[Wijnand]

Zou misschien ook met het vermogen te maken kunnen hebben. Grotere vermogens ondanks het zeer waarschijnlijke goede ontwerp, zullen logischerwijs een grotere impact hebben dan een kleiner vermogen zoals van een DAC.

[Wijnand]

Thanks! Sowieso is het een goed om bewust te zijn dat een schakelende voeding mogelijk een negatief effect met zich meebrengt. Het blijft een kwestie van proberen 🙂

[Wijnand]

No problem 😉

Indeed just as Tobias already mentioned it is not a digital problem. It is an analog problem. Digital signals are interpreted out of the analog. So the better the shapes of the square waves, not even knowing what the best shape will be. Still I’m pretty sure they must be smooth in some way.

Personally I think the time domain is just a factor, maybe not even a very big one. I think noise in general is causing several effects.

[Wijnand]

Op zich kan dit wel. Bedenk dat er best wel wat goede audiomerken zijn die switchmode voedingen gebruiken in hun apparatuur. Waar het denk ik om gaat is dat de secundaire zijde van de Audes de schone zijde is. Met het aansluiten van Switchmode voedingen op de secundaire zijde van de Audes zou de secundaire zijde makkelijk vervuilt kunnen worden. Toch denk ik dat het met goede merken niet zo zwart wit is. Ik gebruik zelf een Moon Dac Streamer op de Audes en dit werkt waanzinnig goed. Moon maakt ook gebruik van switchmode voedingen.

Groeten,

Wijnand

[Wijnand]

Hi Agisthos, Thank you for sharing. What type of powersupply did you use? There’s actually two things that I suspect that causes be the problem. One is clock synchro which actually is the effect and two is that the ethernet signal itself is so precise and well built that the signal itself is getting too aggressive. Think about a more pure blockwave with sharper edges instead of less controlled blockwave with smoothed edges. Clocks normally need to stabilize for some time the more it stabilizes the more trouble it becomes for the DAC. I think and I also have some measurements from Jaap that proves that the DAC clock is influenced.

[Wijnand]

Thanks Paul, you’re the best 🙂 Nice to hear that moving to Sweden is done!

It’s more that most progress we have gained in the past isn’t physically touchable. We are learning and working individually together. I remember the times we started using PC’s as streamers. There were so many steps, but how to thank somebody for progress or knowledge that isn’t the same in every system? Especially when it’s not physically touchable like a powersupply or a cable drossel. The work must be done and we all have the same goal. Better sound 🙂

[Wijnand]

Hi Eric, I just looked into your forums, wow there’s a lot to find!, but also already too much to stumble through. I will register for sure and contribute when I have figured things out. The 5 boxes I have built for some friends brought great reviews and it helps me for better understanding. No network seems to be equal above a certain level and my own network seems to be the most difficult despite it’s a very simple one.

I started with streaming in 2009 with a PS Audio PWD. I will do my best to contribute on your forum as well in the future. I already wrote at least above 1000 articles on several forums to share my findings about thousands of hours testing things and equipment, sometimes it felt like a second job, which kept me very busy :-). It never payed out in terms of thanks for this or for that. Still I’m positive this must have helped many, many people and manufacturers. I have seen so many things that have changed, but it’s always difficult to claim the outcome. Although ferrites are commonly used on signal cables. The drossel cable takes it to a new level for ethernet. Thanks for sharing! Bic time! Like I always do I will first dive into the materie and test everything I can think about, after that I will start reading what others have found out. The other way around never worked for me, because it’s too difficult to understand what others hear and prevere.

[Wijnand]

Hi Eric,

Thanks! Yes, of course sharing and contributing is very important to me. At the moment. I’m still researching the effects of power systems, power supplies, routers, switches, clocks, cables, streamers, galvanic isolators, Quantum fuses and filter shokes. It’s an ongoing process for over a decade now. Everything seems to have an effect on a network, it’s really hard to understand, because the effect takes 3 days to stabilize, still the outcome is still not perfect.

