|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
76.122.133.81
In Reply to: RE: Effect of ferrite beads on dc power lines too P24, P4 and/or too Pico PSU’s posted by hfavandepas on May 27, 2011 at 15:03:22
Based on some dissatisfaction with I2S and the Jkeny's post in Diyaudio on spdif vs I2s I decided to switch back to spdif as an experiment. I preferred spdif immediately and am trying for my 3rd day just to see (hear) if I respond favorably over time. I have so far. Anyway I went out and got more ferrites and placed them over my spdif cable (I should try over I2S cables now that I think about it). Yes same impact. At 1st I was worried that I may be throwing out some bits with the noise but after browsing some of your references it seems as if ferrites take care of 50-500 mhz and I believe spdif (esp with 16/44 signa) is much lower than that. I don't know if it works on glass toslink but maybe you could try it. Ferrites ... the gift that keeps on giving.
Follow Ups:
Hi Theo,
I have no experience myself with I2S, but I can understand the attractiveness of the concept to separate the clock signal from the audio-bit stream. But that is from a conceptual point of view. From a practical, real world point of view I doubt if I’m able too execute it properly myself. Especially knowing how VERY DIFFICULT (!!) it even is for professionals too accurately send, transport, receive and process clock signals. I don’t think I can technically execute that properly myself in such a way that it will outperform a proper executed S/Pdif connection. But these are just my thoughts on this matter. I have no real world practical experience with it.
Also, even on the best studio DAC’s money can buy (Prism, Lavry, Weiss, etc) I don’t see an I2S input. If this where a better way of doing things, it would certainly be fitted by these and other companies.
So I concentrate on more simple things which I can execute properly myself given the limited set of tools I have (lack of knowledge, skills and proper equipment).
In frequency/impedance graphs provided by the ferrite manufacturers I see that ferrites have a bell shaped resistance curve that peaks somewhere between 100 – 300 mHz. The official S/Pif specs say that the frequency can be between 100kHz and 6Mhz. A CD player uses 3mHz.
But still I don’t put ferrites on un-shielded S/Pdif cables. Given the theory what an electrical digital signal is (square block pulses) I fear malformation of the block shapes (misshaped eye patterns)
Of course ferrites can reduce HF noise traveling on the shield of an S/Pdif cable, but that HF noise shouldn’t be there at the shield in first place. It would be my strategy too battle that (possible) noise at some other place.
Ferrites on a glass toslink ???? :-)
I know you said you didn’t feel for trying too ad some caps before (and some smaller behind) the ferrites because your P24 line was already fully occupied with ferrites.
On the risk of being ‘pushy’, I again will give the advice too try and place some caps.
Especially with use of an P24 extension cable it’s very easy too do.
You don’t even have too so solder !! Just use a set of needle-nose pliers too push the metal legs of the caps in place between the plastic outside of the P24 connector and the metal wire snap on inside the plastic connector. (click on pictures below too enlarge and too see how easy this can be done) No soldering needed ! I just check with my multi-meter too see if the metal legs make good contact. That’s all.
Van on filtering, demping, ect
Van on filtering, demping, ect
Tomorrow a variety of some 30 oscon caps will arrive. Caps have such a profound and positive effect on sound quality difference in my setup when placed before and after the ferrites, that I will start searching for the best combinations before and after the ferrites. And than eventually, when I found the right combination, I will solder the caps onto the P24 before and after the ferrites.
By the way: I now know why I can cold boot into 110 mHz busspeed. These days I experimented a lot with caps on and off the P24 before and after the ferrites and I discovered that it is the caps on the P24 mobo socket.
For quite some time already I used as much extra caps on the P24 as the Pico or Antec ATX would tolerate on the 3.3 and 5 volt line. The Pico doesn’t tolerate much. Only some 330 uF per voltage. But the Antec ATX tolerates very large values of 4700 uF per voltage. But I never did try how low I could go. I just put them there as extra current-reservoir and noise shunt. Although I didn’t hear any SO improvement at all (which I had expected because on the P4 extra caps will give some SQ improvement). But I just left the caps on the P24, because I also did not hear any sound quality degradation also.
Only when I started realizing that the lower the swithing-speeds of logic on the MoBo, the lower the HF noise production there will be on the MoBo, I tried how low I could go with busspeeds.
But I never realized that it was due too those caps on the P24, which I already have on the P24 for over 1,5 years now.
Mark
fully optimized cMP2 PC -> ESI Juli@ -> Van den Hul Optocoupler MkII-> Lavry Black DA10 -> XLR Mogami Gold -> Klein & Hummel O300
I guess you are saying ferrites on spdif cable are not recommended but what is the speed of a 16/44 propagating through spdif? If it is < 10mhz ( & I think it is < 10) I tried it anyway and it seems to work.With respect to caps on p24 right now I have zero room but since you so politely encourage me to try I may get a p24 extension and try.
:)
Edits: 06/30/11
Hi Theo,
‘not recommended’ Well, I don’t dare too do any recommendations.
I’m no expert nor educated or work in digital audio electronics, I was more of thinking out loud.
But if it works, than it works.
I trust your equipment and ears.
I only wonder where that noise is coming from.
That is why I thought about fighting that noise at some other place.
Although I don’t have any clue where too look else.
For the S/Pdif specs I just googled and I found them quickly.
This was what google presented too me as first link.
http://www.epanorama.net/documents/audio/spdif.html
Mark
fully optimized cMP2 PC -> ESI Juli@ -> Van den Hul Optocoupler MkII-> Lavry Black DA10 -> XLR Mogami Gold -> Klein & Hummel O300
Based on your comments I took out the spdif ferrites and listened and I prefer them in by a wide margin. Interestingly I tried listening @ 192 sample rate and with ferrites it seems dull and missing info but not @ 44 sr.
Anyway these ferrites are something special in audio pc's. What I don't understand is the original thread from AA that you referenced in your original post on ferrites. Most posters there said it was not recommended to use ferrites period. I guess not knowing any better in these eletrical pc things I just have to try everything.
Hi Theo,
That thread was about putting them on digital lines that transport usb and/or spdif signals.
So I fully agree not putting them on these digital lines unless there is a very very dirty component on either side that might throw extra additional HF dirt into the data line.
But that thread made me aware of that i also see them sometimes power lines (my dell switching loader, some wallwarts, ect) I suddenly realized that may be ferrites might possibly also work on the power lines comming from the ATX.
Mark
fully optimized cMP2 PC -> ESI Juli@ -> Van den Hul Optocoupler MkII-> Lavry Black DA10 -> XLR Mogami Gold -> Klein & Hummel O300
Hey Mark,
I think that is why power cords can make a big difference depending on how they handle emi.
Some manufactures agree i think and when I was fitting my dirty supply for ferrites I noticed that there was already a big one built into the psu. Here is a link to the review and on the sixth picture down you can see the ferite. There is as much sticking out as there is inside the case. Might be why I liked this psu.
No one here remembers the bending of our minds
Hi Dawnrazor,
Thank you for this.
I already was interested for quite a while in ways too filter the output of a switching ATX PSU for my cMP setup. But using a classic filter concept containing a combination of chokes and caps is very hard too calculate. Big chance one will end up with a filter that rings/resonates.
Fortunately I red that the chance of ringing is much less when a combination of ferrites and caps is used. However there still is a chance of ringing/resonance, it is much less and much easier too control.
As ferrites are lossy components the needed calculations are also somewhat lossy. A rough approximation is enough. At least: this is what I roughly understand from articles on the net from TI, Alterra, Analog Devices, EMAC and others.
I see it is an Enermax PSU.
RME used too recommend Enermax PSU’s on there website until some years ago.
When RME changed their website these recommendations didn’t re-appear on their new website.
So I think its not a bad choice at all.
Mark
N.B. Also XP optimizations tips did not re-appear on the new RME website.
I still use one of RME’s XP registry tips too disable CD-rom ‘polling’/ autostart in registry.
fully optimized cMP2 PC -> ESI Juli@ -> Van den Hul Optocoupler MkII-> Lavry Black DA10 -> XLR Mogami Gold -> Klein & Hummel O300
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: