Building a DIY Music Server

[Dev]

19 hours ago, MarcelNL said:

 

and what is the preliminary verdict?

 

 

The problem is I just didn't stop at that. I added couple of more tweaks and more on the way on getting parts. I have also recently added couple of other things to my setup, including a new AM fuse, a new Westminster USB cable, speaker isolation using APs, etc. So its impossible to tell how much Saligny power alone is contribution to it but I have a feeling that it does majorly. All I can tell my system sounds sublime at this time. I saw the high current version a couple of weeks back during a recent discussion with Tiberiu on another topic but it didn't occur to me replace with the standard. Thanks to you for bringing this to the attention. I will let you know more on this when the dust settles a bit, if at all :-)

[Dev]

15 hours ago, NanoSword said:

Can you share full Pic for your power 
I can see that you are adding vcap and Saligny . can you share the improvement .

 

I am still tweaking and experimenting things. I will post a picture when things have settled down a bit. So far its sounding good with the Taiko ATX but my system as a whole has gone through a slew of changes recently as I was saying earlier. It's a bit difficult at this stage to isolate SQ on a particular upgrade but the net result is satisfying to my taste and I am not even using boutique wiring for most parts, except some strategic places. I plan to upgrade to the Mundrof angelique at some point.

[Downtheline]

Hey all, I wanted to share some experiences from the last week. I just bought a used Extreme server and discontinued my DIY dual cpu server project. Life has gotten too busy with work and a little one running around. 

 

I got my Extreme about a week ago and my brain is still burning in while the Extreme was already well burned in. I am comparing it to my diy "mini-extreme" with a single xeon 4210 cpu in an hdplex case, 2 sticks of 8gb ram, jcat usb xe, intel ssd drives on pcie card, powered by dual rail paul hynes sr7t to taiko dc/atx and to the motherboard. Usb powered by farad super3. Using gentooplayer (still) via USB flash drive, mconnect on phone.  

 

So far I am liking what I hear with the Extreme, but I am having to make adjustments. There is a lot more bass and body, at first it was overwhelming until I turned off my sub, then much better. I think that streaming from qobuz is actually otherwise quite close in quality using TAS on the Extreme compared to my old server.

 

Roon with whatever build is on the Extreme is kind of annoying and fatiguing to listen to. I really have taken some steps back in my network recently (due to a mouse) and I hope to improve this soon. I believe this to be the reason that things are close between gentooplayer and TAS when streaming. 

 

On local files, my old server is still enjoyable. This says alot because often when there are improvements, what I previously had been accustomed to I can't tolerate any longer. However the Extreme really sets itself apart here with a much more palpable and realistic presentation. There is cleaner imaging, further wrapping of sound stage. Transients are sharper. I'm really drawn in. High resolution tracks are mind blowing! 

 

So in all, I won't say my sophomore attempt at a diy server was embarrassed, especially given it's much lower cost. I know it's no dual cpu server with hqplayer and project lasso etc, but not too bad. I am done though with Linux based audio systems however. I just never enjoyed that sort of tweaking. 

 

My system is decent I'd say, could probably use some targeted improvements.  Usb cable is Sablon 2020 (nenon's old cable), dac is Holo Audio May KTE, pre is Don Sachs (latest version), amp is First Watt F8, speakers are Spatial Audio X5s, sub is Rythmik F15. Interconnects are Iconoclast tpc, speaker cables are Iconoclast sptpc. Power cables are all Iconoclast BAV, into an Uberbuss conditioner. Resonance isolation from mix of rollerblock jrs and nordost sort cones on bamboo cutting boards or the Taiko platform.  Any comments on what to improve will be considered! 

[acg]

@Downthelinethanks for the write up.

 

Very interesting considering one of my dacs is a May KTE and I consider it as source immune as anything out there.   All that extra bass sounds like it is a big change to frequency response so I really do wonder the cause. 

 

I would start there with any system changes,  try to figure out why two decent servers sound so different into a very well designed dac.  Are you listening NOS on the May.  Grounding changes or perhaps irregularities with the diy server?

[bit01]

@Downtheline - I find the JCAT USB sensitive to its power supply. You can change levels of PRaT, body of the music, bass level etc by changing how you power it. I can change the bass level from light to overwhelming by changing the power cord to its LPS. I am not familiar with the Farad but if you have experimented and you may already know this.

My current favorite LPS is described in the link below (I have a heavy gauge DIY Furutech PC on it and a DIY DC lead with the same Sablon cable on the JCAT) :

 

 

[Exocer]

3 hours ago, bit01 said:

@Downtheline - I find the JCAT USB sensitive to its power supply. You can change levels of PRaT, body of the music, bass level etc by changing how you power it. I can change the bass level from light to overwhelming by changing the power cord to its LPS. I am not familiar with the Farad but if you have experimented and you may already know this.

My current favorite LPS is described in the link below (I have a heavy gauge DIY Furutech PC on it and a DIY DC lead with the same Sablon cable on the JCAT) :

 

 

Really cool power supply. Will have to compare this with the DC4 i’m using. Super reasonably priced.

[Downtheline]

6 hours ago, acg said:

NOS on the May.

Yes NOS on the May.

[Downtheline]

2 minutes ago, Exocer said:

My current favorite LPS is described in the link below (I have a heavy gauge DIY Furutech PC on it and a DIY DC lead with the same Sablon cable on the JCAT) :

I'm also using a diy DC lead, made with mundorf silver gold wire. I think the Farad is decent, but I haven't experimented with others actually. I have an Uptone js2 on hand. And a Keces p3 I could play around with though. 

[Exocer]

7 minutes ago, Downtheline said:

I'm also using a diy DC lead, made with mundorf silver gold wire. I think the Farad is decent, but I haven't experimented with others actually. I have an Uptone js2 on hand. And a Keces p3 I could play around with though. 

I use mostly SG DIY leads everywhere, except for the Opto USB, where I use Mundorf Angelique (ran out of SG).

[bit01]

2 hours ago, Exocer said:

Really cool power supply. Will have to compare this with the DC4 i’m using. Super reasonably priced.

Very opposite ends of the price scale, especially if you find a used one on ebay (should be some as it is an old widely used industrial design)! 

Please report back when/if you ever do compare.

 

ATB

b.

[bit01]

I currently use this one LPS->JCAT 

 

[Nenon]

@Downtheline - congrats on your Extreme. Did you buy Jay's Extreme? 

What you are describing is similar to what I heard in my system. My Extreme got a lot better over time. Yours may get better too. You may want to contact Taiko (if you haven't) to go over all the settings and make sure the previous owner did not mess something up. 

[Nenon]

10 hours ago, acg said:

Very interesting considering one of my dacs is a May KTE and I consider it as source immune as anything out there.   All that extra bass sounds like it is a big change to frequency response so I really do wonder the cause. 

 

I would start there with any system changes,  try to figure out why two decent servers sound so different into a very well designed dac.  Are you listening NOS on the May.  Grounding changes or perhaps irregularities with the diy server?

 

Unfortunately there is no such thing as immune DACs :(. I would have traveled the world the last few years (if DACs were immune) for the time I have spent on DIY servers :). Even the smallest changes to servers in a good resolving system are quite audible. His two servers are quite different and expected to sound different. 

[Exocer]

1 hour ago, bit01 said:

Very opposite ends of the price scale, especially if you find a used one on ebay (should be some as it is an old widely used industrial design)! 

Please report back when/if you ever do compare.

 

ATB

b.

I couldn't resist at that price. I bought one pre-owned from Ebay for ~$25. I happen to have a small Galaxy 2U chassis laying around unused, already cut out for FI-09 IEC inlet with two GX-16 cutouts. Will need to pick up a fuse holder and IEC. I've got the wire on hand as well. I am really happy with what I have but when cheap solutions like this come along I have to hear how they sound. Count me in! Once I get everything built up, I will run some tests and post feedback here. 

 

Aside from that, you're 100% spot on with your description of how the JCAT USB XE is very much reactive to the power you feed it. It is the card that ultimately made me understand that I had to move on from my beloved MPAudio ALS-HPULN, which sounded very very good if you don't get to hear it next to a DC4 in the same exact chain.

 

[Downtheline]

32 minutes ago, Nenon said:

Did you buy Jay's Extreme? 

Nope Anil's from Whatsbest forum. Jay's looks nice though. I actually wanted a black one but his wasn't yet available when I got mine! 

[MarcelNL]

4 minutes ago, Downtheline said:

Nope Anil's from Whatsbest forum. Jay's looks nice though. I actually wanted a black one but his wasn't yet available when I got mine! 

Just out of sheer curiosity, what does one buy when selling an Extreme?

 

[Downtheline]

31 minutes ago, MarcelNL said:

Just out of sheer curiosity, what does one buy when selling an Extreme?

Some reports of a new technology involving diamonds, hardwoods, pulleys... and vinyl. I guess. 

[MarcelNL]

O jeeezzz, so you are saying that once I go mono when the RCA 70A arrives I am done with digital?

[bit01]

1 hour ago, Exocer said:

I couldn't resist at that price. I bought one pre-owned from Ebay for ~$25. I happen to have a small Galaxy 2U chassis laying around unused, already cut out for FI-09 IEC inlet with two GX-16 cutouts. Will need to pick up a fuse holder and IEC. I've got the wire on hand as well. I am really happy with what I have but when cheap solutions like this come along I have to hear how they sound. Count me in! Once I get everything built up, I will run some tests and post feedback here. 

 

Aside from that, you're 100% spot on with your description of how the JCAT USB XE is very much reactive to the power you feed it. It is the card that ultimately made me understand that I had to move on from my beloved MPAudio ALS-HPULN, which sounded very very good if you don't get to hear it next to a DC4 in the same exact chain.

 

Price sounds about right! The frame acts as the power transistor heat sink and has pem nuts in it on the short side so you can mount it on the chassis if you chose too.  It does get warm powering the USB XE (1.5A?) so that would aid cooling. They are not efficient (45%). Looking forward to your experience with it.

 

ATB

b.

[acg]

11 hours ago, Nenon said:

 

Unfortunately there is no such thing as immune DACs :(. I would have traveled the world the last few years (if DACs were immune) for the time I have spent on DIY servers :). Even the smallest changes to servers in a good resolving system are quite audible. His two servers are quite different and expected to sound different. 

 

I'm not really arguing that the server makes no difference,  rather that it seems highly unusual that it should make that sort of gross frequency response change in the bass.   That sort of change indicates that something is/was broken.   A while back I trialled a number of I2S CD transports into the May KTE and thought I could easily tell which was which until I set it up so that I did not know which I was listening to, and surprise surprise I could not reliably pick them even after a few weeks acclimatisation. 

 

You may say that my system is not resolving enough... but that is highly unlikely. 

 

Granted, that was not USB and it was not the Extreme, which I would love to hear in my system, but if I got one and had to turn off my subs due to a server change alarm bells would be ringing.

[LTG2010]

49 minutes ago, acg said:

rather that it seems highly unusual that it should make that sort of gross frequency response change in the bass

Windows has a much fuller/warmer balance than Linux. If using Linux and you have tweaked your equipment to balance out some of that Linux brightness, then switching to a Windows machine (Extreme) would go the other way. Especially if using a Sub woofer, some of that low bass  appearing will be exagerated.

[acg]

@LTG2010I've never really gone into tweaking Linux but one of my digital sources is a simple RPi running roipee and I would never call it bright.  However W7 sounds different to W8 sounds different to W10 to W11, no doubts there.

 

Warmth is really the upperbass region,  say 250Hz to 500Hz which should be at least an octave above anything the subs are offering.  It is the most important region to get right as it sets the foundation for the whole sound of the playback.

 

Anyway,  I don't understand how the change can be so great unless something else is wrong.   Maybe it is a tweaked Linux as you suggest that has caused a playback to be setup to accommodate...I dunno... unlikely... but if so similar in scale results (i.e. the subs becoming redundant) may also have been had with any old windows pc. 

[MarcelNL]

I owuld not doubt the difference, or relate it to something being wrong...IME getting phase aligned or timing correct has a large impact on how our brain suddenly is freed up from doing sound processing allowing us to hear all notes much more clearly and that often has a large impact on bass.

[Downtheline]

Yeah, I think phase is the issue. I turned my sub crossover further down, had been around 50, now down around 35. Much better now.

[ASRMichael]

On 7/7/2022 at 6:06 PM, Gavin1977 said:

Micro-Audio SMPS440-TK3 and Micra p5 up and running.  Been a long time coming - I am super cautious to ensure that everything boots without fault (and sparks)!

 

I am already hear improved clarity and detail retrieval.  200hr break-in now for the Mundorf caps before any critical listening.

 

 

 

 

 

(Yes long leads I know, but design choice limits me to this).

4 days to shoot out! Looking forward to hearing @Gavin1977 PSU! 👍👍