The filter shokes really work great! From now on I will use them and use less galvanic isolators. Additional filtering in an ethernet system works without a doubt. In the past I think it was from 2018 to 2021 I have worked with 8 Pink Faun isolators in my network, which also uses small cores in each transformer. With 8 Pink Faun isolators I reached a very good sound, but It was not perfect, due to side effects. Now I use the ferrite and Iron cores, but I still have some side effects in my network. It will keep me off the forums for a while, but when I’m done I will share my findings 🙂

Maybe in the meantime I have old research which also took me a few years.

https://forum.psaudio.com/t/ethernet-cables-and-sound/11297/400?u=wijnand

Best regards,

Wijnand

[Wijnand]

No it doesn’t matter. it’s completely on another power source and ground system, but also on the same power source it has the same outcome

[Wijnand]

There is an aspect where clocks seem to interfere (sound becomes darker), while everything else is still perfect. And there’s an aspect where the sound just decreases. Less good soundstage, less natural, more unstability, less insight, etc

[Wijnand]

yes of course, but still I think we deal with two different aspects. One general noise related and one clock ‘noise’ related.

The packages are send by a clock. They have a certain pace. No matter the order they are in.

When I add a switching powersupplie to the switch. The noise problem became worse. This is audible, but the clock problem becomes better.

[Wijnand]

Hi Tobias,

Of course you are right about the digital aspect 🙂

But what if the signal itself is the carrier of data which is clocked at 25Mhz, while the dac runs at 24.576Mhz

It actually seems pretty logical that making this 25 Mhz to stable and present that it can cause trouble for the DAC clock. This also happens without the special clock switch.

The filter will smooth the blockwave in a more sinus looking wave, which has less steep edges. Think about this. Also that when I connect things that it sounds near perfect in the first hours. Normal noise would interfere at once.

It’s a pity that a forum cant show actual sound. It starts at a very high level, for streaming terms better than I ever heard. It takes a system to a complete new level. Until the dark sounds kicks in.

[Wijnand]

Good morning Paul,

What makes you think I never have replaced the clock? 🙂 To keep it short the problem also appears without the clock.

Indeed normally burned in cables would be better. Untill you will come to this certain level. It’s about the complete stream, when I only use one or two high grade power supplies then there’s no problem. This makes it better and there ain’t any issue at all, but when you take it a few levels up and use up to 8 Farad power supplies. Then the stream becomes too stable and clocks (even normal crystals) will interfere with each other, They probably will work together and the problem is born.

I’m in close contact with the guys from Pink Faun and also the electronics designer. I also had 3 different Pink Faun streamers, actually four or five when I would count some modifications. I also had an audition of different ultra streamers with different specs. They wanted to know which I thought sounded best. I was pretty clear about my preference and they have followed that advice, which means one clock in the ultra streamer sounds better than more. This is about interference.

In my own Pink Faun streamer I have tested one to 5 clocks. With only one clock it sounded better.

For what concerns the powersupplie of the Pink Faun ultra streamer it’s really on a high level. But still every Farad power supply is superior. I think the Paul Hynes just makes enough noise to prevent problems with dark sound. Feeding the option modules with the internal powersupplie of the Pink Faun streamer solves the problem as well. Also the first edition Farad power supply brought less problems.

 

[Wijnand]

When the sound became dark (DAC crystal out of phase). I think it would make sense when I would give some examples.

Adding a new linear powersupply will solve the problem for 4 days (until some level of break in is reached) . I have tried this at least 8 times 🙂

Replacing a high grade powersupply by a less good powersupply. This prevents the crystals from synchronising, so the dark sound is solved, but it’s not the same as with the better powersupply.

Replacing the ethernet cable will solve the problem for 2 hours. I have tried this more than 10 times.

Adding CAT8 makes things worse. Adding noise by adding for example a switching powersupply somewhere in the network makes things better.

Adding antennas to the ground system solves the dark sound, but on other levels is less good.

Disconnecting the audio system for 8 hours, solves dark sound for 4 hours.

Using less good power cables will solve the Dark sound (to some amount).

Using a new power cable no matter where in the system will solve the problem for 48 hours (probably a certain level of break in) I have tried this at least 20 times.

Adding powernetfilters like Kemp SNS or Furutech NCF clearline will cause dark sound

I think I can give more than 50 other examples that suits my explanation 🙂

 

 

[Wijnand]

Yes, maybe it’s that simple, but how to achieve that?

I think that the blockwave (stream) itself is already noise to a DAC

[Author]

Posts