciously at the coax-cable connected too my TV and cable-box. This coax-cable is also connected too the earth's safety ground in my breaker panel closet, possibly creating another ground loop (with help of the unbalanced connections on the TV and the cable box). But unplugging the coax-cable did not improve sound quality.
Also completely disconnecting and removing the TV and cable-box from my sound system (power + interconnects) and running my computer 'headless' did not improve sound quality.
Because Jim Brown also looks at cleaning 'Ethernet trash' in his tutorial, I also tried cleaning 'Ethernet trash' in my home. First I installed a Baaske LAN isolater, but that did not have any impact on sound quality (waste of money). So like Jim Brown suggests I tried too suppress the noise coming from the PSU's that power my LAN network equipment.
In my breaker panel closet I have five cheap switch-mode wall wart style PSU's that power: a cable-internet-router, a managed Netgear ProSafe switch, a WiFi-acces point, a Philips Hue bridge and a NAS.
I decided too install a whole new 6 outlet (earthed) power strip. Onto the strip's power cord (AC side) I installed an Epcos power line filter too suppress noise that those 5 cheap wall wart style PSU's will throw back into the 230 V AC power line.
On the DC side I installed the same common mode choke coils like your friend also installed.
Unfortunately, the sound quality only improved marginally. I don't think this slight improvement was worth the trouble.
But I just red Ryelands tip on removing connections on/to pin 4&5 and 7&8. Ryelands always has tips that work very well. 50% less wires that can create a ground loops and/or provide a path for noise, And also 50% les antenna's to pick up noise. So I'm going to try Ryelands tip ASAP
So what did I learn from this 'ground loop hunting'?
Computers do not only cause problems because they come with a noisy (ATX) PSU. Computers also come with a build-inn design defect: the so called 'Pin 1 problem'.
I don't know which one of the two is more worse. But when combined together (Pin 1 problem + the noisy PSU) it can have serious negative impact on sound quality.
Especially in systems that use unbalanced connections. Ground loops introduced by using unbalanced connections will create noise and also allow noise too travel almost system wide.
Happy 'ground loop' hunting.
I'm curious how other inmates eliminated ground loops and suppress associated noise.
So please do report your findings here on the audioasylum.
Mark
Mac (with Audirvana + ARC2 Acoustic Room Correction plugin) -> Cat 6 cable -> Merging Hapi -> AES/EBU XLR Vovox SD -> PMC TwoTwo.8
Follow Ups:
I used balanced connections in the main system. In the HT, I once needed to use one of the Jensen isolators on the incoming cable signal to tame a nasty ground loop.
And I found this site which references Jim Brown and many other sources. Now I have much reading to do, and much investigation into some of the gear I own.
Thanks for posting this.
Regards,
Presto
Rane Note 110 - Sound system interconnection guide.Priceless.
Edits: 01/28/17
Hi Mark,
Thanks for your post.
Which Epcos power line filter did you put into the ATX PSU power cable?
Hi BiBo01,I use these Epcos filters because the are readily available and not very expensive.
But many others company's manufacture same types.Epcos EMC Filters Series/Type: B84111A
See http://www.mouser.com/ds/2/400/b84111a-773410.pdf
(Since 2013 Epcos stopt production.)On the ATX PSU power line I put a filter which can handle 3A amps.
2 years ago I still used active Klein & Hummel O300 monitors which have 3 amplifiers build inside with a total max consumption of 435 Watts per speaker. Sow on the power lines too the K&H O300 I put an Epcos (6A amps rating) reversed (line side now is: side with the caps)
You find the schematic's all over the www and they are used by many other filter company's like Schurter, etc.
I first saw the schematic's discussed in Ivor Catt e.a. book Digital Hardware Design, by Ivor Catt, David Walton, Malcolm Davidson.
http://www.ivorcatt.org/digital-hardware-design.htm
Chapter 14 page 84 and on.The Epcos filters I use have a standard, generic design which does not filter the ground-wire.
So depending on where the noise source is located, the ground-wire may still facilitate loops.Mark
Mac (with Audirvana + ARC2 Acoustic Room Correction plugin) -> Cat 6 cable -> Merging Hapi -> AES/EBU XLR Vovox SD -> PMC TwoTwo.8
Edits: 01/30/17
What you have said highlights the basic set up principle in high quality audio replay, analog or digital. Congradulations for doing it.
Which leads to another point. An inmate says he thinks he has a common mode noise problem. He sticks a choke on the PS's lead. He says he has improved the sound. But which is the chicken and where is the egg? For all we know, the new high performance power supply has revealed a flaw in his previous 'optimised' setup, and the problem is elsewhere in his replay chain. So, the hunt for another stick on plaster can start again.
"But which is the chicken and where is the egg?"
Good Question !
Lets consider the chicken = noise source.
The eggs = the noise flow.
First off: The noise source (chicken) does not necessarily needs too be the Sonore Signature PSU.
There may be other noise source candidates along the loop.
Also: system parts that suffer from common noise traveling on the circuit board may be more than 1. The microRendu may suffer, the USB data transport may suffer, the DAC may suffer.
Lets trace the noise flow first (follow the eggs).
When doing so it must be considered that current ALWAYS needs too return to it's source.
So which return path can be used by the noise too travel back?
Is differential mode noise?
(In one wire the noise flows 'up', in the other wire the same noise flows 'down').
Lets hope it is common mode noise.
I don't like too think of an 1.400 dollar Sonore Signature PSU that generously produces differential mode (U)HF noise.
Or is it common mode noise?
(Both wires same noise, same voltage, in same direction).
What direction is the common mode noise flowing?
Towards or from the microRendu?
All I know about Mercman's system is:
? -> Signature PSU -> 2 PSU wires -> microRendu -> USB cable -> Ayre QX-5 Twenty -> ?
? -> UTP LAN cable -> microRendu
- Mercman's friend put's an impedance Z (AC resistance varying according to frequency) on both PSU wires by means of a ferrite choke.
- The PSU wires are between the microRendu mini-pc and the Sonore Signature PSU.
- The impedance on the PSU wires reduces/suppresses the flow of (U)HF noise in the PSU wires
- Mercman enjoys better sound quality.
When trying too find the electrical return path for the common mode noise, I will have too drawn a circle that always includes:
A: the Signature PSU -> 2 PSU wires -> the microRendu
(no matter if the common mode noise is flowing towards or from the microRendu)
B: And the (U)HF noise source. This not necessarily needs to be the Signature PSU.
When trying too draw a circle that closes a loop, it is easy too see this only can be done through:
Circle A: a loop via the USB cable + power cords attached too system parts
Circle B: a loop via the LAN cable + power cords attached too network equipment and systems parts.
Suppose the noise travels back using circle A.
* Bridging the gap.
May be Mercman could also offer an alternative path for the noise too flow along in an attempt to bypass the circuit board inside the microRendu.
He could try too connect a wire between the PSU casing and the microRendu casing. It might work. But also it might not work and even making things worse by creating an extra loop.
All depends on where the noise source is located and if there is also a 'Pin 1'-type of design flaw present in any involved equipment.
* Hunt for another stick on plaster.
Most likely you are right about this.
Because the common mode noise needs an opportunity to get onto a power cord and travel back through the electrical system too the Sonore Signature PSU too close the loop .
Common mode noise on PSU (ref grnd) wire -> (ref. grnd) circuit board microRendo -> (ref. grnd) wire USB cable -> (ref grnd) circuit board DAC -> etc
Somewhere along this route it has to get on a power cord and travel back too the Sonore Signature. It might use the parasitic capacitances between the power line and the equipment ground for this. But it may also exploit yet another design flaw for this.
Mark
Mac (with Audirvana + ARC2 Acoustic Room Correction plugin) -> Cat 6 cable -> Merging Hapi -> AES/EBU XLR Vovox SD -> PMC TwoTwo.8
Thanks for this post Mark. I was hoping that inmates might discuss this interesting issue..
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: