Building a DIY Music Server

[The Computer Audiophile]

Perhaps derBen can even sell his plans so others can have them created elsewhere. I only say this because derBen doesn’t seem to want to get into this business. 

[Soul Analogue]

8 hours ago, Nenon said:

Sorry about that @Soul Analogue. I was mislead by some of your comments, such as:

 

We were talking about preference on general PSU topology and never said choke input was implemented for the ULPS for Taiko ATX

and I clarified on the next day...  could not imagine you were still misled ...

 

On 8/17/2021 at 7:55 AM, Soul Analogue said:

There is no “entirely” different design …

My suggestion of a change of higher amp choke is still valid for Nenon’s cap input supply…

my cap input design is still transformer > rectifier > c1 > choke > c2

Very similar right? Just plain more simple i would say..

i have built over a hundred of this CLC topology psu for audio purpose for myself and my friends and i never encountered the issues mentioned by Nenon.. 

 

 

[Soul Analogue]

17 hours ago, derBen said:

Hi,

 

I just finished my second DIY Audiio Server and I would like to share some impressions.

 

I'm running a Aaeon Mini-ITX board with 16GB Apacer Ram, a 250GB Samsung 980Pro M.2 SSD and a Intel i7 9700T CPU.

For the connections I use JCAT FEMTO cards: USB and NET.

The operating system is Euphony OS running ROON.

 

The server is built into two cases:

- One external PSU regulating 16V + 16V + 9V + 7V from a custom 400 VA toroid

- The main case containing the local PSU (regulating 12V + 12V + 5V + 3,3V) and the computer components

 

Both cases have the same shape and dimensions, only the button in the front is missing in the PSU case.

 

What dio you think?

Ben

 

Nice chassis design 

[austinpop]

That’s a gorgeous chassis!!

 

@derBen is in Der House!

[derBen]

Hello all,

 

I am willing to make my CAD data available to the forum community.
However, I assume that most people will not be able to do much with the data.

 

My cases are built exactly around the existing computer components.

 

Using a different mainboard would require changes to the CAD data.
Using other USB or NET cards instead of the JCAT would also require changes to the CAD data.
In addition, the PSU chamber is also designed exactly to the size of my local PSU.

 

Furthermore, some processing steps that I did myself after laser cutting are not included in the current CAD data (but I could provide this).
These are mainly grinding work and the milling of chamfers on the edges. I did that myself.

After anodising, some more processing steps were necessary:
First of all, some threads had to be cut, which should be electrically conductive (anodised aluminium is insulating).
But also blind holes, which should be drilled only 3.5 mm deep and should not go through the whole sheet.

 

The construction is not trivial and not self-explanatory.

 

If there is still interest in the data, I would be happy to provide it.

Kind regards

Ben

 

[Darryl R]

10 minutes ago, derBen said:

Hello all,

 

I am willing to make my CAD data available to the forum community.
However, I assume that most people will not be able to do much with the data.

 

My cases are built exactly around the existing computer components.

 

Using a different mainboard would require changes to the CAD data.
Using other USB or NET cards instead of the JCAT would also require changes to the CAD data.
In addition, the PSU chamber is also designed exactly to the size of my local PSU.

 

Furthermore, some processing steps that I did myself after laser cutting are not included in the current CAD data (but I could provide this).
These are mainly grinding work and the milling of chamfers on the edges. I did that myself.

After anodising, some more processing steps were necessary:
First of all, some threads had to be cut, which should be electrically conductive (anodised aluminium is insulating).
But also blind holes, which should be drilled only 3.5 mm deep and should not go through the whole sheet.

 

The construction is not trivial and not self-explanatory.

 

If there is still interest in the data, I would be happy to provide it.

Kind regards

Ben

 

 

Hi Ben,

 

If you did the edge breaks, that's a lot of additional work.  Do you have a background in machining?

[derBen]

Hi Darryl,

 

I don't have a machining background.

The milling of chamfers was done using a router mounted into a milling table (DIY) and a sharp HSS cutter.

But much more work than the chamfers was the grinding of the raw laser-cut edges...

 

Best

Ben

[guiltyboxswapper]

18 hours ago, Nenon said:

try the Mundorf HC equivalent capacitance. 

I remember you mentioned last time, but they're HUGE so would require either a far bigger PCB (= stops fitting nicely in a PC) or wires, hence went for 33,000uf 40v x5 with lovely gold PCB traces instead.  

[MarcelNL]

size matters in caps, no joke, somehow size per F seems to also correlate with how they fare sound wise IME

Not as selection criterion, just saying.

[MarcelNL]

Meanwhile: Emile responded by saying you can add caps after the ATX when you observe the max stated on their website, using the caution to beware of oscillations.

 

So while we were emailing I added another 200nF to the second EPS cable, same result and more of it, better dynamics, clearer upper midrange and cleaner treble, all in all a more open and organic sound that appears to have more 'penetration' at lower volume (if that makes sense to anyone).

 

I was listening to the new configuration and thought is sounded only slightly better, that was until I noticed I had not yet loaded the song into RAM to play as WAV from RAM....

 

 

[MarcelNL]

He also mentioned that you can fit AWG 14 in a molex pin, I was planning to give 16AWG a go but will skip that and go for 14 when the new pins arrive for the final cable build.

[Gavin1977]

Molex 46992-0610 - any cable assemblies for this (part number please?)... otherwise it's crimp you own?

[MarcelNL]

I crimp my own, it's easy enough after a few rounds of testing and very cost effective against buying pre made cables.

[michaeltay5871]

I have purchased the Saligny standard but it looks puny. Can I just go ahead and connect it to the transformer, cap and choke as shown in Nenon ULPS?  My transformer is Toroidy extreme TSAS400VA-voltage-to-50V.  Specs of the Saligny standard are :

Operates from DC to 600Hz.

AC operating voltage 6Vac to 46Vac.

DC operating voltage 9Vdc to 65Vdc.

Iq = 1,5mA.

Continuous load current up to 20A or more depending on mosfet configuration.

Over 200A pulsed current at Ta = 25Celsius ( Max Rθjc = 1.0C/W, pulse duration ≤100 μs, duty cycle ≤1%).

Require a minimum output capacitor of 1000μF

Do not support centre tapped transformer.

 

 

[elan120]

15 minutes ago, michaeltay5871 said:

Can I just go ahead and connect it to the transformer, cap and choke as shown in Nenon ULPS?  My transformer is Toroidy extreme TSAS400VA-voltage-to-50V.

Yes to your first question, but what is your secondary output voltage?

[michaeltay5871]

1 hour ago, elan120 said:

Yes to your first question, but what is your secondary output voltage?

24 VAC to give 33 VDC after rectification. 

[elan120]

4 minutes ago, michaeltay5871 said:

24 VAC to give 33 VDC after rectification. 

That should work just fine.

[michaeltay5871]

7 minutes ago, elan120 said:

That should work just fine.

Thanks elan120. 

[Tatomek7]

3 hours ago, michaeltay5871 said:

24 VAC to give 33 VDC after rectification

I have exactly the same and it works well, also transformer 400VA.

[michaeltay5871]

I found out I have ordered three 47000uf and two 22000uf Mundoff HC caps instead of the other way round. Is this still workable and I have to place a choke between a 22000uf and 47000 uf cap or is there a better configuration than this?

[elan120]

36 minutes ago, michaeltay5871 said:

I found out I have ordered three 47000uf and two 22000uf Mundoff HC caps instead of the other way round. Is this still workable and I have to place a choke between a 22000uf and 47000 uf cap or is there a better configuration than this?

If you are building the V3 version of Nenon's ULPS design, then no problem, that is what my V3 build will be, where the final portion is 47KuFx3 and all 3 wiill have bypass cap installed.

[michaeltay5871]

21 minutes ago, elan120 said:

If you are building the V3 version of Nenon's ULPS design, then no problem, that is what my V3 build will be, where the final portion is 47KuFx3 and all 3 wiill have bypass cap installed.

What about placements of the chokes?  One in between two 22kuF and other between 22kuF and 47kuF?

[elan120]

4 hours ago, michaeltay5871 said:

What about placements of the chokes?  One in between two 22kuF and other between 22kuF and 47kuF?

Yes, to form a CLCLCCC configuration.

[michaeltay5871]

1 minute ago, elan120 said:

Yes, to form a CLCLCCC configuration.

Ok thanks. 

[MarcelNL]

So the Taiko ATX arrived (took long enough), and I happen to have some linear 19V PSU lying around that served my NUC...probably best first find out how many VA that LPSU is capable of as most NUCs require only little DC grunt.