Home Computer Audio Asylum

Music servers and other computer based digital audio technologies.

[ Follow Ups ] [ Post Followup ] Thread: [ Display  All  Email ] [ Computer Audio Asylum ]
[ Alert Moderator ]

Initial results/ impressions after one day. (With lots of questions still too be answered)

Hi Emial Tim,

These are my experiences after one day of clamping ferrites. It will be al long copy text and I write this also for others too because I am not educated in electrics and (digital) electronics. So it’s likely that i’m not aware of ‘white spots’ in my knowledge on this. (a white spot is an area of knowlegde that I don’t have, but also I don’t know that there is something I don’t know. If I knew I don’t know, than it’s a black spot. And I can start looking for the knowledge)
So too anyone who reads this, feel free too comment, too correct, and too improve.

So in the text below, I will give some background on what I did and why I did it. So anyone can correct or comment on my approach and can correct false assumptions and lack of knowlegde and give tips for other better approaches.

* Why wanted to try ferrite camp on beads
I wanted too experiment with filtering HF noise on the power supply lines too the mobo. Why?
Using a linear PSU too power the MoBo has a significant positive effect on sound quality. It is said that this is because less noise and ripple.
Putting decoupling caps (smoothing caps) on the P4 power lines has a positive effect on SQ because these caps probably smooth the ripple and may also shunt some HF noise. But using large capacitance too create a current reservoir too smooth ripple, large capacitance will not shunt HF frequencies. Putting litlle by-pass caps on the bigger caps helps somewhat. But still is not very effect on the very high frequency’s I want to filter.
I thought about filtering with 2 nd order L/C filters or R/C filters. But calculations too see if the these 2nd order filters will not start too ring, are much, much too difficult for me.

A more simple approach would be a combination of ferrites and capacitance. As suggested in the article from ferrite manufacturer Steward. Also google on” The Use Of Ferrites In EMI Suppression’ and one will find many more articles on this.
Ferrites provide a high resistance too HF frequency’s with much less chanche on ringing . Ferrite’s provide a high resistance for frequencies in the range from 50 mHz too 500 mHz. For exact data see the data sheet of the specific clamp of the ferrite bead that is chosen.

Things too keep in mind when using Ferrite’s
When reading article on The Use Of Ferrites In EMI Suppression it is pointed out that feeding DC current through a ferrite, will degrade the effect that a ferrite bead has on HF frequency’s. When too much DC current flows through ferrite, it becomes saturated and the filtering effect on HF frequency’s is lost or much less. See articles that discuss the effects of ‘DC & Low Frequency AC Bias Effects And Saturation’. Also inmate Bibo01 was probably pointing at this.


So here’s what I tried so far.
(with some background info, as too ‘why’and ‘how’)

* Starting with the electrical 230 source in my house.
In The Netherlands every home gets 3 times 230 Volt AC delivered. These three 230 AC voltages are delivered with a fase-shift of 120 degree’s. One can recognize these three fase groups, through the grouping of the spurs in the meterboard. In the meterboard one can see that the spurs are arranged in three groups of spurs. Every group of spurs lurks from one of those three 230 AC voltages.
When installing a dedicated spur, it’s wise too take the dedicated spur from the most clean 230 AC voltage fase. In my case I toke it from the 230 AC voltage fase that only feeds the laundry dryer and the washing machine. Nothing else lurks of that fase. So it’s the most clean 230 AC volt fase too put the dedicated audio spur on. Provided that the washing machine and laundry dryer are not running this is the most clean fase in my house. How much HF pollution already is delivered at my doorstep through the electrical net, I can only guess, But putting 2 big ferrite beads on this dedicated 230 AC voltage line coming from the dedicated spur, makes no audible difference at all.

* 230 connections too the used gear.
All cableling of the dedicated audio spur and also all power cords are shielded. The shields are only grounded at one end of the cable too a real earth.
Only the active Klein & Hummel speakers and my Lavry Black DAC are on this dedicated audio spur. The Lavry DAC has Epcos line filter in the power cord. This way I try too prevent the Lavry spoiling the dedicated audio spur. The lavry has a switching PSU inside, so one never knows. Putting the lavry on another spur makes makes no sound quality difference. So probably the Lavry is not pollution so much back inside the dedicated audio spur.

* I use two linear PSU’s and 2 pico’s
They are arranged in 2 groups. A dirty group and a clean group. Just a Cics suggests doing with 2 ATX PSU’s. I use that same concept (dirty and clean), but using 2 pico’s
- clean section: linear PSU -> Pico 160 XT on the P24
- dirty section: linear PSU -> Pico 200 which powers the SSD and USB
(the P4 PSU line is taken directly from the linear PSU and bypasses the Pico).
Both linear PSU’s are on the same 230 AC outlet and on the same 230 AC spur with a real earth. So the two linear supplies both can see each other through the 230 Volt AC connection and also through the safety earth connection.
I don’t know if the 2 linear PSU’s also can see each other trough the DC side through the mobo somehow.
Putting the 2 linear PSU’s on 2 different spurs with real earth connections, has no positive effect on sound quality.
Putting the 2 linear PSU’s on a clean kitchen spur (with only the dishwasher on it) makes no difference in SQ.

* other sphericals connected with might inject HF noise.
- NAS
I use a NAS with might inject noise onto the MoBo.
Connecting the LAN cable makes no difference in sound quality.
This might be prevented, because LAN sockets are transformer decoupled by design. May be this feature prevents HF noise. But being AC, the HF noise should be able too get trough this transformer.
So I’m a little puzzled here.

Too my utmost surprise: connecting my Sony TV too the VGA port, also makes no difference in sound quality. I had expected the TV too inject much HF noise into the MoBo.
However the TV is earthed trough a real earth connection via the coax cable. The coax cable has a real earth connection in the meterboard when entering the house.

* HIGH QUALITY optical out to Lavry DAC made a real improvement
As suggested by Cics using an optical connection which provides galvanic isolation between DAC and MoBo yields a significant SQ improvement. But only when high quality toslink cables are used.
Using the standard toslink optical cables that came with my RME sound cards makes no difference at all. Standard quality toslink cable sound the same as balanced AES/EBU from my Lynx AES16 PCI card.
Surprisingly by far the best is: optical out by means of a HIGH QUALITY optical toslink.
Just as Cics recommends in his cMP recipe.

* Why I tried ferrites in this setup.
So I still had the nagging feeling that there still must be al lot of HF noise around or a ground loop somehow. Otherwise: why would using a high quality toslink connecting make such a significant SQ improvement. Right ?

This is why I wanted too see if blocking HF noise traveling around, would yield any sound quality improvement.
So I just started (more or less without any rational strategy) placing ferrites on power cords.

* Ferrite clamp on P4
To my surprise placing a large ferrite clamp on the P4 psu line coming from a linear PSU (!!), made a significant SQ improvement.
More improvement than placing smoothing caps on the P4. So definitely worth taking the trouble.
So is it a lousy linear PSU? I don’t think so as it already had made a sound quality improvement when not using the ATX P4.
Placing ferrites just makes a second significant SQstep. In my setup: equal too the one not feeding the P4 from a Aerth watts ATX switching PSU.

* many ferrite clamps on individual P24 voltage lines.
Just one big ferrite clamp on the P24 has hardly any effect on SQ.
But clamping many each on individual + 3.3, +5, + 12 volt wires (each combined with 1 black 0 Volt line) also makes a nice SQ improvement.
The P24 extension cable has now 9 ferrite clamps on it. 5 clamps on each individual red 5 volt DC lines, 3 on each orange 3,3 line and 1 on the combined yellow +12volt lines.

* DC bias.
I probably make a lot DC bias faults by clamping ferrites with out knowing the currents through each wire. so I want to start measuring the current in all lines of the P24.
Knowing all currents in each wire from the P24 extension cable, I might be able too make better +/- combinations which cause less DC bias.

* Is there really much more HF noise in my setup than others?
So I really don’t know if my setup has lots of HF noise traveling around, compared too other cMP setups. So if anybody sees any basic flaws in my setup, please let me know what you think.
As I can’t think of any basic flaws which would create a lot of HF noise in my cMP setup or setup flaws that create unwanted ground loops.

* optimizing the use of Ferrites
Until now I just clamped around with no real strategy. If anybody has tips on where and when ferrites are more effective, let me know. Any suggestions welcome.

* adding caps on both side?
What about placing extra caps on both side of the ferrites, to shunt HF frequency’s that see the high resistance of the ferrite ahead and thus will prefer traveling through the caps?

I found it encouraging that TheoB also found the same SQ improvements in his setup.
Especially because I know his Martin Logans are extremely revealing but neutral, electrostatic speakers which will let Theo hear every minute change in ambiance, micro details, etc.

So everybody start clamping!
Ferrites only cost 1 to 3 euro each.
Just try them and report back.

Mark

This post is made possible by the generous support of people like you and our sponsors:
  Parts Connexion  

Follow Ups Full Thread
Follow Ups
  • Initial results/ impressions after one day. (With lots of questions still too be answered) - hfavandepas 08:32:10 06/02/11 (0)
FAQ

Post a Message!

 Forgot Password?
Moniker (Username):
Password (Optional):
  Remember my Moniker & Password  (What's this?)    Eat Me
E-Mail (Optional):
Subject:
Message:   (Posts are subject to Content Rules)
Optional Link URL:
Optional Link Title:
Optional Image URL:
Upload Image:
E-mail Replies:  Automagically notify you when someone responds.

[ Top Top of Page ] [ Contact Us ] [ Support/Wish List ] [ Copyright © 2024, Audio Asylum, all rights reserved ]
[ General ] [ Speakers ] [ Tubes ] [ Vinyl ] [ Digital ] [ Hi-Rez ] [ Video Asylum ] [ Cables ] [ Tweaks/DIY ] [ Music ] [ Films ]