Jump to content
IGNORED

Building a DIY Music Server


Recommended Posts

10 minutes ago, ASRMichael said:

Do you know that for sure? Have you had a peak under the capacitors? 

Quite sure, from reading the mosterthread on the Extreme at WBF and the fact the ATX only was made available MUCH later, not sure I want to reread all in that thread though ;-)

ISP, glass to Fritz!box 5530, another Fritz!box 5530 for audio only in bridged mode on LPS, cat8.1, Zyxel switch on LPS, Finisar <1475BTL>Solarflare X2522-25G, external wifi AP, AMD 9 16 core, passive cooling ,Aorus Master x570, LPSU with Taiko ATX, 8Gb Apacer RAM, femto SSD on LPS, Pink Faun I2S ultra OCXO on akiko LPS, home grown RJ45 I2S cable, Metrum Adagio DAC3, RCA 70-A and Miyaima Zero for mono, G2 PL519 tube amps. 

Link to comment
30 minutes ago, MarcelNL said:

Quite sure, from reading the mosterthread on the Extreme at WBF and the fact the ATX only was made available MUCH later, not sure I want to reread all in that thread though ;-)

I’ve read that too…Emile has been asked on several occasions what’s under there. He said he didn’t want to disclose. That’s after the ATX came out to Diy. This make me think it has the ATX. 

Link to comment

I read something like that knowing what they now they would never go for a full linear regulated PSU for the Extreme. 

Anyway I totally get that they do not disclose all their tricks and tweaks, it;s pretty amazing what he is sharing already!

 

ATX or not, I do not see much merit in making clones, following design principles is one thing but there is a need to listen to stages of development IMO...many crossings where things can go horribly wrong, and they usually do because there are too many factors at play.

ISP, glass to Fritz!box 5530, another Fritz!box 5530 for audio only in bridged mode on LPS, cat8.1, Zyxel switch on LPS, Finisar <1475BTL>Solarflare X2522-25G, external wifi AP, AMD 9 16 core, passive cooling ,Aorus Master x570, LPSU with Taiko ATX, 8Gb Apacer RAM, femto SSD on LPS, Pink Faun I2S ultra OCXO on akiko LPS, home grown RJ45 I2S cable, Metrum Adagio DAC3, RCA 70-A and Miyaima Zero for mono, G2 PL519 tube amps. 

Link to comment
1 hour ago, RickyV said:


So I looked a little closer and I now think I see 4 connecting wires.

 

BED6E3F0-1527-47CA-8EFC-4037016FA867.jpeg.15cf9b19f7b6999620f425d9e2fde601.jpeg

 

red for coil 1. Yellow for coil 2, see only one wire.

 

AD6FC2E1-E472-400A-BE47-7768D464A68A.thumb.jpeg.3e6e0426348145aa9bbf65c1c0ba6915.jpeg

 

So @NEMO777, not sure if this is helpful or makes things more confusing.

 

In my opinion it is pretty necessary to know the values of al these coil if you want to duplicate this power supply.

yes, I now also saw that the wires go to two coils of the drosel, apparently it is switched on in three-phase mode. now the question arises in which part of the scheme it is included. yes, taiko is increasingly puzzling me.

Link to comment

image.png.325b1960c783218e558e623a52655037.png

The big Lundhal Chokes are dual. Theres a number of ways to connect up but it might be like the third example: Serial connection for improved common mode rejection. eg. the first 47uf cap and bpass just after the rectifier circuit. 2 chokes on +ve and return path completely decoupling capacitor bank which would follow eg, your 160uf or 720 uf depending what supply ou are making. Its easy to try it this way and compare to 2nd choke spliting the second capacitor bank. I mean like below, but one dual choke eg Lundhal or 2 single hammonds etc.

image.jpeg.a147982b6ab799e9e9a252f9be298f28.jpeg

 

Link to comment
27 minutes ago, LTG2010 said:

image.png.325b1960c783218e558e623a52655037.png

The big Lundhal Chokes are dual. Theres a number of ways to connect up but it might be like the third example: Serial connection for improved common mode rejection. eg. the first 47uf cap and bpass just after the rectifier circuit. 2 chokes on +ve and return path completely decoupling capacitor bank which would follow eg, your 160uf or 720 uf depending what supply ou are making. Its easy to try it this way and compare to 2nd choke spliting the second capacitor bank. I mean like below, but one dual choke eg Lundhal or 2 single hammonds etc.

image.jpeg.a147982b6ab799e9e9a252f9be298f28.jpeg

 


I think it looks like that, so it improve common noise rejection before enter the dc atx

Link to comment

someone (same guy I see now) put a comparison across many PSUs online, in the end it will also make a large difference for what purpose a PSU is used....

ISP, glass to Fritz!box 5530, another Fritz!box 5530 for audio only in bridged mode on LPS, cat8.1, Zyxel switch on LPS, Finisar <1475BTL>Solarflare X2522-25G, external wifi AP, AMD 9 16 core, passive cooling ,Aorus Master x570, LPSU with Taiko ATX, 8Gb Apacer RAM, femto SSD on LPS, Pink Faun I2S ultra OCXO on akiko LPS, home grown RJ45 I2S cable, Metrum Adagio DAC3, RCA 70-A and Miyaima Zero for mono, G2 PL519 tube amps. 

Link to comment

the transformation is ridiculous, how separation between instruments and voices is improved, the timbre of each note more clear, cleaner, more nuances..it's probably going to be a very long night!

 

ISP, glass to Fritz!box 5530, another Fritz!box 5530 for audio only in bridged mode on LPS, cat8.1, Zyxel switch on LPS, Finisar <1475BTL>Solarflare X2522-25G, external wifi AP, AMD 9 16 core, passive cooling ,Aorus Master x570, LPSU with Taiko ATX, 8Gb Apacer RAM, femto SSD on LPS, Pink Faun I2S ultra OCXO on akiko LPS, home grown RJ45 I2S cable, Metrum Adagio DAC3, RCA 70-A and Miyaima Zero for mono, G2 PL519 tube amps. 

Link to comment
On 9/11/2021 at 9:56 PM, Tatang said:


I think it looks like that, so it improve common noise rejection before enter the dc atx

Yes...the princple of common mode choke sounds like very good for noise suppression... it also avoid saturation of the core....so that it can work at higher current...

HOWEVER, there were several times that I tried common mode choke in my power supply...i never liked it.... no matter at digital, analog and DHT filament....

it sounded like the connection to emotion of the music had lost

Revert back to differential mode choke, everything just came back

So differential mode choke for me all the way =)

Builder of Linear Power Supplies

Link to comment
35 minutes ago, Soul Analogue said:

Yes...the princple of common mode choke sounds like very good for noise suppression... it also avoid saturation of the core....so that it can work at higher current...

HOWEVER, there were several times that I tried common mode choke in my power supply...i never liked it.... no matter at digital, analog and DHT filament....

it sounded like the connection to emotion of the music had lost

Revert back to differential mode choke, everything just came back

So differential mode choke for me all the way =)

 

I find it interesting how peoples experiences differ.  For instance I will only power a DHT filament one of two ways:  raw AC; or high quallity CCS followed by a high quality common mode choke.  Both methods have worked very well for 2.5V and 4V DHT's in power amps and the 7.5V of 10/801a in preamps.  I've not been happy with differential mode chokes in these posistions, nor more run-of-the-mill DC/CCS in general.

 

What I have experienced with DHT's though is that if you are going to have a super low noise filament supply the remainder of the circuit and in particular how components are laid out makes a difference to the sound...it needs to be very good and not at all cramped.

 

Link to comment

Thanks to the DIY community for contributing to the discussion on the original Taiko Audio SGM Extreme unregulated power supply circuit. I am especially grateful for the personal messages from the participants who helped me with information on the real denominations of the components and the scheme for their inclusion. In my opinion, the circuit has original solutions for the parts used and the circuit for their inclusion. I will definitely try to collect it completely, as well as various options for including individual elements in order to influence the final sound. Now I am waiting for all the components of the server to come to me in order to assemble it and start experimenting with the power supply. As distinguished members of the community noted that it is better to look for your own path to implementation, I think that at this stage my path is to start from the already implemented model, and on its basis find the best solution for myself.

DIY-Taico-2.jpeg

Link to comment

Having carefully studied the available sources for Taiko Audio SGM Extreme, I noticed that the rectifier circuit no doubt uses dual Schottky diodes. There are 2 of them with a common anode and a common cathode for voltage rectification. But in one of the photos, I saw that Taiko uses 4 elements. Does the community have an explanation of how Taiko uses them?

1388342585_TaikoAudioSGMExtremeInsideLundahlMundorfPanzerholtz1.thumb.jpg.373f7d37ab1375585b978edbf61eeafa (3).jpg

WeChat 圖片_20191007212418 (2).jpg

WeChat 圖片_20191007212418.jpg

Link to comment
9 hours ago, AudioDoctor said:

So, this might be another dumb question, but...

 

How in the heck are you guys installing an OS on your headless servers?

 

My MB needs a graphics card at boot anyway, I connect a monitor and keyboard/mouse to it when I need to tinker in the BIOS or install Daphile for the first time. After that I control everything from the laptop, I can install updates or even transfer the OS, prepare a new SSD-install OS and change the settings. 

Daphile switches off the graphics card after a minute or two. 

ISP, glass to Fritz!box 5530, another Fritz!box 5530 for audio only in bridged mode on LPS, cat8.1, Zyxel switch on LPS, Finisar <1475BTL>Solarflare X2522-25G, external wifi AP, AMD 9 16 core, passive cooling ,Aorus Master x570, LPSU with Taiko ATX, 8Gb Apacer RAM, femto SSD on LPS, Pink Faun I2S ultra OCXO on akiko LPS, home grown RJ45 I2S cable, Metrum Adagio DAC3, RCA 70-A and Miyaima Zero for mono, G2 PL519 tube amps. 

Link to comment
1 hour ago, Nenon said:

 

@NEMO777 if my comments that you are doing it all wrong were not enough, Emile from Taiko, who is checking this thread occasionally, saw your posts and reached out to me this morning. Here is what he said:

“That is not the Extreme power supply schematic. If you are undertaking the effort and invest considerable funds to build a ULPS for the DIY Taiko DC-DC ATX, you will have far better results using a 24V AC transformer secondary with 32-33V DC out. If you want to use larger value chokes, or chokes with a lower DCR, do simulate and adjust capacitance value to match.”

 

I tried telling you the same thing in my post, which I highly recommend you re-read:

To repeat again - I believe you are wrongfully assuming that the Extreme is using the DIY Taiko ATX and doing things that will not sound so good with the DIY Taiko ATX. 

 

We used the opportunity to discuss different chokes in this unregulated LPS design. For those who want to play with the Hammond 195J10 or the even bigger Hammond 195K12 choke, here is what you need to do:

  • 22000uF - Hammond 195J10 - 22000uF - Hammond 195J10 - 220000uF; or
  • 22000uF - Hammond 195K12 - 22000uF - Hammond 195K12 - 330000uF

These designs will sound tighter and a bit more mechanical than the v2 and v3 I have shared. These larger values chokes will have a sound signature more typically associated with “digital sound” when used with the DIY DC-DC ATX. The difference in sound between these two versions is very minimal. But the Hammond 195K12 is very big and won't fit in the upcoming DIY chassis. You will need an external chassis for that. 

 

My recommendation still remains the V2 or the V3, but if you want to play with different chokes, at least make sure you have the right capacitance for the choke you are using. 

Thank you Nenon for your comment. English is not my first language. It is difficult for me to convey all the nuances of my thought and idea through machine translation. I'm not going to blindly copy the Taiko circuit, it is important for me to take the initial concept and, on its basis, conduct a series of experiments, find that unique combination of physical power parameters such as voltage, capacity and number of capacitors, shunt elements, and more. The power supply is just one piece of the puzzle of a good audio server. An extremely important element is software customization. I really hope that you, as you promised us, will share the intricacies of setting up a two-processor server in due time.
I want to ask you, can you have any thoughts on how the dual Schottky diodes are included in the Taiko circuit?

Link to comment

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now



×
×
  • Create New...